Search

Rainfall–runoff modelling is at the core of any hydrological forecasting system. The high spatio-temporal variability of precipitation patterns, complexity of the physical processes, and large quan...

Sustainability requires increased energy efficiency with minimal waste. The future power systems should thus provide high levels of flexibility iin controling energy consumption. Precise projections of future energy demand/load at the aggregate and on the individual site levels are of great importance for decision makers and professionals in the energy industry. Forecasting energy loads has become more advantageous for energy providers and customers, allowing them to establish an efficient production strategy to satisfy demand. This study introduces two hybrid cascaded models for forecasting multistep household power consumption in different resolutions. The first model integrates Stationary Wavelet Transform (SWT), as an efficient signal preprocessing technique, with Convolutional Neural Networks and Long Short Term Memory (LSTM). The second hybrid model combines SWT with a self-attention based neural network architecture named transformer. The major constraint of using time-frequency analysis methods such as SWT in multistep energy forecasting problems is that they require sequential signals, making signal reconstruction problematic in multistep forecasting applications.The cascaded models can efficiently address this problem through using the recursive outputs. Experimental results show that the proposed hybrid models achieve superior prediction performance compared to the existing multistep power consumption prediction methods. The results will pave the way for more accurate and reliable forecasting of household power consumption., Under consideration at Pattern Recognition Letters

Accurate wind speed forecasting is a fundamental requirement for advanced and economically viable large-scale wind power integration. The hybridization of the quaternion-valued neural networks and stationary wavelet transform has not been proposed before. In this paper, we propose a novel wind-speed forecasting model that combines the stationary wavelet transform with quaternion-valued neural networks. The proposed model represents wavelet subbands in quaternion vectors, which avoid separating the naturally correlated subbands. The model consists of three main steps. First, the wind speed signal is decomposed using the stationary wavelet transform into sublevels. Second, a quaternion-valued neural network is used to forecast wind speed components in the stationary wavelet domain. Finally, the inverse stationary wavelet transform is applied to estimate the predicted wind speed. In addition, a softplus quaternion variant of the RMSProp learning algorithm is developed and used to improve the performance and convergence speed of the proposed model. The proposed model is tested on wind speed data collected from different sites in China and the United States, and the results demonstrate that it consistently outperforms similar models. In the meteorological terminal aviation routine (METAR) dataset experiment, the proposed wind speed forecasting model reduces the mean absolute error, and root mean squared error of predicted wind speed values by 26.5% and 33%, respectively, in comparison to several existing approaches.

In this paper, the octonion neural network is investigated to forecast the short-term solar irradiance. The previous and the next eight values solar irradiance are organized into two octonion values; thereby the network could be constructed. This method not just gives the opportunity to forecast eight values ahead solar irradiance using one octonion input but also takes all the advantages of the octonion domain. The octonion input contains the past values solar irradiance which produces dynamics naturally to the network and decreases the input dimension vector. The octonion training algorithm has eight dimensions rather than one dimension in the real-valued neural networks. Comparison with the real-valued neural networks for forecasting solar irradiance shows that the proposed method is promising to deal with such problem. The optimal structure is used to manage the an autonomous photovoltaic (PV) system that contains the PV modules and the battery bank. The use of the proposed method presents benefits for the number of the used modules and for the battery energy requested as well.

Semantic segmentation necessitates approaches that learn high-level characteristics while dealing with enormous amounts of data. Convolutional neural networks (CNNs) can learn unique and adaptive features to achieve this aim. However, due to the large size and high spatial resolution of remote sensing images, these networks cannot analyze an entire scene efficiently. Recently, deep transformers have proven their capability to record global interactions between different objects in the image. In this paper, we propose a new segmentation model that combines convolutional neural networks with transformers, and show that this mixture of local and global feature extraction techniques provides significant advantages in remote sensing segmentation. In addition, the proposed model includes two fusion layers that are designed to represent multi-modal inputs and output of the network efficiently. The input fusion layer extracts feature maps summarizing the relationship between image content and elevation maps (DSM). The output fusion layer uses a novel multi-task segmentation strategy where class labels are identified using class-specific feature extraction layers and loss functions. Finally, a fast-marching method is used to convert all unidentified class labels to their closest known neighbors. Our results demonstrate that the proposed methodology improves segmentation accuracy compared to state-of-the-art techniques.

A Meta-cognitive Quaternion Valued Neural Network (Mc-QVNN) learning algorithm and its forecasting applications is proposed in this paper. The Mc-QVNN has two parts, the cognitive part that contains the QVNN and a meta-cognitive part, which self-regulates the learning algorithm. At each epoch, when the Mc-QVNN receives a new sample, the meta-cognitive part makes a decision about the manner, the time and the need to learn this sample or not. In this case, the algorithm deletes the unneeded samples and keeps just the necessary ones for learning. The meta-cognitive component makes the decision according to the quaternion magnitude and phases. Three forecasting problems, which are MackeyGlass time series, Lorenz attractor and the real home's power in the city of Honolulu in Hawaii, USA, are taken to test the performance of the proposed algorithm. Comparison with other existing methods shows that the Mc-QVNN is promising for forecasting chaotic systems.

In this paper, a metacognitive octonion-valued neural network (Mc-OVNN) learning algorithm and its application to diverse time series prediction are presented. The Mc-OVNN is comprised of two components: the octonion-valued neural network that represents the cognitive component and the metacognitive component that serves to self-regulate the learning algorithm. At each epoch, the metacognitive component decides if, how, and when learning occurs. The algorithm deletes unneeded samples and only stores those that will be used. This decision is determined by the octonion magnitude and the seven phases. To evaluate the Mc-OVNN algorithm’s performance, it is applied to five real-world forecasting problems: the power consumption of a home in Honolulu, HI, USA, Box and Jenkins J series, Euro to Algerian Dinar (DZ) real-time conversion rates, the Mackey–Glass equation, and Europe Brent oil price prediction in a time series. When comparing the Mc-OVNN to other relevant techniques, Mc-OVNN displays its capability for efficient time series prediction. The real-time evaluation of the proposed algorithm is presented using the power consumption of a home in Boumerdes, Algeria, as a case study.

High penetration of renewable energy poses a significant challenge in operation of power system. A potential solution for this problem is utilizing Battery Energy Storage System (BESS). The purpose of this paper is to analyze the effectiveness of BESS ability to peak shave and smooth the load curve of an actual circuit on the island of Maui in Hawaii. The distribution circuit has about 850 kW of installed rooftop Photo-Voltaic (PV) generation. Higher penetration of PV increases the concern about the potential impacts on the transmission system. At first, we will present two different methods for load forecasting. Reliable forecasting of daily load is required to effectively utilize the BESS system. We have employed two different methods for load forecasting in order to achieve two main purposes including peak shaving and smoothing. For reaching these goals, two approaches are analyzed. The first approach is utilizing a nonlinear programming method in terms of load shifting and smoothing. The second approach includes a real time control strategy to have smoothing and peak shaving at the same time. As a real case study, these proposed methods have been applied within 108-day data collection period and pros and cons of these methods will be discussed.

In this paper, the development of an improved Takagi Sugeno TS fuzzy model for identification and chaotic time series prediction of nonlinear dynamical systems is proposed. This model combines the advantages of fuzzy systems and Infinite Impulse Response IIR filters, which are autoregressive moving average models, to create internal dynamics with just the control input. The structure of Fuzzy Infinite Impulse Response FIIR is presented, and its learning algorithm is described. In the proposed model, the Butterworth analogue prototype filters are estimated using the obtained membership functions. Based on the founding orders of the analogue filters, the IIR filters could be constructed. The IIR filters are introduced to each TS fuzzy rule which produces local dynamics. Gustafson--Kessel GK clustering algorithm is used to generate the clusters which will be used to find the number of the IIR parameters for each rule. The hybrid genetic algorithm and simplex method are used to identify the consequence parameters. The stability of the obtained model is studied. To demonstrate the performance of this modeling method, three examples have been chosen. Comparative results between the FIIR model on one hand, and the traditional TS fuzzy model, the neural networks and the neuro-fuzzy network on the other hand. The results show that the proposed method provides promising identification results.

In this chapter, the prediction of the daily global solar irradiation of the great Maghreb region using the complex-valued wavelet neural network (CVWNN) is presented. Both multi-input single output (MISO) and multi-input multi-output (MIMO) strategies are considered. The meteorological data of the capitals of the great Maghreb, which are Tripoli (Libya), Tunis (Tunisia), Algiers (Algeria), Rabat (Morocco), El Aaiun (Western Sahara), and Nouakchott (Mauritania), are used like samples from each country. To test the applicability and the feasibility of the CWNN to predict the daily global irradiation for the great Maghreb case, several models are presented. Results obtained throughout this chapter show that the CWN technique is suitable for prediction of the daily solar irradiation of the great Maghreb region.

In this chapter, the short-term forecasting of the global solar irradiation using the fuzzy modeling technique is proposed. The multi-input multi-output (MIMO) fuzzy models are used to predict the next 24 h ahead based on the mean values of the daily solar irradiation and the daily air temperature. The measured meteorological data of Tamanrasset City, Algeria (altitude, 1,362 m; latitude, 22°48 N; longitude, 05°26 E) is used, where the 2 years (2007–2008) are used for modeling and the year 2009 is used to validate the developed model. Several models are presented to test the feasibility and the performance of the fuzzy modeling technique for forecasting hourly solar irradiation in the MIMO strategy. Results obtained throughout this chapter show that the fuzzy modeling technique is suitable for a short-time forecasting of the solar irradiation.

In this paper, the fuzzy modeling technique is used to estimate the photovoltaic (PV) maximum power generation. The equivalent circuit parameters of a PV module essentially depend on solar irradiation and temperature. This dependence on environmental factors is investigated and it is shown that it is a nonlinear relationship that cannot be easily expressed by any analytical equation. Therefore, the fuzzy modeling technique is utilized to overcome these difficulties. The Gustafson-Kessel algorithm is used to obtain clusters of the dataset which produces the fuzzy rules to identify the nonlinear photovoltaic maximum power generation and the irradiation. The developed architectures can approximate the maximum power and predict the irradiance parameter. To demonstrate the feasibility and the performance of the proposed architectures, several simulation results have been presented.

In this study, the identification of bioprocesses using Random Search (RS) and Simulated Annealing (SA) algorithms is considered. The high nonlinearity and the large application of bioprocesses give them the opportunity to be challenging systems for several conventional techniques that cannot be used for the identification process. The derivative free algorithms, in which the gradient of the problem to be optimized is not required, could be used for the identification of bioprocesses parameters. Two systems have been chosen for identification using the RS and the SA algorithms which are the fed batch fermentor and the bioreactor. Simulation results are presented for the above cases and are found to be satisfactory compared to the real bioprocesses parameters.

The approximation of general continuous functions by nonlinear networks is very useful for system modeling and identification. Such approximation methods can be used, for example, in black-box identification of nonlinear systems, signal processing, control, statistical data analysis, speech recognition, and artificial intelligence. Recently neural networks have been established as a general approximation tool for fitting nonlinear models from input/output data due to their ability of learning rather than complicated process functions (Gao, 2002). Their attractive property is the self-learning ability. A neural network can extract the system features from historical training data using the learning algorithm, requiring little or no a priori knowledge about the process (Patan, 2008). This is why during the past few years the nonlinear dynamic modelling of processes by neural networks has been extensively studied (Narendra & Parthasarathy, 1990; Nerrand et al., 1993; Levin, 1992; Rivals & Personnaz, 1996). In standard neural networks, the nonlinearities are approximated by superposition of sigmoidal functions (Cybenko, 1989). In the other hand, the wavelet theory has found many applications in function approximation, numerical analysis and signal processing. Though this attractive theory has offered efficient algorithms for various purposes, their implementations are usually limited to wavelets of small dimension. The reason is that constructing and storing wavelet basis of large dimension are of prohibitive cost. In order to handle problems of larger dimension, it is necessary to develop algorithms whose implementation is less sensitive to the dimension. And it is known that neural networks are powerful tools for handling problems of large dimension. Due to the similarity between wavelet decomposition and one-hidden-layer neural networks, the idea of combining both wavelets and neural networks has been proposed in various works (Zhang & Benveniste, 1992; Pati & Krishnaprasad, 1993; Hong, 1992; Bakshi & Stephanopoulos, 1993; Tsatsanis & Giannakis, 1993;et al., 1994; Delyon et al., 1995; Saad Saoud & Khellaf, 2009). For example, in (Zhang & Benveniste, 1992) wavelet network is introduced as a class of feedforward networks composed of wavelets, in (Pati & Krishnaprasad, 1993) the discrete wavelet transform is used for analyzing and synthesizing

The high nonlinearity and the high cost of the Continuous Stirred Tank Reactor (CSTR) system need to realize a low cost embedded system based on an inexpensive microcontroller using the full neural networks. The two parts, training and validation are used to adapt the network parameters in real time as proposed in this paper. The realized card simulates the CSTR model. The well-known backpropagation algorithm is implemented to train a neural network model. Both the training and the validation parts are shown through an alphanumeric liquid crystal display. A comparison was made between the realized embedded system and the computer results.