Search

Smartphones are a vast category of consumer electronic products. Design For Disassembly (DFD) is a methodology that can effectively reduce the disassembly, recycling, and maintenance costs of smartphones. However, due to the diverse constraints posed by the connections among smartphone components, the variability of disassembling tools makes it challenging to objectively characterize the disassemblability of the smartphone. Therefore, information entropy is introduced to characterize the complex state of the system. Disassemblability can be expressed by calculating the operating time of the disassembling tool through information entropy. After multiplying by a constrained quantity factor, the Improved Disassembling Tool Entropy (IDTE) responds to changes in disassemblability as the structural level changes. According to the structure and recycling direction, the disassembling level of smartphones can be divided into module level, part level, and hybrid part-module level. Based on the Maynard Operational Sequencing Technique (MOST), the basic unit of operating time of the disassembling tool is calculated. For the hybrid selective Disassembly Sequence Planning (DSP) of the part and module levels, the Improved Double Genetic Algorithm (IDGA) model is established to compute the optimal disassembly sequence corresponding to each program level. This model calculates the IDTE corresponding to the optimal disassembly sequences at each disassembling level. The validity of IDTE was verified by a coupled comparison test between IDTE and theoretical disassembly time. Finally, an analysis of the disassemblability variations was conducted for two different models of smartphones based on their structure. [ABSTRACT FROM AUTHOR]

For hydrogen fuel cell vehicles, energy management strategies (EMS) are vital for balancing fuel cell and battery power, limiting fuel cell power, maintaining state of charge (SOC) fluctuation range and mitigating degradation. Reinforcement learning-based EMS, especially using deep Q-network (DQN) and deep deterministic policy gradient (DDPG), have demonstrated potential for enhancing the fuel economy of hybrid electric vehicles (HEV) and fuel cell electric vehicles (FCEV). This research proposes mutation protection DQN (MPD)-based EMS to improve hydrogen fuel economy and reduce fuel cell degradation under driving cycles with plenty of mutations. By quantifying mutation, exploring its relationship with driving conditions and integrating a mutation protection module with DQN, MPD-based EMS achieves approximately 11% and 6% better fuel economy compared to the other two learning-based EMS. Additionally, it also reduces fuel cell degradation by approximately 21% and 13%. • Quantifying the mutation of driving conditions through variance. • Utilizing information entropy to process mutations. • Introducing a dynamic mutation threshold value. • Designing mutation protection module and integrating it into the DQN. [ABSTRACT FROM AUTHOR]

An efficient multi-level design optimization method is developed in this research based on information entropy to select the configurations with high potential to achieve the optimal design solution. The generic design is modeled by an AND-OR tree based on design requirements. A node in the AND-OR tree is used to describe partial design solution such as a component or an assembly with design parameters. The different design configurations are created from the AND-OR tree through tree-based search, and each design configuration is modeled by design parameters. Optimization is conducted at two levels. Parameter optimization is carried out to identify the optimal parameter values for each design configuration, while configuration optimization is carried out to identify the optimal configuration. Information entropy is employed to evaluate the partial configuration candidates modeled as branches in the AND-OR tree to eliminate the branches that are unlikely to lead to the optimal solution. [ABSTRACT FROM AUTHOR]

Aiming at the problem of magnetic anomaly detection under low signal-to-noise ratio, a full magnetic gradient detection via minimum entropy and gradient orthonormal basis function is proposed in this paper. Firstly, the mathematical model of magnetic dipole gradient tensor and the experimental method of collecting magnetic gradient tensor signals are introduced. Secondly, a new minimum entropy detector is designed, the basic functions which are processed by new minimum entropy detector can be orthogonalized easily. Thirdly, a group of new standard orthogonal basis functions are constructed according to the characteristics of magnetic gradient anomaly signals. Finally, the detector and orthogonal basis functions are used to detect the magnetic anomaly. Both numerical simulation signals and experiment signals are used to prove the effectiveness of the methods. • A new magnetic gradient minimum entropy detector is designed on the orthonormal basic function. • A group of new standard orthogonal basis functions are constructed on the characteristics of magnetic gradient anomaly signals. • The detector and orthogonal basis functions are used to detect the magnetic anomaly. [ABSTRACT FROM AUTHOR]

Adaptable design is an increasingly utilized design paradigm that can effectively and efficiently create a new design by adapting an existing design to maintain, upgrade and extend the functionality of a product by enhancing its adaptability. This paper presents an approach integrating soft sets and entropy theories to develop a mapping model between design requirements and design solutions to evaluate the adaptability when the design or product has insufficient design information. The design adaptabilities of cylindrical gear machine tools were used to validate the feasibility of the proposed approach and a modified design with improved adaptability is realized. [ABSTRACT FROM AUTHOR]

Cloud computing offers a variety of on-demand services to users and has gained significant prominence in the contemporary era. The security of information stored in cloud data centers has become a central concern, especially for sensitive data like medical images, videos, and multimedia content that require heightened protection when stored in data centers. Cloud users have a responsibility to ensure the security of their data through strategic measures. Focusing on maintaining the privacy of images stored in the cloud, this research introduces an innovative image encryption technique that is based on a modified Skew Tent chaotic map. The suggested modification to the chaotic map includes combining the Skew Tent map with both the sine trigonometric function and perturbation technology. This results in enhanced randomness, more intricate dynamical behavior, a broader chaotic range, and increased sensitivity to initial values in contrast to alternative chaotic maps. The incorporation of this adapted map into a stepwise procedure, involving two rounds of permutation followed by diffusion, effectively accomplishes image data encryption for cloud storage systems. These consecutive operations collectively enhance the encryption method's randomness and robustness. Through simulations conducted using Cloudsim, the cipher-image exhibits a uniform distribution and achieves a commendable information entropy score of 7.996749. The encryption algorithm significantly reduces correlation coefficients from 1 in the original image to 0 in the encrypted image, while maintaining NPCR (Number of Pixel Change Rate) and UACI (Unified Average Changing Intensity) values within the critical range. Additionally, both theoretical analysis and practical evaluations confirm the algorithm's resilience against exhaustive, occlusion, and classical attacks. Moreover, extending this encryption framework to video data, a novel video encryption approach is proposed. • Novel image encryption with modified Skew Tent map for cloud privacy. • Method merges Skew Tent map, sine function, and perturbation technology. • The technique boosts chaos, complexity, range, and initial value sensitivity. • Cloudsim simulations show cipher-image uniformity & high entropy of 7.996749. • Encryption lowers correlation from 1 to 0, keeps NPCR and UACI in critical range. [ABSTRACT FROM AUTHOR]

Developing a new spectrometric analytical method based on a fingerprinting approach requires optimisation of the experimental stage, particularly with novel instruments like benchtop low-field NMR spectrometers. To ensure high-quality LF-NMR spectra before developing the multivariate model, an experimental design to optimise instrument conditions is essential. However, difficult-to-control factors may be critical for optimisation. Taguchi methodology addresses these factors to obtain a system robust to variation. This study uses the Taguchi methodology to optimise instrument settings for acquiring high-quality 1H and 13C LF-NMR signals in a short time from virgin olive oil (VOO). Two experimental trials (for 1H and 13C signals, respectively) were carried out and analysed to find an optimal and robust combination of instrument settings against changes in two difficult-to-control factors: ambient temperature and small deviations of the NMR tube volume (700 ± 50 μL). The responses to be optimised, run time and spectral information quality, were analysed separately and jointly, as some factors showed opposite behaviour in the effect on the responses. Multiple response analysis based on suitable desirability functions yielded a combination of factors resulting in desirability values above 0.8 for 1H LF-NMR signals and almost 1.0 for 13C LF-NMR signals. In addition, a novel approach to assess the information quality of an analytical signal was proposed, addressing a major challenge in analytical chemistry. By applying information theory and calculating information entropy, this approach demonstrated its potential for selecting the highest quality (i.e. most informative) analytical signals. The acquisition instrument conditions of LF-NMR were successfully optimised using Taguchi methodology to acquire highly informative 1H and 13C spectra in a minimum run time. The importance lies in the future development of non-targeted analytical applications for VOO quality control. In addition, the innovative use of information entropy to a priori assess the signal quality represents a significant advance and proposes a solution to a long-standing challenge in analytical chemistry. [Display omitted] • Instrument settings for 1H and 13C LF-NMR signals acquisition were optimised. • Taguchi experimental designs were performed to optimise a robust system. • A novel proposal to a priori assess the information quality of an analytical signal is presented. • Information theory was applied to select the most informative fingerprint. [ABSTRACT FROM AUTHOR]

[Display omitted] • This paper uses NIR spectroscopy combined with intelligent analysis algorithms to improve the accuracy of similarity measurement. • The method effectively extracts sample information of NIR spectra in high-dimensional space, retaining the original data features. • The Sinkhorn distance is introduced to improve the t-SNE algorithm, which can focus more on the global probability distributions. • To mitigate multi-attribute impacts on similarity measure, a multi-attribute distance matrix is constructed by information entropy. • The method can assist in the maintenance and design of the product formula. Due to the high-dimensionality, redundancy, and non-linearity of the near-infrared (NIR) spectra data, as well as the influence of attributes such as producing area and grade of the sample, which can all affect the similarity measure between samples. This paper proposed a t-distributed stochastic neighbor embedding algorithm based on Sinkhorn distance (St-SNE) combined with multi-attribute data information. Firstly, the Sinkhorn distance was introduced which can solve problems such as KL divergence asymmetry and sparse data distribution in high-dimensional space, thereby constructing probability distributions that make low-dimensional space similar to high-dimensional space. In addition, to address the impact of multi-attribute features of samples on similarity measure, a multi-attribute distance matrix was constructed using information entropy, and then combined with the numerical matrix of spectral data to obtain a mixed data matrix. In order to validate the effectiveness of the St-SNE algorithm, dimensionality reduction projection was performed on NIR spectral data and compared with PCA, LPP, and t-SNE algorithms. The results demonstrated that the St-SNE algorithm effectively distinguishes samples with different attribute information, and produced more distinct projection boundaries of sample category in low-dimensional space. Then we tested the classification performance of St-SNE for different attributes by using the tobacco and mango datasets, and compared it with LPP, t-SNE, UMAP, and Fisher t-SNE algorithms. The results showed that St-SNE algorithm had the highest classification accuracy for different attributes. Finally, we compared the results of searching the most similar sample with the target tobacco for cigarette formulas, and experiments showed that the St-SNE had the highest consistency with the recommendation of the experts than that of the other algorithms. It can provide strong support for the maintenance and design of the product formula. [ABSTRACT FROM AUTHOR]

Decision tree algorithms have gained widespread acceptance in machine learning, with the central challenge lying in devising an optimal splitting strategy for node sample subspaces. In the context of continuous data, conventional approaches typically involve fuzzifying data or adopting a dichotomous scheme akin to the CART tree. Nevertheless, fuzzifying continuous features often entails information loss, whereas the dichotomous approach can generate an excessive number of classification rules, potentially leading to overfitting. To address these limitations, this study introduces an adaptive growth decision tree framework, termed the fuzzy neighborhood decision tree (FNDT). Initially, we establish a fuzzy neighborhood decision model by leveraging the concept of fuzzy inclusion degree. Furthermore, we delve into the topological structure of misclassified samples under the proposed decision model, providing a theoretical foundation for the construction of FNDT. Subsequently, we utilize conditional information entropy to sift through original features, prioritizing those that offer the maximum information gain for decision tree nodes. By leveraging the conditional decision partitions derived from the fuzzy neighborhood decision model, we achieve an adaptive splitting method for optimal features, culminating in an adaptive growth decision tree algorithm that relies solely on the inherent structure of real-valued data. Experimental evaluations reveal that, compared with advanced decision tree algorithms, FNDT exhibits a simple tree structure, stronger generalization capabilities, and superior performance in classifying continuous data. • Existing methods often split node samples either by preset fuzzy sets or dichotomy. • We propose a novel node-cutting method which only depends on data structure itself. • We introduce a fuzzy neighborhood decision model for constructing adaptive decision trees. • The proposed decision tree has a minimal rule set and robust generalization ability. [ABSTRACT FROM AUTHOR]

• Crucial time exists in the accumulation of monitoring data for Bayesian inversion. • Section number and location impact crucial time, but frequency and error level do not. • Relative crucial time determined by Peclet number and mainly controlled by dispersion. • Spatial structure of crucial time uncovered and explained by information entropy theory. • Novel design method on emergency monitoring largely improve PSI practicality. The Pollution source identification (PSI) is an important issue on river water quality management especially for urban receiving water. Numerical inversion method is theoretically an effective PSI technique, which employs monitored downstream pollutant breakthrough curves to identify the pollution source. In practice, it is important to know how much monitoring data should be accumulated to provide PSI results with acceptable accuracy and uncertainty. However, no literature reports on this key point and it seriously handers the numerical PSI technology to mature practical applications. To seek a monitoring guideline for PSI, we conducted extensively numerical experiments for single-point source instantaneous release taking Bayesian-MCMC method as the baseline inversion technique. The crucial time (T c) phenomenon was found during the data accumulation process for Bayesian source inversion. After T c , estimated source parameters subsequent sustained low error levels and uncertainty convergence. Results shown the presence of T c impacted by the number and location of monitoring sections, while monitoring frequency and data error do not. Under different river hydrodynamic conditions, relative crucial time (Λ) is determined by the river's Peclet number, and minimum effective Λ was controlled by dispersion coefficient (D x). Analytic spatial structure of Λ(U, D x) was uncovered and this relationship successfully explained by the information entropy theory. Based on these findings, a novel design method of PSI emergency monitoring network for preparedness plan and a practical framework of PSI for emergency response were established. These findings fill the important knowledge gap in PSI applications and the guidelines provide valuable references for river water quality management. [Display omitted] [ABSTRACT FROM AUTHOR]

• A fast real-time method for measurement of gas holdup characteristics with image information entropy. • The global and local profiles of gas holdup could be collected simultaneously. • Sieve plate resistance and bubbling state were analyzed with the use of image information entropy. Wet scrubbers are often used in various types of industrial workshops to separate dust particles from the air. Non-uniform gas–liquid flow patterns above the sieve plate can result in significant losses in purification efficiency for particles. To improve the flow pattern via sieve plate design and to more accurately predict purification efficiency, a better understanding of the gas holdup is required. Various methods have been developed to measure and analyze the gas holdup from local and global perspectives. However, for guiding the operation and design of reactors, it is obviously insufficient to obtain the phase holdup characteristics under a steady state. The spatial and temporal characteristics of gas holdup in a wet scrubber are rather difficult to measure in real time with the current techniques. In this study, a fast real-time method for measurement of gas holdup characteristics was proposed, based on image information entropy analysis. The axial and radial profiles of gas holdup under different bubble washing operation statuses were obtained and are discussed here. More detailed information about the gas holdup above the sieve plate under various operation conditions has been attained. Sieve plate resistance and bubbling state were analyzed with the use of image information entropy. [ABSTRACT FROM AUTHOR]

• A mask-guided semantic fusion method for optimizing the results of semantic segmentation. • A closed mask generation (CEG) module for post-processing of the edge detection algorithm. • A semantic confidence fusion (SCF) module to fuse the semantic segmentation results. • This method does not need to obtain the target data in advance. • This method has been verified in two scenarios. Though many high-precision semantic segmentation models have been proposed, how to improve the generalization ability of these models is still an urgent problem. Recently, a great number of unsupervised domain adaptation (UDA) algorithms around domain adaptation problems have been studied in semantic segmentation. These methods require labeled source domain data and unlabeled target domain data. In this paper, we propose a closed-mask-guided semantic fusion method (CSF) to improve the semantic segmentation results of unknown scenes, where the target domain data is not obtained in advance. First, a Closed Mask Generation (CMG) module is designed to convert the edge detection result into a mask that can segment the image into several image blocks. Then, a Semantic Confidence Fusion (SCF) module based on information entropy and voting method is introduced, which can select reliable semantic segmentation results for each image block by comparing the confidence of several semantic segmentation networks. In addition, the experimental results on both KITTI and COCO Stuff datasets validate the effectiveness of this method. The code is publicly available at https://github.com/tryhere/CSF. [ABSTRACT FROM AUTHOR]

Probabilistic summarization is the process of creating compact statistical representations of the original data. It is used for data reduction, and to facilitate efficient post-hoc visualization for large-scale patch-based data generated in parallel numerical simulation. To ensure high reconstruction accuracy, existing methods typically merge and repartition data patches stored across multiple processor cores, which introduces time-consuming processing. Therefore, this paper proposes a novel probabilistic summarization method for large-scale patch-based scientific data. It considers neighborhood statistical properties by importance-driven sampling guided by the information entropy, thus eliminating the requirement of patch merging and repartitioning. In addition, the reconstruction value of a given spatial location is estimated by coupling the statistical representations of each data patch and the sampling results, thereby maintaining high reconstruction accuracy. We demonstrate the effectiveness of our method using five datasets, with a maximum grid size of one billion. The experimental results show that the method presented in this paper reduced the amount of data by about one order of magnitude. Compared with the current state-of-the-art methods, our method had higher reconstruction accuracy and lower computational cost. [Display omitted] • A novel probabilistic summarization method for large-scale patch-based scientific data. • The neighborhood statistical properties are considered by importance-driven sampling guided by the information entropy. • Compared with the current state-of-the-art methods, our method had higher reconstruction accuracy and lower computational cost. [ABSTRACT FROM AUTHOR]

Clustering is a fundamental task in data mining, aiming to group similar objects together based on their features or attributes. With the rapid increase in data analysis volume and the growing complexity of high-dimensional data distribution, clustering has become increasingly important in numerous applications, including image analysis, text mining, and anomaly detection. DBSCAN is a powerful tool for clustering analysis and is widely used in density-based clustering algorithms. However, DBSCAN and its variants encounter challenges when confronted with datasets exhibiting clusters of varying densities in intricate high-dimensional spaces affected by significant disturbance factors. A typical example is multi-density clustering connected by a few data points with strong internal correlations, a scenario commonly encountered in the analysis of crowd mobility. To address these challenges, we propose a Self-adaptive Density-Based Clustering Algorithm for Varying Densities Datasets with Strong Disturbance Factor (SADBSCAN). This algorithm comprises a data block splitter, a local clustering module, a global clustering module, and a data block merger to obtain adaptive clustering results. We conduct extensive experiments on both artificial and real-world datasets to evaluate the effectiveness of SADBSCAN. The experimental results indicate that SADBSCAN significantly outperforms several strong baselines across different metrics, demonstrating the high adaptability and scalability of our algorithm. [ABSTRACT FROM AUTHOR]

• Entropy-based causality learning framework is proposed for building fault diagnosis. • The proposed synchronicity concept is employed to measure building symptom correlations. • Causal learning aims to determine Bayesian network structure. • Modelica test cases demonstrate the framework's effectiveness. Faults, such as malfunctioning sensors, equipment, and control systems, significantly affect a building's performance. Automatic fault detection and diagnosis (AFDD) tools have shown great potential in improving building performances, including both energy efficiency and indoor environment quality. Since modern buildings have integrated systems where multiple subsystems and equipment are coupled, many faults in a building are cross-level faults, i.e., faults occurring in one component that trigger operational abnormalities in other subsystems. Compared with non-cross-level faults, it is more challenging to isolate the root cause of a cross-level faults due to the system coupling effects. Bayesian networks (BNs) have been studied for the root cause isolation for building faults. While promising, existing BN-based diagnosis methods highly rely on expert domain knowledge, which is time-consuming and labor expensive, especially for cross-level faults. To address this challenge, we propose an entropy-based causality learning framework, termed Eigen-Entropy Causal Learning (EECL), to learn BN structures. The proposed method is data-driven without the use of expert domain knowledge; it utilizes causal inference to determine the causal mechanisms between faults status and symptoms to construct the BN model. To demonstrate the effectiveness of the proposed framework, three fault test cases are used for evaluation in this study. Experimental results show that the BN constructed by the proposed framework is able to conduct building cross-level faults diagnosis with a comparable isolation accuracy to those by domain knowledge while maintaining less complexed BN structure. [ABSTRACT FROM AUTHOR]

We study quantum dynamics of a wave packet on a class of one dimensional decorated aperiodic lattices, described within a tight binding formalism. We look for the possibility of finding extended single particle states even in the absence of any translational periodicity. The chosen lattices are stubbed with one or more atoms, tunnel coupled to the backbone, thereby introducing a minimal quasi-one dimensionality. It is seen that, for a group of such lattices a certain correlation between the numerical values of the hopping amplitudes leads to a complete delocalization of single particle states. In some other cases, a special value of a magnetic flux trapped in the loops present in the geometries delocalize the states, leading to a flux driven insulator to metal transition. The mean square displacement, temporal autocorrelation function, the time dependence of the inverse participation ratio, or the information entropy – the so-called hallmarks of studying localization based on dynamics – all of them indicate such a complete turnover in the nature of the single particle states and the character of the energy spectrum under suitable conditions. The results shown in this work using quasiperiodic lattices of the Fibonacci family are much more general and hold good even for a randomly disordered arrangement of the building blocks of the systems considered, and indicate a subtle universality class under which these lattices can be grouped. • We examine the quantum dynamics of a wave packet on a family of one-dimensional aperiodic lattices. • It is observed that a specific connection between the hopping amplitudes for a set of such lattices results in a total delocalization of single particle states. • In other instances, the states are delocalized due to a unique value of a magnetic flux that is trapped in the loops found in the geometries. • We analyze the mean square displacement, temporal autocorrelation function, time dependence of the inverse participation ratio, and information entropy. • The findings presented here with quasiperiodic lattices from the Fibonacci family are far more broadly applicable. [ABSTRACT FROM AUTHOR]

Concept lattice theory, which is one of the key mathematical models of granular computing, is capable of successfully dealing with uncertain information in nominal data. It has been applied to machine learning tasks such as data reduction, classification, and association rule mining. For the problem of outlier detection in nominal data, this paper presents a concept lattice theory-based approach for detecting outliers in nominal data. First, subcontexts and concept lattices based on subsets of objects are discussed. Then, information entropy is introduced into the formal context, and an object entropy based on attribute granules is proposed. Finally, a nominal data-oriented outlier detection method is explored based on the proposed object entropy. The experimental results show that the proposed detection method can effectively detect outliers in nominal data. Besides, the results of the hypothesis testing indicate that the proposed method is statistically significantly different from the other methods. The code is publicly available online at https://github.com/from-china-to/OEOD. [ABSTRACT FROM AUTHOR]

The distinguishable deep features are essential for the 3D point cloud recognition as they influence the search for the optimal classifier. Most existing point cloud classification methods mainly focus on local information aggregation while ignoring the feature distribution of the whole dataset that indicates more informative and intrinsic semantic relationships of labeled data, if better exploited, which could learn more distinguishing inter-class features. Our work attempts to construct a more distinguishable feature space through performing feature distribution refinement inspired by contrastive learning and sample mining strategies, without modifying the model architecture. To explore the full potential of feature distribution refinement, two modules are involved to boost exceptionally distributed samples distinguishability in an adaptive manner: (i) Confusion-Prone Classes Mining (CPCM) module is aimed at hard-to-distinct classes, which alleviates the massive category-level confusion by generating class-level soft labels; (ii) Entropy-Aware Attention (EAA) mechanism is proposed to remove influence of the trivial cases which could substantially weaken model performance. Our method achieves competitive results on multiple applications of point cloud. In particular, our method gets 85.8% accuracy on ScanObjectNN, and substantial performance gains up to 2.7% in DCGNN, 3.1% in PointNet++, and 2.4% in GBNet. Our code is available at https://github.com/YangFengSEU/CEDR. [ABSTRACT FROM AUTHOR]

• Notation of region of interest with critical contribution to failure probability is proposed. • A new active-learning function is devised based on the theory of information entropy. • Sufficient PC-Kriging accuracy in the range of the region of interest can be secured. • Computational burden can be significantly alleviated by the new active-learning function. • The proposed approach outperforms other adaptive surrogate-assisted simulation methods. A new active-learning function is developed to integrate Polynomial-Chaos Kriging with probability density evolution method. First, the relative importance of each representative point to the probability of failure is separately measured, thereby the region of interest is defined for the probability density evolution method, which covers the representative points exerting vital contributions to the probability of failure. Then, a new learning function called the probability density evolution method-oriented information entropy is readily devised based on the information theory, and the stopping condition is defined by specifying a threshold for the learning function values. Two examples are studied to show the efficacy of the adaptive Polynomial-Chaos Kriging probability density evolution method, and the recommended values of two key parameters associated with this new learning function are provided. Moreover, comprehensive comparisons are conducted against several existing reliability methods. The results highlight the advantage of the proposed active-learning function for structural reliability analysis in terms of both computational accuracy and efficiency. [ABSTRACT FROM AUTHOR]

Product competitiveness is highly influenced by its related design specifications. Information retrieval of customers preferences for the specification determination is essential to product design and development. Big sales data is an emerging resource for mining customers preferences on product specifications. In this work, information entropy is used for customers preferences information quantification on product specifications firstly. Then, a method of information mining for customers preferences estimation is developed by using big sales data. On this basis, a density-based clustering analysis is carried out on customers preferences as a decision support tool for the determination and selection of product design specifications. A case study related to electric bicycle specifications determination using big sales data is reported to illustrate and validate the proposed method. [ABSTRACT FROM AUTHOR]

To explore the trend of fatty acid methyl esters (FAMEs) during the oxidation process of biodiesel, this study uses gas chromatography - mass spectrometry (GC-MS) combined with information entropy theory to propose a new method to analyze the effect of FAMEs on the stability of biodiesel (especially with different amounts of unsaturated FAMEs). The induction periods of the main FAMEs in biodiesel were determined by the Rancimat method. The results showed that the longest induction period of methyl stearate was 248 h, while the induction of methyl linoleate was only 0.12 h. Accelerated oxidation of biodiesel, using GC-MS analysis showed that the oxidation trends of FAMEs was different. The oxidation rate of methyl linoleate was the fastest. A mathematical equation was proposed to calculate the trend of FAMEs. The results showed that during the 5 h accelerated an oxidation process, the content of methyl linoleate decreased from 28.35% to 2.10%, while the content of methyl oleate decreased by 18.21%. Using information entropy, the weighting coefficients of FAMEs in the oxidation process of biodiesel were calculated. The weighting coefficient of methyl linoleate was as high as 0.6797. Methyl linoleate was the main reason why biodiesel was easily oxidized. • GC-MS and information entropy methods was employed for FAMEs oxidation evaluation. •The weight coefficients of different FAMEs in the oxidation process are obtained. •Methyl linoleate is the main internal reason that affects the oxidation of biodiesel. [ABSTRACT FROM AUTHOR]

• Information entropy is first used for Kriging-based system reliability assessment. • The prediction uncertainty of the system state is quantified by information entropy. • A learning function is proposed using the prediction uncertainty of the system state. • Sample with the largest uncertainty get more attention for model reconstruction. Structural reliability assessment is a popular topic in engineering problems, particularly for the larger and more complex systems with implicit performance functions, also called black-box problems. The reliability assessment for a black-box problem must continuously be computed using simulation models, such as the finite element model, a highly time-consuming process with a high computational cost. The adaptive Kriging has gained considerable attention over the past decade. The Kriging-based reliability assessment method reduces the computational cost to a great extent, on the premise of ensuring the accuracy of reliability assessment. However, many of the currently published system reliability assessment methods, construct adaptive Kriging models by reducing the probability of incorrect prediction of the system state, and do not make full use of the uncertainty of the system state prediction information. To this end, a new Kriging-based method for structural system reliability assessment is proposed in this study. First, the probabilities of incorrect or correct system state predictions were understood from the perspective of information entropy. Second, an active learning strategy is proposed based on information entropy theory. Finally, the advantages of the proposed method are demonstrated and highlighted through several numerical examples. The results show that the proposed method achieves a good balance between the accuracy and computational cost, and the numerical magnitude effect does not affect the computational cost. Moreover, this is an effective method for assessing the reliability of complex systems. [ABSTRACT FROM AUTHOR]

The government service platform, exemplified by the government hotline, has to handle extensive volumes of business documents that contain rich and timely public opinion information and citizens' demands. However, manual processing struggles to process large-scale text data, adversely impacting operating costs and the quality of government services. This study proposes a comprehensive system for business document classification of the government hotline (BDCGHS) in China to address these challenges. BDCGHS leverages information entropy fused with term frequency-inverse document frequency (TF-IDF) weight to mine new words from business documents of the government hotline, and store them in a new word repository. These new words optimize Chinese word segmentation and text representation for text classification. We introduce a novel data structure called nested balanced binary tree to expedite new word mining, yielding a computational speed of almost five times than the Trie trees. Comparative experiments on the THUNews and government hotline datasets validate our proposed improvement BDCGHS algorithm's superior performance 3 % over text classification algorithms. Compared to the latest bidirectional encoder representations from the transformers (BERT) model, BDCGHS enhances the accuracy of order dispatch based on business documents by almost 3 %. It has also demonstrated stable operations in two Chinese cities for over a year, yielding favorable results. [Display omitted] • An embedded balanced binary tree structure is proposed for new word discovery. • A new word database is constructed for the government hotline. • The effects of mainstream classification algorithms are compared based on a new word database. • An intelligent text classification system for the government hotline is constructed and the results have promising results. [ABSTRACT FROM AUTHOR]

High-accuracy parameter learning in Bayesian Networks (BNs) is a key challenge in real-time decision support applications, particularly when the available data are limited. Prior/Expert knowledge was introduced to eliminate the drawbacks of insufficient information; however, this method is subjective. In this study, we explored the use of monotonicity constraints to control the causal relationships between the nodes and their parents-nodes in BNs and proposed a new learning algorithm called global domain monotonicity based on maximum information entropy (GDM-MIE), which was designed for parameter learning in uncertain discrete BNs with nonlinear equality constraints when only finite data are available. In the proposed algorithm, a class of monotonicity is encoded as a constraint on information entropy and the parameter learning problem is transformed into constraints among the node parameters based on a known network. Furthermore, we considered the parameters as uncertain entropy information and discussed the monotonicity among parameters in the global spatial domain, proving the accuracy of the logical relationships of the model, system reliability, and time complexity. Finally, the proposed method was validated using standard BNs, and its performance was analyzed by comparing the proposed method with the existing learning algorithms. The results showed that the proposed method is more accurate and has better Kullback–Leibler divergence. To revalidate the rationality of the proposed method, the Alarm and Asia networks were employed as special cases. The GDM-MIE was found to achieve the intended goal of observing the estimated parameters by closely approximating the original real parameters with a small sample size, indicating that the proposed algorithm can served as an efficient and feasible method for learning Bayesian parameter. [ABSTRACT FROM AUTHOR]

Fuzzy rough set is a highly effective mathematical method for feature selection, which offers clear interpretability without expert knowledge. However, most of fuzzy-rough feature selection methods are to rely on all samples and candidate features during the selection of a best feature at each iteration. This often shares high computation complexity and is inefficient for large datasets. Therefore, a two-way accelerator for feature selection is presented by stepwise narrowing the search space for both samples and features. As the foundation for our accelerator, a monotonic fuzzy conditional entropy, named parameterized fuzzy granule-based conditional entropy, is first proposed to guide the feature selection process. After identifying a best feature, a sample-based accelerator is then designed to disregard redundant samples for the calculation of the newly defined entropy. A feature-based accelerator is further proposed to eliminate redundant candidate features, of which the inclusion cannot change the proposed entropy of a currently selected feature subset. Our accelerator is finally developed by integrating the sample-based accelerator with the feature-based accelerator. Experimental comparisons demonstrate the effectiveness and efficiency of the proposed two-way accelerator for feature selection. [ABSTRACT FROM AUTHOR]

The efficient diagnosis of the bearing fault categories is vitally significant to enhance the overall safety of rotating machinery and decrease the maintenance cost. This paper introduces an innovative sparse measure approach for the bearing fault characteristics using Legendre wavelet multi-scale multi-mode (LWMSMM) frequency domain. The motivation behind this approach lies in the need for an effective method that overcomes complexities associated with extracting transient characteristics and frequency-related bearing fault features. The core principle of the proposed method is to leverage information Entropy (IE) to pick up the most sensitive fault features and utilize the genetic algorithm (GA) to optimize the control parameters of LWMSMM frame, including the decomposition level and the number of wavelet base, so as to thoroughly match the more salient fault characteristics of the bearing. The essence of the effectiveness of the method lies in two aspects. One is that the rich properties especial diverse regularities of LW can effectively approximate the complex fault characteristics. The other is that IE can effectively represent the dynamic characteristics of the faults at each resolution scale and each mode frequency domain. Finally, experiments are conducted on three datasets to validate the effectiveness and robustness of the presented method. The experimental results demonstrate that the developed method can accurately identify different fault categories with the simpler classifiers and achieves a diagnosis accuracy of 100 %, surpassing cutting-edge approaches, and it provides a new and promising method for rotating machinery in real industrial applications. [ABSTRACT FROM AUTHOR]

The shift of paradigm from reductionist to holism in the field of thermal adaptation promoted a boom in thermal adaptation theory and provided a systemic perspective on this issue. The thermal adaptation theory confronts twin difficulties of scale and complexity. The complexity and entropy of occupants' thermal adaptation emerging. This paper aims to investigate the information entropy exchange in the coupled climate-building-occupants system. First, we justified that the building-occupants system is an open system and exchanges energy, matter, and entropy with its surrounding climate system. Then, correlation analysis results suggested that climate entropy could influence thermal adaptation entropy. Thermal adaptation entropy fluctuated with climate entropy and the difference varied between −0.5 and 0.5 below 31 °C. A generalized entropy balance model was proposed for quantifying the entropy exchange among coupled climate-building-occupants systems for both free-running and air-conditioning buildings. The quantitative relation between the entropy of thermal adaptation and H_DTR reveals the dynamic occupant's thermal adaptation under the process of entropy exchange with the climate system. The H_DTR and entropy relation models provide a new perspective on thermal adaptation coupled with climate beyond outdoor daily mean temperatures. Control strategies that integrate the dynamic characteristics of DTR can bring more economical, energy-efficient and comfortable effects to the operation of air conditioning or heating systems. Our findings shed some light on the dynamical and complex thermal adaptation. In further studies, more results and methods from meteorological research could be used for thermal adaptation research and interdisciplinary theories could be adopted to help us better understand occupants' thermal adaptation. • The building-occupants system is an open system and exchange energy, matter, and entropy with the climate system. • The entropy of SCV and Clo show a strong positive relationship with DTR entropy. • Thermal adaptation entropy fluctuated with DTR entropy and the difference varied between −0.5 and 0.5 below 31 °C. • The entropy balance model was proposed for quantifying the entropy exchange. • Different H_DTR may lead to the difference in acceptable temperature ranges. [ABSTRACT FROM AUTHOR]

Takagi–Sugeno–Kang (TSK) classifiers have achieved great success in many applications due to their interpretability and transparent model reliability for users. At present, however, how to evaluate classification results is still an unsolved issue for TSK classifiers. This study designs a fuzzy rule-based classifier based on TSK classifiers, the outputs of which for an instance can be considered as the membership grades that the instance belongs to all classes. Then, an information entropy-based method is proposed to estimate the certainty of the outputs, which facilitates the further evaluation of the classification results of the instance for users. If the confidence level is not high, users can reject the classification results, and use other more advanced classifiers or collect more information about the instance. Moreover, the developed mechanism is suitable for handling large data since the adaptive moment estimation algorithm is used to identify the parameters of it. Experimental results demonstrate that the developed mechanism outperforms several rule-based classifiers. • The aim is to form a fuzzy rule-based classifier through information entropy. • The developed mechanism can provide the confidence of classification results. • The proposed classifier can handle large data due to the adaptive moment estimation. [ABSTRACT FROM AUTHOR]

As time passes, the data within information systems may continuously evolve, causing the target concept to drift. To ensure the effectiveness of data-driven decision making, it is crucial to detect drift in a timely manner and gather relevant information. In this paper, we introduce two methods that can directly detect concept drift in the provided information system, by considering a new perspective on uncertainty. First, using entropy under a single attribute constraint, we define the uncertainty of the target concept in an information system. By integrating the uncertainty of each attribute, the overall uncertainty of the target concept in the information system is obtained. Subsequently, two concept drift detection methods are proposed, namely EBTBM (Entropy-Based Threshold-Based Method) and EBSBM (Entropy-Based Sampling-Based Method). These methods utilize the defined uncertainty of the target concept as a statistical measure of the difference between two data blocks. Finally, extensive experiments on artificial and real-world data sets are conducted to validate the effectiveness of the proposed concept drift detection methods. [ABSTRACT FROM AUTHOR]

Underwater object detection is considered as one of the most challenging issues in computer vision. In this paper, a weighted multi-error information entropy based YOLO (You Only Look Once) network is proposed to address underwater illumination noise affecting the detection accuracy. First, underwater illumination is essentially structural and non-uniform, and it is modeled as an independent and piecewise identical distribution, which is a generic noise model to describe the complex underwater illuminating environment and accommodates the traditional Gaussian distribution as a special case. Second, assisted by the proposed illumination noise model, a minimum weighted error entropy criterion, which is an information-theoretic learning method, is introduced into the loss function of YOLO network, and then the network parameters are trained and optimized to improve the detection performance. Furthermore, a multi-error processing strategy is simultaneously used to handle vector errors during information back-propagation in order to accelerate convergence. Experiments on underwater object detection datasets including URPC2018, URPC2019 and Enhanced dataset, show the proposed weighted multi-error information entropy based YOLOv8 network gets mean average precision (MAP) of 88.7%, 91.8% and 96.7% respectively, and average frames per second (FPS) of 116.6. These two evaluation metrics are better than the baseline YOLOv8 and the existing advanced non-YOLO approaches by at least 5.2% and 5.3% respectively. The results verify the effectiveness and superiority of the proposed network for underwater object detection in complex underwater environment. [ABSTRACT FROM AUTHOR]

Differential network analysis has become an important approach in identifying driver genes in development and disease. However, most studies capture only local features of the underlying gene-regulatory network topology. These approaches are vulnerable to noise and other changes which mask driver-gene activity. Therefore, methods are urgently needed which can separate the impact of true regulatory elements from stochastic changes and downstream effects. We propose the differential network flow (DNF) method to identify key regulators of progression in development or disease. Given the network representation of consecutive biological states, DNF quantifies the essentiality of each node by differences in the distribution of network flow, which are capable of capturing comprehensive topological differences from local to global feature domains. DNF achieves more accurate driver-gene identification than other state-of-the-art methods when applied to four human datasets from The Cancer Genome Atlas and three single-cell RNA-seq datasets of murine neural and hematopoietic differentiation. Furthermore, we predict key regulators of crosstalk between separate networks underlying both neuronal differentiation and the progression of neurodegenerative disease, among which App is predicted as a driver gene of neural stem cell differentiation. Our method is a new approach for quantifying the essentiality of genes across networks of different biological states. [ABSTRACT FROM AUTHOR]

What is lacking from the literature and practice of library print collection assessment is a mathematical function that can integrate variables and quantify the overall measurement of the collection as a final score for the library science practitioners and administrators. The study uses the information entropy to explain a mathematical procedure on how to define variables from the regular circulation statistics. Based on the assessment concepts of circulation statistics and user behavior, the study explores to develop mathematical functions and proposes a Collection-User Suitability index to quantify the suitability measurement of the overall collection and the user intention. The study then considers the extreme conditions to validate the functions and verifies the application of the function via empirical and simulated data. Finally, the authors developed a computing tool and uploaded it to GitHub for free access based on the study's results. It is hoped that library science practitioners and administrators can utilize the mathematical function. [ABSTRACT FROM AUTHOR]

Many-objective optimization problems (MaOPs) are challenging tasks involving optimizing many conflicting objectives simultaneously. Decomposition-based many-objective evolutionary algorithms have effectively maintained a balance between convergence and diversity in recent years. However, these algorithms face challenges in accurately approximating the complex geometric structure of irregular Pareto fronts (PFs). In this paper, an information entropy-driven evolutionary algorithm based on reinforcement learning (RL-RVEA) for many-objective optimization with irregular Pareto fronts is proposed. The proposed algorithm leverages reinforcement learning to guide the evolution process by interacting with the environment to learn the shape and features of PF, which adaptively adjusts the distribution of reference vectors to cover the PFs structure effectively. Moreover, an information entropy-driven adaptive scalarization approach is designed in this paper to reflect the diversity of nondominated solutions, which facilitates the algorithm to balance multiple competing objectives adaptively and select solutions efficiently while maintaining individual diversity. To verify the effectiveness of the proposed algorithm, the RL-RVEA compared with seven state-of-the-art algorithms on the DTLZ, MaF, and WFG test suites and four real-world MaOPs. The results of the experiments demonstrate that the suggested algorithm provides a novel and practical method for addressing MaOPs with irregular PFs. • A novel RL-RVEA addresses many-objective optimization with irregular pareto fronts. • A reinforcement learning-based adaptive reference vector to guide the direction of convergence. • A scalarization approach preserves the diversity of solutions for next generation. • The RL-RVEA outperforms seven advanced many-objective optimization algorithms. [ABSTRACT FROM AUTHOR]

• Analyzes the spring drought in the cold region of Northeast China from 1980 to 2020. • Considering the impact of early spring snowfall, the spring drought was analyzed. • A novel cutoff radius approach for the PP information entropy model. Northeast China is an important grain-producing area, and spring is the sowing time for crops. Drought hinders crop germination and leads to crop failure. Spring drought analysis helps to understand the characteristics of regional spring drought and provides a reference for formulating disaster prevention and reduction measures. Considering the characteristics of seasonally frozen soil regions in northeastern China, sampling at a resolution of 0.5 degrees, accounting for the snow cover factors in the early spring, as well as precipitation, temperature, evapotranspiration, and soil moisture during the spring planting period, and uses the projection pursuit information entropy model to analyze the spatial and temporal variability of spring drought in Northeast China. In the model, a new method was proposed to calculate the cutoff radius by adjusting the radius size with the parameter. The optimal parameters were ultimately determined using the Shapiro-Wilk test, Kolmogorov-Smirnov test, and Epps-Pulley test. The results are as follows: drought is relatively severe in the eastern parts of Inner Mongolia, as well as in the northwest of Liaoning. The central and northern parts of Inner Mongolia, the northern and northeastern parts of Heilongjiang, the southeastern parts of Jilin, and the eastern parts of Liaoning are relatively humid, while the drought situation in other regions is relatively moderate. Regional spring drought shows an increasing trend, and most regions show statistical significance. Areas with high temperatures and low precipitation, as well as low snow cover in the early spring, are prone to spring drought. Moist regions are characterized by high precipitation, low evapotranspiration, and abundant snow resources. [ABSTRACT FROM AUTHOR]

When hazardous substances leak into the atmosphere, Source Term Estimation (STE) can quickly determine the information of the leaked substance to control the threat of the leakage event to the environment and public safety. Reasonable sensor configuration is the foundation of STE. Existing sensor configuration schemes have limited applicability in STE processes. When urban neighborhood has complex building structures and airflow conditions, there is a lack of a method that can provide reasonable sensor configurations to ensure good STE performance. In this study, a new sensor configuration design method is proposed, which weights the joint entropy of adjoint concentration information of multi wind directions to obtain the cost function, and then uses Simulated Annealing (SA) algorithm to minimize the cost function. This article selects a real complex urban neighborhood for simplified modeling, and designs the optimal sensor array for it. STE results of several leakage scenarios are compared with the optimal and uniform sensor configurations at different leakage locations and wind directions. Compared to traditional uniform sensor configurations, the optimal sensor configuration has significantly improved estimation performance in both source location and strength, with a reduction of 27 %–51 % in the bias of source location identification and 17 %–39 % in the bias of source strength estimation. • A design method for sensor configuration in solving Source Term Estimation (STE) problems is proposed. • This method can cope with STE problems in urban neighborhood facing complex structures and wind conditions. • A complex urban neighborhood is selected to validate the rationality of the proposed method. • This method is making possible to apply sensor optimization strategies in a real urban neighborhood. [ABSTRACT FROM AUTHOR]

The protection and utilization of traditional villages have evolved from the individual to the regional level and have gradually formed a network. China has introduced a policy of centralized contiguous protection and utilization areas for traditional villages. Intervention in modern relationships has led to the reconstruction of connections between villages. It is necessary to explore the modern bonds between villages with historical connections. Using the existing 85 centralized contiguous protection and utilization areas and 8155 traditional villages as the research objects, this study depicts the geographical, cultural, and industrial ties of the centralized contiguous protection and utilization areas in 2010 and 2020 based on the information entropy model. In addition, it studies the traditional village network using the gravity model of spatial interaction to explore the evolution and boundary rationality of the centralized contiguous protection and utilization areas. Research has shown that (1) the cultural relationship between centralized contiguous protection and utilization areas has fluctuated over the past decade, and the industrial relationship has generally strengthened. (2) Historical and cultural relics such as architectural relics, intangible heritage sites, and cultural corridors are the main forces that maintain the cultural and industrial ties of villages. Socioeconomic factors generally have a negative impact, and a mechanism exists for the modern bond between centralized contiguous protection and utilization areas to change from cultural connections to industrial connections. (3) Existing centralized contiguous protection and utilization areas have good efficiency. However, there are problems with delineating boundaries in terms of limited county boundaries and large municipal boundaries. Therefore, building a governance framework of a supercluster city and county coordination-centralized contiguous protection and utilization areas is proposed. • Measuring multiple ties of centralized contiguous areas by information entropy. • Historical and cultural relics maintain the kinship and industrial ties of village. • Industrial bonds can easily form separately, but kinship bonds can generate both. • About 40 % of traditional villages include about 70 % of village connections. • A better boundary for contiguous areas should be between city and county levels. [ABSTRACT FROM AUTHOR]

Phthalocyanine derivative nanostructures are highly organized organometallic structures that exhibit two-dimensional polymeric phthalocyanine frameworks. These conjugated macrocycles, characterized by their versatility, have gained significant attention in the development of technologies due to their structural resemblance to naturally occurring porphyrin complexes. The growth of these structures is analyzed to explore their topological characteristics by employing connectivity-based bond additive descriptors. The generated results are utilized in conjunction with the selected samples from Chem3D software to establish predictive models for calculating log P , log S , polar surface area, and total energy. Moreover, the structural complexity of these frameworks is compared through classical, spectral, and topological index based entropies, with the identification of the advantages and disadvantages associated with each of the three approaches. [Display omitted] [ABSTRACT FROM AUTHOR]

Accurate wind power prediction is significant to the stability of power system. Existing deterministic prediction methods unable to describe the uncertainty of wind power while both the point and probabilistic models have difficulty in tracking the abrupt changes in wind power accurately. To settle these problems, a point-interval prediction model based on combination of improved preprocessing methods and entropy weighted GRU quantile regression network (QR-EGRU) is proposed. Firstly, an improved wavelet threshold denoising (IWTD) is applied to reduce noise interference. An optimized variational mode decomposition (OVMD) based on sparrow search algorithm (SSA) is proposed to decompose the series into subsequences. Secondly, two update gate matrices based on information entropy (IE) are introduced to replace the traditional update gate matrix of the GRU to construct the EGRU. Point prediction results are obtained by using the EGRU model. Furthermore, the QR algorithm with nonlinear loss function is derived to realize the interval prediction of the EGRU. Finally, the proposed model is validated on real wind power data from the Kaggle competition. Experimental results demonstrate that the proposed model performs well in both point and interval prediction. It can track the mutation series more precisely than other models and improve the prediction accuracy. • An improved WTD and optimized VMD algorithms are proposed in this paper to improve the data quality. • Two update gate weight matrices are introduced to replace the traditional update gate weight matrix of the GRU network. • The information entropy is used to emphasize the impact of the update gate weight matrices. • EGRU model is proposed to realize wind power point prediction. • A novel nonparametric interval prediction model based on QR algorithm and EGRU is proposed. [ABSTRACT FROM AUTHOR]

Unpaired image-to-image translation among category domains has achieved remarkable success in past decades. Recent studies mainly focus on two challenges. For one thing, such translation is inherently multi-modal (i.e. many-to-many mapping) due to variations of domain-specific information (e.g., the domain of house cat contains multiple sub-modes), which is usually addressed by predefined distribution sampling. For another, most existing multi-modal approaches have limits in handling more than two domains with one model, i.e. they have to independently build two distributions to capture variations for every pair of domains. To address these problems, we propose a Hierarchical Image-to-image Translation (HIT) method which jointly formulates the multi-domain and multi-modal problem in a semantic hierarchy structure by modeling a common and nested distribution space. Specifically, domains have inclusion relationships under a particular hierarchy structure. With the assumption of Gaussian prior for domains, distributions of domains at lower levels capture the local variations of their ancestors at higher levels, leading to the so-called nested distributions. To this end, we propose a nested distribution loss in light of the distribution divergence measurement and information entropy theory to characterize the aforementioned inclusion relations among domain distributions. Experiments on ImageNet, ShapeNet, and CelebA datasets validate the promising results of our HIT against state-of-the-arts, and as additional benefits of nested modeling, one can even control the uncertainty of multi-modal translations at different hierarchy levels. • Propose a Hierarchical Image Translation (HIT) framework for categorical translation. • Propose a nested distribution space to formulate multi-domain and multi-modal issues. • Propose a nested distribution loss to characterize the domains' inclusion relations. • Demonstrate promising results in image quality, target accuracy and model efficiency. [ABSTRACT FROM AUTHOR]

• The original data is reconstructed to better extract demand features and filter noise. • The spatio-temporal interactive clustering is more suitable for large-scale bike-sharing system. • The information entropy based on random matrix is introduced to measure the stability of model. • The new clustering method in this paper can greatly reduce the rebalancing cost. As an important part of green travel mode, operation service of bike-sharing system is increasingly intelligent and refined. Operators can effectively match supply to demand through reasonable delivery and rebalancing methods. One of the most important foundations of these strategies is zone management. Therefore, this paper establishes a new framework of management area division, which includes three parts: data reconstruction, clustering model and model performance. Firstly, we reconstruct the original demand data through Coarse Grain and Visibility Graph methods. The reconstructed data highlights the characteristics of demand fluctuation and filters the noise. Secondly, we build a new spatio-temporal interactive clustering model. Through time dimension clustering analysis, we not only divide the clustering window, but also give three types of temporal labels of the demand series. At the same time, we give the clustering results of spatial dimensions. We interact the spatial clustering results of the demand series with the temporal labels under the combination of the Integration and Constraint criterions in each clustering window. Thirdly, we use the information entropy index to measure the stability of the clustering results and compares the distance cost of rebalancing before and after clustering. Empirical results show that after the spatio-temporal interaction clustering, demand fluctuation in each cluster has high consistency and stability in both time and space dimensions. The average information entropy not only decreases by more than 69 % compared with the results before clustering, but also smaller than the average level of merging without spatio-temporal interaction. Moreover, the rebalancing cost after spatio-temporal interactive clustering is only 61.26 % of the actual rebalancing cost. This helps operators give more efficient rebalancing strategies and more stable rebalancing routing than before. [ABSTRACT FROM AUTHOR]

Multi-energy microgrids deploy distributed power sources with flexible generation characteristics to meet the changing needs of users. Because they comprise renewable as well as traditional energy sources, they can play an important role in the transition to clean and efficient low-carbon power generation. However, both the complexity of coordinating different power sources and random fluctuations of renewable-energy generation capacity pose significant challenges. To solve these problems, a risk-limiting dispatching strategy for multi-energy microgrids is proposed that considers net load-demand response. First, a bi-directional demand-response mechanism is introduced, and a flexible load is used as a dispatchable resource to optimize the net load curve of the system. Second, entropy-based reliability modeling is conducted to address the multiple uncertainties in system operation. Third, a risk-limiting dispatching model of the multi-energy microgrid is established that considers three optimization objectives: economy, environmental protection, and reliability. Finally, the NSGA-III algorithm is used to obtain the set of optimal solutions for the model, and a compromise solution that meets the system operation requirements is selected based on the technique for order preference by similarity to an ideal solution (TOPSIS) with a criteria importance though intercriteria correlation (CRITIC) and analytic hierarchy process (AHP, CRITIC–AHP) hybrid assignment. The simulation results show that the proposed optimal scheduling strategy improves the operational reliability index while ensuring system economy and environmental protection. This study provides a new approach for controlling system risk and ensuring continuous energy-supply capability under uncertain conditions. • The microgrid reliability indicators based on information entropy theory is proposed to limit potential risks. • A risk-limiting dispatching strategy for multi-energy microgrids is proposed. • An optimal scheduling strategy is proposed to improve the operation efficiency. [ABSTRACT FROM AUTHOR]