Search

Network intrusion detection is a vital component in contemporary cybersecurity defense frameworks. As network traffic data experiences exponential growth, deep learning-based intrusion detection algorithms have become a hot research topic. However, existing intrusion detection methods commonly exhibit low accuracy, high false positive rates, and poor performance in detecting minority class attacks. To address these challenges, this paper presents an efficient two-stage intrusion detection model, MARNet. MARNet consists of two stages: the first stage is multi-scale feature-classification extraction, where numerical and categorical features in the dataset are processed separately and then fused. The second stage is the classification phase, which employs residual networks and channel attention to automatically and effectively extract features from network traffic, while identifying crucial features. Furthermore, we implemented EQL v2 loss to tackle the problem of class imbalance within the dataset. Experiments were conducted using the UNSW-NB15 and NSL-KDD datasets, After a thorough and comprehensive evaluation on two widely used datasets, the model achieved a multi-class accuracy of 90.55% and a detection rate of 99.51% with a false positive rate of 0.61% on the UNSW-NB15 dataset. On the NSL-KDD dataset, the model’s multi-class accuracy was 99.96%, detection rate was 99.96%, and the false positive rate was 0.01%. The experimental results show that the model surpasses existing ones in accuracy, detection rate, false alarm rate, and detection of minority class attacks, effectively meeting the requirements of modern large-scale intrusion detection systems.

Accurate building semantic segmentation in remote sensing imagery is essential for urban planning, environmental monitoring, and map creation. While deep learning has achieved significant advancements in this field, precisely segmenting building edges and shadows in complex scenarios remains challenging. Shadows often introduce boundary ambiguities, affecting the local shape and texture information of buildings. Current methods do not fully perceive or utilize shadows. To address these challenges, we propose an advanced high-resolution image segmentation network, high-resolution network, integrated with a shadow-inclusive edge perception module. Our approach involves introducing a shadow-inclusive contour transition module (SCTM) during the feature extraction stage to enhance the features of blurry boundaries. The proposed SCTM and shadow-aware attention module significantly enhance attention maps, improve responses in blurry boundary regions, and increase consistency between predictions and ground truth, setting a new benchmark for building semantic segmentation in remote sensing imagery. This enriched information is then fed into an attention module that concurrently focuses on boundary and channel features, surpassing traditional semantic segmentation methods. We validated our method on three datasets: Massachusetts, WHU, and Inria. Our approach outperformed state-of-the-art methods on the WHU Building Dataset across all metrics, including mIoU, Accuracy, Kappa, and Dice coefficient.

In complex environments, mobile robots performing tasks with different hazard levels need to consider different road factors, this paper proposes a functional model correlating task hazard levels with road factors, proposing an innovative Hybrid Adaptive Genetic Algorithm (HAGA). The HAGA integrates an optimized two-optimization (2-opt) operator* with an enhanced Adaptive Genetic Algorithm (AGA) for efficient path planning in diverse tasks and complex road conditions. Firstly, pre-optimize the initial paths is performed by introducing a new domain knowledge-based operator that duplicates paths in the path are deleted to avoid the redundant paths, and then they are divided into the TOP layer and the ordinary layer, the TOP layer is optimized by using the adaptive 2-opt* operator that satisfies the hyperbolic tangent function (TANH), and the crossover and variability of the ordinary layer are optimized by using the S-type function (Sigmoid function, Sigmoid) and TANH AGA for the optimization treatment for the crossover and variance of the ordinary layer, respectively, to establish a robot path planning algorithm suitable for multitasking in complex environments. The experiment proves that the improved HAGA has strong global search ability and also improves the local search ability, has good generality and robustness, and reduces the optimal path distance by a minimum of 2.74% and a maximum of 10.86% compared with the comparison algorithm in the experiment. The experimental results showed that the method enables mobile robots to perform tasks with different hazard levels in complex environments with good generalization and robustness, and the coefficient of variation (CV) of repeated experiments at five safety levels was kept within 2 %. which had good universality and versatility. This research has the potential to enhance the safety and efficiency of mobile robot operations in challenging environments.

With the rapid advancements in deep learning technology, various deep learning-based object detection algorithms have found extensive applications in UAV-related tasks. However, motivated by the fact that current object detection algorithms for unimodal aerial remote sensing images fail to achieve around-the-clock object detection. To tackle this, we propose an around-the-clock object detection algorithm YOLOFIV that fuses infrared and visible features. First, we design a dual-stream backbone network based on the attention mechanism to adequately extract the features of both modalities. Moreover, the ECA attention mechanism is integrated into the feature enhancement network to amplify attention toward challenging detection scenarios. Finally, we improve the horizontal detection head to a rotating one to preserve object orientation. We evaluate the proposed method YOLOFIV on the widely used drone vehicle dataset, YOLOFIV achieves an accuracy of 64.71% (in terms of $\text{mean average precision}_{0.5}$), accuracy improvement of 8.32% over baseline bimodal model, similar performance to UACMD designed for ARSI object detection but with 92.35% reduction in parameter count, and 17.87 times speedup. The results show that our approach achieves round-the-clock object detection while maintaining a favorable accuracy-speed tradeoff.

Functional magnetic resonance imaging (fMRI) has emerged as a prevalent tool for investigating motor deficits and rehabilitation in the context of stroke. Particularly, the exploration of functional connectivity (FC) through resting-state fMRI has the potential to unveil the neural connectivity mechanisms underlying post-stroke motor impairment and recovery. Despite the significance of this approach, there exists a gap in the literature where a comprehensive review dedicated to post-stroke functional connectivity analysis is lacking. In this paper, we undertake an extensive review of both static functional connectivity network analysis (SFC) and dynamic functional connectivity network analysis (DFC) in the context of post-stroke motor dysfunction. Our primary goal is to present comprehensive methodological insights and the latest research findings pertaining to motor function recovery after stroke. We commence by providing a succinct overview of SFC and DFC methods, elucidating the preprocessing and denoising techniques essential to these analyses. Subsequently, we summarize the application of two methods in stroke disease, mainly focusing on the extracted insight into post-stroke brain dysfunction and rehabilitation. Our review indicates a prevalence of SFC as the method of choice for post-stroke functional connectivity investigations. Specifically, SFC studies reveal a reduction in FC between motor areas due to stroke lesions, with increased FC correlating positively with functional recovery. Nevertheless, the DFC for post-stroke analysis has only begun to unveil its potential due to its ability in temporal dynamics. In summary, this review paper presents a thorough understanding of post-stroke functional connectivity analysis and its implications for the study of motor function recovery, offering valuable insights for future research and clinical applications.

In this paper, the scattering characteristics of polarized light are studied based on the microsurface element theory, and the Polarization Bi-directional Reflection Distribution Function (pBRDF) model applicable to rough surfaces is established according to the three-component model, and the expression of polarization of reflected light from the target surface is derived based on the model, and the accuracy of the model is verified by comparing the model with the conventional model. The modeling is also carried out from the scattering perspective, and the relationship between different artificial target particle diameters and polarization characteristics is analyzed. The results show that the Degree Of Polarization (DOP) is strongly related to the detection zenith angle. The DOP of the target is largest in the direction of the mirror reflection of the source. The polarization characteristics of different materials are different, and the relative error of the established model is significantly smaller than that of the conventional model. And the scattering model can make preliminary classification of different man-made targets according to different particle diameters, which can provide theoretical support for the analysis of the polarization characteristics of the target surface.

In the routine inspection process of railway catenary systems, the primary task is to find out the locations of various components accurately. The complex composition of the components in the catenary system and their large dimensional differences make the inspection of small components considerably difficult. Aiming at the problem of the difficulty in locating small components, a new locating method, named asymmetrically effective decoupled head-you only look once (AED-YOLO), for locating small components of the catenary has been proposed in this study. In this method, firstly, a small object detection layer has been added to improve the detection accuracy of the small-sized components such as fastener nuts and bracing wire. Secondly, to reduce missed and false detection errors of small components, the improved bidirectional feature pyramid network with high-order spatial interactions and recursive gated convolution has been used to fuse the features of different scales to further enhance the ability to detect small objects. Finally, an asymmetrically effective decoupled head has been proposed using different decoupled branches to decouple the classification and localization processes, thus further reducing the error in small-sized object classification and location. Experiments performed on the railway catenary dataset collected on-site show that the proposed localization method can efficiently improve detection accuracy, achieving a mean average precision of 93.5%. Thus, compared to the other methods, the proposed method can accurately locate small-sized components.

Rehabilitation is essential for post-stroke body function recovery. Supported by the mirror neuron theory, motor imagery (MI) has been proposed as a potential stroke therapy capable of facilitating the rehabilitation. However, it is often quite difficult to estimate the degree of the participation of patients during traditional MI training as well as difficult to evaluate the efficacy of MI based rehabilitation methods. The goal of this paper is to develop a virtual reality (VR) based MI training system combining electromyography (EMG) based real-time feedback for poststroke rehabilitation, with the immersive scenario of the VR system providing a shooting basketball training for bilateral upper limbs. Through acquiring electroencephalography (EEG) signal, the brain activity in alpha and beta frequency bands was mapped and the correlation analysis could be achieved. Furthermore, EMG data of each patient was collected and calculated as threshold with root-mean-square algorithm for feedback of the performance score of the shooting basketball training in virtual environment. To investigate the feasibility of this newly-built rehabilitation training system, four experiments namely initial assessment experiment, motor imagery (MI), action observation (AO), and combined motor imagery and action observation (MI+AO) were carried out on stroke patients at different recovery stages. The result shows that MI+AO can generate more pronounced event-related desynchronization (ERD) in alpha band compared to other cases and induce relatively obvious ERD in beta band compared to AO, which demonstrates that VR-based observation has ability to facilitate MI training. Furthermore, it has been found that the muscle strength from MI+AO is the highest through the EMG analysis. This proves that the feedback of EMG can be used to quantify patient’s training engagement and promote MI training at a certain extent. Hence, by incorporating such an EMG feedback, a VR-based MI training system has the potential to achieve higher efficacy for post-stroke rehabilitation.

Goal: The purpose of this work is to improve malaria diagnosis efficiency by integrating smartphones with microscopes. This integration involves image acquisition and algorithmic detection of malaria parasites in various thick blood smear (TBS) datasets sourced from different global regions, including low-quality images from Sub-Saharan Africa. Methods: This approach combines image segmentation and a convolutional neural network (CNN) to distinguish between white blood cells, artifacts, and malaria parasites. A portable system integrates a microscope with a graphical user interface to facilitate rapid malaria detection from smartphone images. We trained the CNN model using open-source data from the Chittagong Medical College Hospital, Bangladesh. Results: The validation process, using microscopic TBS from both the training dataset and an additional dataset from Sub-Saharan Africa, demonstrated that the proposed model achieved an accuracy of 97.74% ± 0.05% and an F1-score of 97.75% ± 0.04%. Remarkably, our proposed model with AlexNet surpasses the reported literature performance of 96.32%. Conclusions: This algorithm shows promise in aiding malaria-stricken regions, especially those with limited resources.

The tremendous development of the Internet of Things (IoT) technology in the last decades has fostered advancement in automatic medical assistive devices to support the existing Electronic Health Record (EHR) system. As an integral part of the EHR electronic model, public cloud servers store medical data. Unfortunately, public cloud servers are prone to security and privacy breach. This paper introduces a novel non-cryptographic approach to preserve electrocardiograph (ECG) data confidentiality and integrity in the EHR environment. The main objective of the proposed anonymization algorithm is to obscure the patient’s cardiac information during transmission and to protect information stored in the cloud database. Although we focus on ECG data, generalization to other types of clinical data can be derived using the proposed method. Performance evaluation of the proposed scheme showed that the algorithm conceals both fiducial and non-fiducial features of the data. Therefore, confidentiality feature is preserved. This paper examined confidentiality of the anonymized data using the Percentage Residual Difference (PRD) and investigated the integrity of the reconstructed data in terms of cross-correlation. Security analysis carried out using the PRD, brute force attack, and performance comparison between the proposed algorithm and existing methods. Evaluation showed that the proposed scheme offers a secure non-cryptographic model for transmission and storing clinical data in the cloud. Moreover, in terms of processing time, the proposed algorithm is ten times faster than the existing wavelet packet method when processing long ECG data, 65,536 sample points. In a real-time experimental testbed, the implemented proposed system was successful.

This paper addresses trajectory optimization problems for underwater data muling with mobile nodes. In the underwater data muling scenario, multiple autonomous underwater vehicles (AUVs) sample a mission area, and autonomous surface vehicles (ASVs) visit the navigating AUVs to retrieve the collected data. The optimization objectives are to simultaneously maximize fairness in data transmissions and minimize the travel distance of the surface nodes. We propose an nearest- $K$ reinforcement learning algorithm, which chooses only from the nearest- $K$ AUVs as candidates for the next node for data transmissions. We use the distance between AUVs and the ASV as the state, selected AUVs as the action. A reward is designed as the function of both the data volume transmitted and the ASV travel distance. In the scenario with multiple ASVs, an AUV association strategy is presented to support the use of multiple surface nodes. We conduct computer simulations for performance evaluation. The effects from the number of AUVs, the size of the mission area, and the state number are investigated. The simulation results show that the proposed algorithm outperforms traditional methods in terms of the fairness and ASV travel distance.

Effective detection of small targets plays a pivotal role in infrared (IR) search and track applications for modern military defense or attack. However, IR small targets are very difficult to detect because of their weak brightness, small size, and lack of shape, structure, texture, and other information elements. In order to simultaneously satisfy the robustness and timeliness of target detection, inspired by density peak clustering and the human visual system, an idea combining an improved density peak global search and local contrast calculation is proposed. First, the positions of candidate targets are determined in the preprocessed image using the improved density peak global search method (IDPGSM). Second, the saliency map is obtained using the double-weights enhanced local contrast method (DWELCM) for the candidate target neighborhood. Finally, adaptive threshold segmentation is used to detect IR small targets. Through comprehensive analysis of five evaluation indicators, the experimental results on seven real sequences and three hundred IR images of different scenes that the proposed method has better detection performance compared with six baseline methods. It can quickly and accurately determine the small target position in the case of severe background clutter and noise interference.

The difficulty of Multi-Agent Reinforcement Learning (MARL) increases with the growing number of agents in system. The value function decomposition is an effective way to alleviate the curse of dimension. However, the existing methods usually either can only provide the low-order approximate decomposition of no more than the second-order, or need to spend a lot of effort manually designing the high-order interaction among agents according to experience. Therefore, the existing methods either tend to bear large decomposition error or are not convenient to use. In this paper, a high-order approximate value function decomposition method is proposed, which has the following prominent characteristics: the low-rank vector is exploited to represent value function, the low-order and high-order components share the same input (i.e., the embedding vector), the attention mechanism is used to select the agents participating in the high-order interaction, and all agents share the model parameters if the agents are homogeneous. To our knowledge, this is the first MARL method modeling low- and high-order interaction simultaneously among agents that can be trained end-to-end. Extensive experiments on two different multi-agent problems demonstrate the performance gain of our proposed approach in comparison with strong baselines, particularly when there are a large number of agents.

Deep Reinforcement Learning (DRL) methods are inefficient in the initial strategy exploration process due to the huge state space and action space in large-scale complex scenarios. This is becoming one of the bottlenecks in their application to large-scale game adversarial scenarios. This paper proposes a Safe reinforcement learning combined with Imitation learning for Task Assignment (SITA) method for a representative red-blue game confrontation scenario. Aiming at the problem of difficult sampling of Imitation Learning (IL), this paper combines human knowledge with adversarial rules to build a knowledge rule base; We propose the Imitation Learning with the Decoupled Network (ILDN) pre-training method to solve the problem of excessive initial invalid exploration; In order to reduce invalid exploration and improve the stability in the later stages of training, we incorporate Safe Reinforcement Learning (Safe RL) method after pre-training. Finally, we verified in the digital battlefield that the SITA method has higher training efficiency and strong generalization ability in large-scale complex scenarios.

To locate microseismic sources, the reverse double-difference time imaging method is proposed in this paper. The traditional time-reversal imaging method requires an accurate velocity model. To reduce the influence of the velocity model errors, this method uses the reverse travel time difference from adjacent receivers to each grid point of the velocity model and the arrival time difference of the microseismic event for imaging. We propose two imaging conditions for single source imaging and multi-source imaging to obtain high-resolution images of the sources. These imaging conditions are robust and can eliminate the error caused by outliers in the reverse double-difference time. The k-means clustering algorithm is used to optimize positioning results. Numerical experiments for source location estimation demonstrate that the reverse double-difference time imaging method is stable and reliable, and can locate multiple seismic sources with high-precision when the velocity model has errors. Some factors affecting the accuracy of source location estimation in the reverse double-difference time imaging are tested using 2D numerical experiments.

During the long-term operation of high-voltage cross-linked polyethylene (XLPE) cables, the cable will gradually deteriorate under the effect of various aging factors. Thermal aging is one of the key factors directly related to the long-term operation reliability of the XLPE cables. In this paper, the thermal aging samples are prepared for PDC measurement. Based on the polymer trap theory and the extended Debye model, the shape of PDC curve, depolarization charge, parameters of the extended Debye model, aging factor (A), elongation at break retention rate (EB%) and their relationships under different thermal aging degrees are analyzed. The results show that the depolarization current curve and the depolarization charge can preliminarily judge the aging condition of the XLPE, but it is not accurate enough. $R_{3}$ and $\tau _{3}$ of the branch, with the largest time constant must be the most sensitive to cable aging. With the increase of aging degree, the A shows an upward trend, and has a nonlinear relationship with EB%. Under the aging temperature of 140 °C, after 15–25 days of aging, the A suggests the samples are in the moderate to severe aging stage, and the PDC test should be performed on carried out for the cables with that have been used for 21 years of operation to evaluate the insulation aging condition. The study reveals that the PDC-based parameters can be serviced as an effective tool for the aging condition prediction of XLPE cable insulation.

In this paper, a design of a silicon mode (de)multiplexer based on cascaded counter-tapered couplers is proposed and investigated. By using the particle swarm optimization algorithm and finite difference time domain method, structural parameters of each counter-tapered coupler in our proposed mode (de)multiplexer are optimized so that high conversion efficiencies can be obtained and coupling lengths can be significantly shortened. The coupling lengths of counter-tapered couplers in the proposed silicon four-mode (de)multiplexer operating in transverse-electric polarization are 8 μm, 9 μm, and 10 μm. The corresponding conversion efficiencies are respectively 98.98%, 99.98%, and 98.68%. Experimental results show that, the fabricated device exhibits a demultiplexed crosstalk of -32.76 dB, -27.10 dB, -32.45 dB, or -33.29 dB at 1550 nm wavelength, when the fundamental transverse-electric mode is respectively launched into the ports Input0, Input1, Input2, or Input3. The measured demultiplexed crosstalk is less than -15.15 dB from 1525 to 1596 nm and lower than -18.12 dB from 1525 to 1561 nm.

The environment and climate significantly affect the land surface temperature (LST) of a city. Previous studies have revealed that LST exhibits significant spatial heterogeneity primarily caused by a combination of natural factors and human activities. Based on this, the introduction of point of interest data of the “production–living–ecological space” divides the influencing pattern into a comprehensive description of human activities supplemented by natural factors, resulting in the precise influencing factors of spatial heterogeneity of LST. Taking Nanjing (Jiangsu Province, China) as a case study, this study uses Landsat-8 remote sensing images, point of interest data, and other data to establish a geographically weighted regression model that combines natural factors and human activities. The main research results are as follows: First, the LST of Nanjing ranged from 19.9 °C to 47.6 °C, whereas the distribution trend was “low at both ends and high in the middle.” Second, there is no multicollinearity of the influencing factors, the fitting degree of LST and each influencing factor reached 0.87. The regression coefficients were high and exhibited both positive and negative values, implying that spatial heterogeneity exists among the influencing factors and LST. Finally, the ranking of how all factors influence the LST followed the order of water area > forest and grassland > ecological space > slope > production space > elevation > living space. The research results have practical significance for improving the quality of life of urban residents and providing a critical theoretical basis for optimizing urban human settlements.

Automatic Feature Recognition (AFR) is considered as the key connection technique of the integration of Computer Aided Design (CAD) and Computer Aided Process Planning (CAPP). At present, there is a lack of a systematic method to identify and evaluate the local features of 3D CAD models. The process information such as topological structure, shape and size, tolerance and surface roughness should be considered. Therefore, a novel Model Based Definition (MBD) based on 3D CAD model AFR and similarity evaluation are proposed in this paper. A Multi-Dimensional Attributed Adjacency Matrix (MDAAM) based on MBD is established based on the fully consideration of the topological structure, shape and size, surface roughness, tolerance and other process information of the B-rep model. Based on the MDAAM, a two-stage model local feature similarity evaluation method is proposed, which combines the methods of optimal matching and adjacency judgment. First, the faces of source feature and target model are used as independent sets to construct a bipartite graph. Secondly, supplement the vertices in the independent set of source feature to make the number of vertices in two independent sets equal. Thirdly, based on MDAAM data, the weighted complete bipartite graph is constructed with the face similarity between two independent sets as the weight. Fourthly, Kuhn-Munkres algorithm is used to calculate the optimal matching between the faces of source feature and target model. Fifthly, the adjacency between matching faces in target model is judged. Finally, the similarity between matching faces of the two models is calculated, which is used as the similarity evaluation result. The effectiveness of this method is verified by three applications.

Well-testing stage analysis is an important way for oilfield operation state decision-making and reservoir management. However, due to the variability and nonlinearity of the downhole data caused by the complex exploration activities and the differences of petroleum type and geological conditions, the classical methods are ineffective in feature extraction, learning network construction, and classifier optimization. In this work, we propose a new well-testing stage classification method based on a deep vector learning model (DVLM). The novelty of this study lies in the combination of multi-feature extraction, deep learning and feature vector mapping. The proposed method can overcome the problems of poor feature representation ability and poor classification model generalization ability in the existing machine learning methods, which mainly caused by the non-optimized training network structure and the unreasonable classifier design,. Firstly, the initial features are obtained by four classical methods. Then a five layers deep belief network embedded with the mutual information coefficient method is implemented for further feature extraction and purification. Finally, the optimized learning vector quantization classifier outputs the predicted tags. For model training and testing 572 field samples total of 4004 data streams are used. By considering the classification errors and accuracy metrics, the neurons number of deep learning network and the classifier are tuned, and an optimal and stable framework is obtained. Comparative experimental results with several classical integration models show that the proposed model achieves the highest classification accuracy of 98.065% as well as the least of features (nine). The results demonstrate that the proposed model has excellent performance in improving the classification accuracy and completing the feature compression. Moreover, the proposed model has very important practical significance for guiding the automatic analysis and processing oil and gas data.

Owing to the population aging and a change in life patterns, the number of stroke incidents in China has increased dramatically with three million new occurrences every year. While the demand for rehabilitation services in China is enormous, the existing rehabilitation resources are not only disparately distributed but also unable to cope with the ever-increasing number of patients. In this paper, we present a study that focuses on a long-term limb function recovery program that uses a wearable body area network (WBAN) for in-home people with stroke discharged from hospital. The investigation lasted for 60 weeks and involved 12 people with stroke sparsely located in Jiaxing, Zhejiang province, China. These patients, who were all within their first 3 months from stroke event, were evenly and randomly split into the experiment group and the control group. The Brunnstrom Stage (BS) and the Mobility Index (MI) were used to monitor the patients' recovery processes. The results demonstrated that the patients from the experimental group had experienced a steady increase in MI throughout the program. They also had improved BS by at least one stage, which outperformed the control group in average. This longitudinal efficacy investigation indicates that the supported in-home rehabilitation system has the potential to reduce the cost and the effort for patients and doctors while still maintaining the quality and effectiveness of rehabilitation; hence could be a possible solution in alleviating the great stress that the healthcare system is currently experiencing in China nationwide.

In this paper, a chamfer distribution antenna array is designed to generate a deflecting circular-polarized vortex beam carrying orbital angular momentum (OAM) mode. When the vortex beam is designed with a deflection angle, the sidelobe of the vortex beam will deteriorate. In order to solve this problem, an optimized array design is presented in this paper. Inspired by the Chebyshev distribution array synthesis method, the antenna elements in the four corners of the antenna array are cut out to realize the chamfer distribution array. Therefore, the sidelobe level of the deflected vortex beam can be effectively suppressed. Moreover, a simplified feeding network is designed to reduce the feeding network complexity by using circularly polarized structural center antenna element. The designed antenna array is simulated, fabricated, and measured at center frequency of 4.25 GHz in C-band. The far-field radiation pattern of the deflected vortex beam and the near E-field OAM vortex beam distribution are simulated and measured to verify the effectiveness of the proposed design method for sidelobe suppression array.

In this paper, a fuzzy logic control strategy is proposed for solving trajectory tracking control issues of an uncertain manipulator. Fuzzy logic is utilized to compensate for nonlinear uncertainties in manipulator dynamics and full-state constraints are involved in full-state feedback controller design for ensuring motion constraints during movement processes. Disturbance observer (DO) is designed to counteract the effects of unknown nonlinear disturbances caused by friction force or other various forms of disturbance. Combining with Lyapunov theory and back-stepping method, the proposed method can guarantee error signals in closed-loop system semi-globally uniformly bounded (SGUB). In view of safe operation, tangent barrier Lyapunov functions (tBLFs) are chosen to maintain joint angle and velocity in a predefined constrained region in tracking processes. Finally, simulation results are carried out to show the effectiveness of our proposed control strategy.

A novel method of using the spiking neural networks (SNNs) and the electroencephalograph (EEG) processing techniques to recognize emotion states is proposed in this paper. Three algorithms including discrete wavelet transform (DWT), variance and fast Fourier transform (FFT) are employed to extract the EEG signals, which are further taken by the SNN for the emotion classification. Two datasets, i.e., DEAP and SEED, are used to validate the proposed method. For the former dataset, the emotional states include arousal, valence, dominance and liking where each state is denoted as either high or low status. For the latter dataset, the emotional states are divided into three categories (negative, positive and neutral). Experimental results show that by using the variance data processing technique and SNN, the emotion states of arousal, valence, dominance and liking can be classified with accuracies of 74%, 78%, 80% and 86.27% for the DEAP dataset, and an overall accuracy is 96.67% for the SEED dataset, which outperform the FFT and DWT processing methods. In the meantime, this work achieves a better emotion classification performance than the benchmarking approaches, and also demonstrates the advantages of using SNN for the emotion state classifications.

Wireless body area network is a low-power and high-security wireless communication protocol specially used in the medical environment. It requires a highly secure encryption algorithm to ensure that information related to a patient's vital signs cannot be accessed by an attacker. In this article, a security scheme based on chaotic mapping for WBAN is proposed. This algorithm makes use of the hybrid chaotic system obtained by the combination of Logistic and Tent mapping as the sequence generator, which can select the effective precision according to the importance of the transmitted information to ensure the reasonable allocation of computing resources, including floating-point encryption, fixed-point 32-bit encryption, fixed-point 24-bit encryption, and fixed-point 16-bit encryption. In the randomness test of the sequence, it can be seen that the sequence autocorrelation coefficient generated by the combined chaotic system is close to 0 and has a flat power spectrum. At the same time, the sequences produced by the floating-point system and 32-bit fixed-point system have passed the NIST test, which proves that combined chaos overcomes the influence of sequence length and finite precision. Experimental analysis and performance comparison show that the proposed encryption algorithm has sufficient security to resist common attacks.

The ultimate goal of military intelligence is to equip the command and control (C2) system with the decision-making art of excellent human commanders and to be more agile and stable than human beings. Intelligent commander Alpha C2 solves the dynamic decision-making problem in the complex scenarios of air defense operations using a deep reinforcement learning framework. Unlike traditional C2 systems that rely on expert rules and decision-making models, Alpha C2 interacts with digital battlefields close to the real world and generates learning data. By integrating the states of multiple parties as input, a gated recurrent unit network is used to introduce historical information, and an attention mechanism selects the object of action, making the output decision more reliable. Without learning human combat experience, the neural network is trained in fixed- and random-strategy scenarios based on a proximal policy optimization algorithm. Finally, 1,000 rounds of offline confrontation were conducted on a digital battlefield, whose results show that the generalization ability of Alpha C2 trained using a random strategy is better, and that it can defeat an opponent with a higher winning rate than an Expert C2 system (72% vs 21%). The use of resources is more reasonable than Expert C2, reflecting the flexible and changeable art of command.

We present demonstration of an erbium-doped fiber laser based on the nonlinear amplifying loop mirror mechanism, with a stable phenomenon of flat-top Q-switched mode-locked (FT-QML) dissipative soliton resonance (DSR) pulse, for the first time. In the experiment, we can observe both the flat-top Q-switched envelope and square-wave pulse by the peak power clamping effect inside the cavity. The characteristics of both continuous DSR pulse and flat-top Q-switched mode-locked square-wave are analyzed comprehensively. With the maximum pump power of 363 mW, single pulse energy in the FT-QML envelope can be up to 14 nJ as three times high as the one of continuous DSR. Essentially, the temporal width of the Q-switched mode-locked envelope can reach 4.6 μs with flat-top duration of 3.5 μs. The amplitude fluctuation of the sub-pulse inside the flat-top is stable at about 0.12% by using the formula of RMSE. Therefore, the laser can provide flat-top burst mode pulse sequence, and the pulse width can be controlled by tuning the pump power.

Vehicular fog and cloud computing (VFCC) system, which provides huge computing power for processing numerous computation-intensive and delay sensitive tasks, is envisioned as an enabler for intelligent connected vehicles (ICVs). Although previous works have studied the optimal offloading scheme in the VFCC system, no existing work has considered the departure of vehicles that are processing tasks, i.e., the occupied vehicles. However, vehicles leaving the system with uncompleted tasks will affect the overall performance of the system. To solve the problem, in this paper, we study the optimal offloading scheme that considers the departure of occupied vehicles. We first formulate the task offloading problem as an semi-Markov decision process (SMDP). Then we design the value iteration algorithm for the SMDP to maximize the total long-term reward of the VFCC system. Finally, the numerical results demenstrate that the proposed offloading scheme can achieve higher system reward than the greedy scheme.

Due to manufacturing errors and material deteriorations in the metal rubber of clamps, the clamp stiffness of the pipeline is uncertain. This paper presents a non-intrusive multi-dimensional Chebyshev polynomial approximation method (M-CPAM) to evaluate uncertain characteristics of the frequency response function (FRF) of the clamp-pipeline system (CPS), where the clamp stiffness parameters are taken as unknown but bounded interval variables. Firstly, a finite element model of the clamp-pipeline system is established. Secondly, the variance-based global sensitivity analysis is implemented to determine significant stiffness parameters. Then, the uncertain intervals of the clamp stiffness are measured by experiments and the dispersion of the clamp stiffness is described. Finally, based on the measured stiffness interval, the uncertain frequency responses of the CPS under different tightening torques are analyzed by the proposed M-CPAM, and the effectiveness of simulation results is verified by experiments. Compared with the results obtained from the Monte Carlo simulation, the experimental measurements, and the polynomial chaos expansion, the proposed M-CPAM provides a more accurate, time-saving and practical method for solving the uncertain frequency responses of the CPS with interval stiffness variables. The results show that the clamp stiffness has great dispersion under the same tightening torque. A frequency shift phenomenon will be observed when the clamp stiffness is uncertain. Moreover, the dispersion of the frequency response of the CPS tends to be concentrated with the increase of the tightening torques.

Full waveform inversion (FWI) is a powerful method to reconstruct the properties of the subsurface media. However, the standard FWI is a non-unique and ill-posed inversion problem, which requires proper techniques to avoid cycle skipping phenomena. Sufficient low-frequencies in the observed data and a good initial model are helpful to mitigate cycle skipping problem, but they are hard to be provided in real cases. Therefore, to reduce the non-linearity and improve the convergence of FWI, we developed a novel approach inspired by using the convolutional neural network to mitigate the cycle skipping problem. We use the 1-dimentional (1-D) convolution kernels of different lengths to convolve each seismic trace of the synthetic and observed data to extract the different features of each time sample, and then we use the Sigmoid function to encode each time sample of the synthetic and observed data according to the polarity of the features. By comparing the coding similarity for the time sample of the synthetic and observed data at the corresponding time, we can identify which part of the synthetic data is well matched with the observed data and which part is mismatched. For the mismatched synthetic data, we attenuate them to reduce their interference on the gradient, thereby the cycle skipping problem can be mitigated. In this case, we use the global-correlation misfit function which behaves better in mitigating the interference of the incorrect amplitude information and highlighting the phase information with weaker non-linearity. In addition, the convolution coding and amplitude attenuation-based method has a strong anti-noise capability and can be combined with the encoded multisource scheme to save the computational costs. Marmousi model tests demonstrate that convolution coding and amplitude attenuation-based FWI behaves better than the standard FWI in generating convergent inverted result.

In this paper, we propose a method to generate a high-order Bessel orbital angular momentum (OAM) vortex beam by using a single-layer reflective metasurface, which integrates the advantages of small size, high efficiency, insensitive polarization and stable incident angles. An offset-fed horn configuration is adopted to overcome the feed-blockage effect, and a subwavelength unit cell with rotational symmetry is designed to cover the reflection phase variation range of 360° in different incident angles. The metasurface is simulated, fabricated and measured at the center frequency of 10 GHz, and the results validate that a second-order Bessel OAM vortex beam can be generated effectively. By comparing the field distributions between the Bessel OAM vortex beam and the conventional OAM vortex beam, we find that the generated Bessel OAM vortex beam is obviously more convergent and has more stable field distributions. The single-layer reflective metasurface is simple and flexible to shape the wavefront to be a Bessel OAM vortex wave, which has the potential to be used in a wide range of applications.

In order to synchronize the motions of the above-knee amputee's limbs and smart prosthetic knee in practical applications, the principle of a Rayleigh oscillator-based reference angle generator (RORAG) for the motion control of smart prosthetic knees is proposed. The central pattern generator (CPG) model of the lower limb in this paper uses a pair of phase-coupling Rayleigh oscillators to imitate the human's lower limb's biological CPG, where the used Rayleigh oscillators are improved to exactly represent the motion characteristics of the lower limb. A frequency control method for the CPG model is proposed to synchronize the thigh reference angle, which is generated by the lower limb's CPG model, with the swing angle of the above-knee amputee's residual thigh. The RORAG model is developed by using the real-time simulation system. An experimental system is developed to verify the performance of the RORAG model in two terms of the curve shape and motion synchronization. Further, in order to estimate the application value of the proposed RORAG model in two terms of the real-time and accuracy, a bio-guided motion platform system is developed and fabricated. On this motion platform system, the RORAG model is used to control a prosthetic knee prototype. The experimental results show that the developed RORAG model can imitate the subject's knee joint's swing angle accurately in real time. Therefore, the proposed RORAG is effective in practical applications.

The Industrial Control System (ICS) is widely used in industrial processes, such as power grids, water conservancy, natural gas, petrochemical and so on. More and more cyber attacks are targeting the ICS worldwide. This paper presents a novel honeynet-based epidemic model in ICS network. The honeynet is an active approach that can attract malware attacks and provide sample information and immunization strategy of the malware. An epidemic model with immunization and quarantine in ICS network is formulated to explore the dynamics of the malware propagation, and the honeynet potency is analyzed as well. Theoretical analysis reveals the disease-free and endemic equilibrium of our model, then the local and global stability of the disease-free (endemic) equilibrium are examined by the basic reproduction number. Furthermore, numerical experiments show that the honeypot with more system vulnerabilities is conducive to suppress the malware epidemic, and the honeynet with lower average degree power low index can be more effectively. In addition, simulation experiments provide the actual behavior of malware propagation in the ICS network and verification of our derivations.

Pose estimation is an important issue for the visual servoing of multicopters. With a process model customized for multicopters and a generic camera model, an extended Kalman filter (EKF) method is proposed in this paper to estimate the pose of multicopters based on off-board multiple cameras. In consideration of EKF implementation, a general point correspondence method that can deal with any number of feature points and cameras is further designed. The proposed EKF method is applicable to not only a single fish-eye (or conventional) camera but also multiple fish-eye (or conventional) cameras. Observability analysis and experiment results show the proposed EKF method is more robust against noise and marker missing than existing filtering methods.

A switchable polarization beam splitter utilizing GST-on-silicon waveguides is introduced and comprehensively investigated in this paper. Based on the phase matching and mode hybridization, structural parameters of the presented device are optimized by using the particle swarm optimization algorithm and finite difference time domain method, so that the output fundamental transverse-electric (TE0) and transverse-magnetic (TM0) modes can be switched by controlling the phase state of GST and the device exhibits broad bandwidth, low crosstalk, and high polarization extinction ratio. For the designed device with the amorphous GST, the crosstalk is lower than -23.40 dB and the polarization extinction ratio is larger than 21.35 dB over the C-band. In the case of the crystalline GST, within the C-band, the crosstalk of less than -27.37 dB and the polarization extinction ratio of greater than 21.37 dB are achieved.

Locally repairable codes (LRCs) are implemented in distributed storage systems (DSSs) due to their low repair overhead. A linear code $\mathcal {C}$ is said to have $(r,\delta)$ -locality if for each coordinate $i$ , there exists a punctured subcode of $\mathcal {C}$ with support containing $i$ , whose length is at most $r+\delta -1$ , and whose minimum distance is at least $\delta $ . An LRC is called optimal if its minimum distance attains Singleton-type bound was proposed. In this letter, optimal LRCs are considered. We first determine $(r,\delta)$ -locality of three dimensional Simplex code, then using anticode strategy, a class of $[{3q,3,2q-1}]_{q}$ LRCs with $(2,q)$ locality are derived for general $q$ . Finally, using an ovoid in $PG(3, q)$ , we construct $[q^{2}+1,4,q(q-1)]_{q}$ and $[{4q-4,4,3q-5}]_{q}$ LRCs with $r=3$ and $\delta =q-1$ . All LRCs constructed in this letter attain the Singleton-type bound.

In this short paper, our main objective is to construct binary locally repairable codes (LRCs) with good properties. Two constructions of LRCs with short lengthes are proposed. The first one is that define a $u$ -linearly independent set ( $u$ -LIS) of an LRC with disjoint repair groups (DRGs) and enlarge it into another one with bigger size to construct new LRCs. The second is puncturing check matrices of known codes to construct new LRCs. As an application, many new binary LRCs are constructed from distance optimal linear codes, which are also locality optimal according to the Cadambe-Mazumdar (C-M) bound.

In this paper, we present a computer-supported method of searching for quantum caps. By means of this method and relevant knowledge of combinatorics, many quantum caps in $PG(3,9)$ and $PG(4,9)$ are constructively proven to exist. Then, according to the theorem that each quantum cap corresponds to a quantum error-correcting code with $d=4$ , we obtain 278 quantum error-correcting codes. Most of these results break the GV bound, and a number of them are optimal quantum codes or have improved parameters.

The data-driven convolutional neural networks (CNNs) have achieved great progress in Synthetic Aperture Radar automatic target recognition (SAR-ATR) after being trained in a large scale of labeled samples. However, the insufficiency of labeled SAR data always leads to over-fitting, causing significant performance degradation. To solve the mentioned problem, a semi-supervised transfer learning method based on generative adversarial networks (GANs) is presented in the present paper. The discriminator of GAN with an encoder and a discriminative layer is redesigned to make it capable of learning the feature representation of input data with unsupervised settings. Instead of training a deep neural network with the insufficient labeled data set, we first train a GAN with varieties of unlabeled samples to learn generic features of SAR images. Subsequently, the learned parameters are readopted to initialize the target network to transfer the generic knowledge to specific SAR target recognition task. Lastly, the target network is fine-tuned using both the labeled and unlabeled training samples by a semi-supervised loss function. We evaluate the proposed method on the MSTAR and OpenSARShip data set with 80%, 60%, 40%, and 20% of the training set labeled, respectively. The results suggest that the proposed method achieves up to 23.58% accuracy enhancement over the random-initialized model.

The accuracy performance of traditional direction of arrival (DOA) estimation algorithms is seriously affected by the reverberation. Considering the advantage of the sparse characteristic of speech signal in time-frequency (T-F) domain, this paper presents a new blind DOA estimation method based on integrated deep learning and convolutional non-negative matrix factorization (NMF). Firstly, mathematic models of microphone array and room impulse response are built. In addition, we extracted blindly initialization parameters of 2-D convolutional NMF using k-means clustering algorithm and singular value decomposition algorithm, which can be used to accurately estimate the main components of desired sound source in the reverberation environment of multi-path propagation. Moreover, the feedback mechanism is introduced into deep 2-D convolutional NMF and correlation coefficient between the signal decomposed by NMF and the signal to be decomposed is used to select the best separated signal for DOA estimation, which make the separation algorithm simpler and more efficient. Finally, test of orthogonality of projected subspaces (TOPS) algorithm is used to validate the DOA estimation capability of this algorithm. Compared with the unprocessed reverberation speech, the estimation error is reduced, which shows that the proposed algorithm can effectively improve the estimation accuracy of DOA estimation when the received signals are in a reverberant environment.

Permanent magnet synchronous motors (PMSMs) are widely used in the field of industrial servo control, especially in high-precision applications. Owing to the periodic torque ripple caused by the cogging torque, flux harmonics, and current offsets, the speed output of the system has a periodic ripple, which affects the control accuracy of the servo system. The conventional proportional-integral controllers cannot reject torque ripple and are highly dependent on motor parameters. This limits the control performance when a PMSM is used as a high-precision servo system. Thus, this paper proposes a combination of model predictive control (MPC) and iterative learning control (ILC) to not only speed up the response time of the system but also effectively reduce the speed ripples. MPC updates the predictive model in real time through feedback and evaluates the system output and control rate according to the cost function. It obtains an optimal control sequence for the next moment and has good parameter robustness and fast response. ILC records the speed of ripple signals over an entire cycle and then uses those signals to compensate for the control signal in the next cycle. It is capable of reducing the periodic speed ripples. The experimental verification of the schemes was conducted on a digital signal processor-field programmable gate-array-based platform. The experimental results obtained confirm the effectiveness of the proposed MPC-ILC scheme.

In this paper, we propose a kind of variable scale aperture sampling reception (VSASR) method for multiple orbital angular momentum (OAM) modes vortex wave reception and demodulation. The VSASR method overcomes the restrictions of equal interval OAM modes which is necessary for the existing partial aperture sampling reception method for keeping the OAM modes orthogonality. Thus, it has higher OAM modes multiplexing utilization efficiency than the partial aperture sampling reception method at the same reception aperture. It is also well for vortex wave consistency reception. Although the orthogonality is a little damaged and there are some other crosstalk OAM modes, the main transmission OAM modes can be kept and distinguished by introducing some specific digital signal processing methods. Finally, for a verification of the presented VSASR method, several reflective metasurface antennas working at 10 GHz are designed, fabricated, and measured to generate different OAM vortex waves. Both the electromagnetic simulation and the measurement experiments are consistent well with the proposed theory analysis, which demonstrates the correctness and effectiveness of the proposed method.

Regaining the lost functionality of limbs is the top priority for people with motor skills impairment as it directly affects their ability to execute activities of daily living and hence, worsens their quality of life. In the last two decades, a great deal of research has focused on error-related potential (ErrP) based brain-computer interfaces (BCIs). Many applications have been developed to assist motor-impaired people in their rehabilitation and among these are robots, spellers, gesture recognition systems, and brain-controlled wheelchairs. In this paper, we present a review of various ErrP based BCI that can potentially aid motor-disabled people in their rehabilitation and execution of their daily activities. First, we describe the ErrP phenomenon and its characteristics followed by a comprehensive application-driven discussion on ErrP based rehabilitation and assistive strategies for motor-impaired people, including studies conducted since the inception of ErrP to the current state-of-the-art applications. Lastly, we discuss the potential issues and challenges being faced by current state-of-the-art applications as well as important future pathways and research directions that might be adopted for advanced ErrP-BCIs used in clinical settings.

Space solar cell components always suffered from the complex and hostile space environment. The reliability of space solar cell components is important to ensure the mission success of space satellites. However, the current research mostly focuses on solar cells' degradation mechanisms. There is less research based on the performance parameters of space solar cell component's degradation to analyze its reliability. Therefore, a reliability analysis method based on the method of modeling the performance degradation of space solar cell components is proposed. By combining general stress-strength interference model and Gamma process, the reliability model of space solar cell components is established. The accelerated degradation test based on the equivalent particle dose is performed to simulate the radiation influence and verify the proposed method. Compared with the displacement damage dose method, the proposed method can further develop the reliability analysis based on performance degradation analysis. Moreover, the cover glass and interconnection are also taken into account together to form a completed component analysis. Through analyzing the cover glass and interconnection experiment results, the cover glass can affect the component reliability; whereas, the interconnection has little influence at the particle radiation condition. Based on the proposed method, the reliability is easier to calculate with the degradation of output power.

Binocular intersection is practically indispensable for stereo visual systems that extract 3D information from images of two distinct cameras. This paper proposes a multi-UAV binocular intersection approach where the two distinct cameras are fixed separately on multiple flying unmanned aerial vehicles (UAVs), rather than physically connected with each other as in the pre-existing works. Eventually, both modeling and algorithms are developed for the multi-UAV binocular intersection with extremely limited communication. As the formulated model states, once one flying vehicle detects the target and shares related information only in a one-shot manner, the other vehicles steer their pan-tilt units (PTUs) to reconstruct a spatial binocular intersection automatically. The binocular intersection algorithm is further improved from two aspects to maximize the probability of target capture. On one hand, the search scope is narrowed by developing an iterative algorithm with priori terrain height and variance. On the other hand, an approaching navigation law is designed to reduce view occlusion caused by terrain. Extensive simulations with different target speeds and flight altitudes are carried out to evaluate the effectiveness of the proposed method. The results show that the probability of binocular intersection is over 90% and is subject to little fluctuation against episodic parameter changes. Furthermore, the developed algorithms are verified localization-free and time-saving as well, and hereafter are potentially practicable for micro aerial swarm applications with extremely limited onboard computation and communication.

In this paper, a plasmonic Feynman gate based on suspended graphene nano-ribbon waveguides at THz wavelengths is proposed and numerically investigated. The proposed plasmonic Feynman gate is composed of two cascaded microring resonators. The resonant states of microring resonators are flexibly controlled by tuning chemical potentials of graphene in rings, and thus, the logic operation results of Feynman gate can be performed successfully. Compared to the plasmonic Feynman gate based on graphene nano-ribbons deposited on the Si substrate and the plasmonic Feynman gate based on graphene nano-ribbons deposited on the SiO2/Si substrate, our designed plasmonic Feynman gate can have better extinction ratio and crosstalk. Calculation results exhibit that the extinction ratio is larger than 15.39 dB and the crosstalk is lower than -15.55 dB when input logic states of our designed device are `00', `01', `10,' and `11'.

A complex human-machine system (CHMS) consists of heterogeneous components with extensive human-machine interactions. CHMSs are typical multistate systems with the ability to adapt to disturbances such as machine failures. These characteristics must be considered comprehensively to accurately evaluate the reliability and performance of a CHMS. However, the existing literature scarcely considers both the reliability and performance simultaneously. In this paper, we propose an agent-based approach to model and evaluate a CHMS. First, a general agent-based modeling framework for a CHMS is generated by analyzing the structure and operations of a CHMS. Then, a dual-clock mechanism is introduced to describe the behaviors of the machine failures and human errors. Two environmental disturbance modeling methods are proposed based on the state transitions of the agent and random events. The methods to model the repair and reconfiguration behaviors are presented based on the contract network. A Monte Carlo-based method is developed to evaluate the reliability and performance of the CHMS simultaneously. Finally, a deck scheduling process for an aircraft carrier is used as a case study to verify the approach. The results show that the reliability and performance of a CHMS can be effectively evaluated.

Deep convolutional neural networks (CNNs) have made a breakthrough on supervised SAR images classification. However, SAR imaging is considerably affected by the frequency band. That means a neural network trained on a SAR image set of one band is not suitable for the classification of another band images. As manually labeling the training samples of each band is always time-consuming, we propose an unsupervised multi-level domain adaptation method based on adversarial learning to solve the problem of multi-band SAR images classification. First, we train a discriminative CNN using samples of one frequency band data set that contains labels to map the data to a latent feature space. Then, we adjust the trained CNN to map the unlabeled samples of another frequency band data set to the same feature space through alternately optimizing two adversarial loss functions. Thus, the features of these two band images are fused and can be classified by the same classifier. We checked the performance of our method using both simulated data and measured data. Our method made a breakthrough in the classification of multi-band images with accuracies of 99% on both data sets. The results are even very close to the supervised CNN trained using a large number of labeled samples.

Camera extrinsic calibration is an important module for robotic visual tasks. A typical visual task is to use a robot and a color camera to pick an object from a variety of items and place it in a designated area. However, the noise of multi-sensor processing may have a significant impact on the results when running a full-process visual task; in addition, checkerboards are inconvenient or unavailable in pick-and-place scenarios. In this paper, we propose and develop a task-oriented markerless hand-eye calibration method by using nonlinear iterative optimization. The optimization employs a transfer error to construct cost function, which is necessarily observable and estimable for visual tasks. Our method does not require a calibration checkerboard and only uses an available saliency object in the task scene as a marker. It provides an end-to-end method that converts extrinsic parameters into variables that are optimized with the cost function, making it not only robust to sensors with noise but also able to meet the requirements of the tasks' reconstruction accuracy. Different from classic methods detecting a known size calibration pattern, the input of our method is a batch of image points and the corresponding world points. The results show that the accuracy of our extrinsic calibration method is sufficient for the robot's pick-and-place tasks. The experiments of the competition demonstrate that our method is definitely effective in the desired tasks of vision-in-the-loop automatic pick-and-place scenarios.

This paper reports the optimal speed control of a permanent magnet synchronous motor (PMSM) system. The predictive control method is an effective strategy for a fast dynamic response. The undesirable performance in the presence of system disturbances, including internal model uncertainties and external load disturbances, is analyzed. To achieve a fast response and ensure stronger robustness and improved disturbance rejection performance simultaneously, a robust generalized predictive controller (GPC) with a high-order terminal sliding-mode observer (HOTSMO) is proposed for a PMSM control system. The proposed observer can estimate the unknown disturbances with chattering elimination. A feed-forward compensation based on the estimated disturbances is provided to the predictive speed controller. The simulation and experimental results show that the proposed control method can achieve a better speed dynamic response and a stronger robustness.