The enhancement of photovoltaic (PV) arrays through reconfiguration presents a promising avenue for increasing the global maximum power (GMP) and improving overall array performance. This enhancement is achieved by mi...
详细信息
The enhancement of photovoltaic (PV) arrays through reconfiguration presents a promising avenue for increasing the global maximum power (GMP) and improving overall array performance. This enhancement is achieved by minimizing differences between rows, thereby reducing the computational load on Maximum Power Point Tracking (MPPT) systems. However, many existing reconfiguration methods face various challenges, including scalability issues, inadequate shading dispersion, distortion of array characteristics, emergence of multiple power peaks, increased mismatch, and more. In order to overcome these obstacles, this study presents a novel method for array reconfiguration that is modelled after the widely used Kolakoski Sequence Transform in picture encryption. The suggested approach is assessed in eight different scenarios with 9 x 9 and 5 x 5 PV arrays shaded differently. Its performance is compared against seven established techniques. Due to its intelligent reconfiguration aimed at minimizing shade dispersion, the suggested approach consistently outperforms alternative methods. It results in substantial improvements in GMP, enhancing it by 32.79%, 14.98%, 10.15%, and 4.13% for 9 x 9 arrays, and 37.10%, 14.36%, and 9.88% for 5 x 5 arrays across diverse conditions. Furthermore, this study comprehensively investigates three separate Artificial Neural Network algorithms, specifically the Levenberg-Marquardt (LMB), Scaled Conjugate Gradient, and bayesian regularization algorithms for MPPT. A Levenberg-Marquardt Backpropagation-based MPPT controller for a 250Wp standalone PV system is used to validate the effectiveness of the recommended configuration. This integrated approach, which combines reconfiguration and LMB-based MPPT utilizes only two sensors regardless of array size. It achieves accelerated convergence tracking within a short 0.13 s, displaying minimal steady-state oscillations.
In order to solve the problem of low accuracy of traditional artificial neural networks in approximating functional, a coupling-loop nonlinear autoregressive with exogenous inputs neural network estimation model is pr...
详细信息
In order to solve the problem of low accuracy of traditional artificial neural networks in approximating functional, a coupling-loop nonlinear autoregressive with exogenous inputs neural network estimation model is proposed and applied to the state of health forecasting of lithium-ion batteries. Firstly, eight health indicators related to the state of health of lithium-ion batteries are mined. In order to eliminate the harm of weakly correlated independent variables and redundant independent variables to the estimation results, principal component feature extraction is applied to the feature extraction of the eight health indicators. bayesian regularization algorithm is used to learn the weights of the neural network, which solves the problems of slow convergence speed and easy to fall into local extremum of back propagation learning algorithm, and improves the generalization ability of the neural network. In order to verify the rationality of the proposed model and algorithm, the coupling-loop nonlinear autoregressive with exogenous inputs neural network model is used to estimate the state of health of lithium battery under the condition of complete charge discharge test, and the estimation results are compared with other neural network models. The simulation results show that the coupling-loop nonlinear autoregressive with exogenous inputs estimation model using feature extraction method and bayesian network weight algorithm can approach the functional with higher accuracy, and the absolute error and relative error of lithium-ion battery's state of health estimation can be reduced by more than 50% on average, while the mean square error can be reduced by more than 80%. (c) 2021 Elsevier Ltd. All rights reserved.
The heat transfer mechanism and temperature distribution in laser welding applications have a great impact on the quality of the weld bead geometry, mechanical properties and the resultant microstructure characterizat...
详细信息
The heat transfer mechanism and temperature distribution in laser welding applications have a great impact on the quality of the weld bead geometry, mechanical properties and the resultant microstructure characterizations of the welding process. In this study, the effects of pulsed laser welding parameters including the frequency and pulse width on the melt velocity field and temperature distribution in dissimilar laser welding of stainless steel 420 (S.S 420) and stainless steel 304 (S.S 304) was investigated. A comprehensive comparison was conducted through the numerical simulation and artificial neural network (ANN). The results of numerical simulation indicated that buoyancy force and Marangoni stress are the most important factors in the formation of the flow of liquid metal. Also, increasing the pulse width from 8 to 12 ms due to increasing the pulse energy, the temperature in the center of the melt pool increased about 250 C-o. This leads to increasing the convective heat transfer in the molten pool and heat affected zone (HAZ). The temperature difference at a distance of 1 mm from the beam center at both metals at a frequency of 15 and 20 Hz is bout 58 and 75 C, respectively. Furthermore, reducing the frequency to 5 Hz, due to diminishment of thermal energy absorption time, has clearly decreased the weld penetration depth in the workpiece. According to the ANN results, increasing both pulse duration and frequency has the significant effect on increasing melting ratio from 0.4 to 0.8 compared to the other input parameters. The ANN results confirmed that under the same input conditions, because of the differences in thermal conductivity coefficient, absorption coefficient and melting point of the two pieces, S.S 304 has experienced higher temperatures about 10% more than S.S 420. Also, among the 13 back propagation learning algorithms, the bayesian regularization algorithm had the best performance. Among the number of different neurons in the hidden layer, com
Accurate prediction of hydrocarbon production is crucial for the oil and gas industry. However, the strong heterogeneity of underground formation, the inconsistency in oil-gas-water distribution, and the complex flow ...
详细信息
Accurate prediction of hydrocarbon production is crucial for the oil and gas industry. However, the strong heterogeneity of underground formation, the inconsistency in oil-gas-water distribution, and the complex flow mechanisms make hydrocarbon production forecasting (HPF) difficult, which leads to a high level of uncertainty in the prediction results. The explosion of machine learning (ML) methodologies that are capable of analyzing big data shed new light on HPF using production data. In this article, an in-depth review is provided regarding HPF using ML methodologies. Firstly, the merits and drawbacks of traditional HPF methods are analyzed and summarized. Then, the applications of ML algorithms in HPF are reviewed in detail, especially concentrating on artificial neural network, support vector machine, and ensemble learning. For each algorithm, the basic theory and its variants are first introduced, and its applications in HPF are comprehensively demonstrated subsequently. Finally, this article presents the challenge and prospects of machine-learning-based HPF. Sophisticated ML proxy models can be con-structed and employed to deal with an extended type of input data such that improving the efficacy of data utilization. On the other hand, deep learning models designed to handle time-series data can gain more attention. Modeling approaches for multivariate time-series hydrocarbon production data using deep neural networks with similar functionality to LSTM may lead to more accurate and computationally efficient production forecasting.
The accurate cooling load prediction of an air conditioning system is the basis for energy saving optimization. To solve the problems of low accuracy of prediction, and most load predictions focusing on short-time pre...
详细信息
The accurate cooling load prediction of an air conditioning system is the basis for energy saving optimization. To solve the problems of low accuracy of prediction, and most load predictions focusing on short-time prediction that causes reducing the practical significance, the application of improved BP neural networks prediction model is presented in this paper. Training and testing data for prediction model have been generated from DeST (Designer's Simulation Toolkits) with climate data of Beijing. The generalization ability of the model has been strongest based on bayesian regularization algorithm to train data. A case study shows that high accuracy is achieved by using the BPNN prediction model based on bayesianregularization method with the prediction error of 1.18% in predicting the building load for longer time.
The production of biofuel from waste has become an important topic for waste management and reducing its environmental hazard. Tea factory waste is a strong candidate due to its availability and sourceability. This st...
详细信息
The production of biofuel from waste has become an important topic for waste management and reducing its environmental hazard. Tea factory waste is a strong candidate due to its availability and sourceability. This study aimed to reveal the biochemical methane potential (BMP) of tea factory waste (TFW) and spent tea waste (STW). Additionally, the results revealed that both substrates had high biodegradability due to high VS removal. The BMP tests took 49 days under mesophilic conditions with a batch reactor and the cumulative methane yields were 249 +/- 3, and 261 +/- 8 mL CH4/g VS for TFW and STW, respectively. According to prediction data with the selected ANN model, which was 50 hidden layer sizes, trained with bayesian regularization algorithm, the maximum cumulative specific methane yield of the co-digestion was simulated as 468.43 mL CH4/g VS when the ratio of 65 and 35% (w/w by VS) of TFW and STW, respectively. The predicted methane yield for co-substrates was 183% higher than mono substrates. This result revealed that TFW can be a good candidate for biogas production as biofuel for not only its availability and sourceability but also the synergistic effect possible for codigestion.
暂无评论