A Heuristic Diagnostic Method for a PV System: Triple-Layered Particle Swarm Optimization–Back-Propagation Neural Network

Abstract

This paper proposes a heuristic triple layered particle swarm optimization-backpropagation (PSO-BP) neural network method for improving the convergence and prediction accuracy of the fault diagnosis system of the photovoltaic (PV) array. The parameters, open-circuit voltage (V oc ), short-circuit current (I sc ), maximum power (P m ) and voltage at maximum power point (V m ) are extracted from the output curve of the PV array as identification parameters for the fault diagnosis system. This study compares performances of two methods, the back-propagation neural network method, which is widely used, and the heuristic method with MATLAB. In the training phase, the back-propagation method takes about 425 steps to convergence, while the heuristic method needs only 312 steps. In the fault diagnosis phase, the prediction accuracy of the heuristic method is 93.33%, while the back-propagation method scores 86.67%. It is concluded that the heuristic method can not only improve the convergence of the simulation but also significantly improve the prediction accuracy of the fault diagnosis system.

Experiment‐based supervised learning approach toward condition monitoring of PV array mismatch

“…In contrast, the existing results are not general at all. For example, the computer simulation models with SIMSCAPE and MATLAB [28] or simple laboratory apparatuses [21] are far cry from the real case scenarios which undermine the study of condition monitoring; field tests [20,26,43] cannot generate repeatable operating conditions due to the stochastic outdoor environments; some cases lack the process of experimental verification [30]. It is noted that our experiments have envisaged different types of mismatch from a systematic level whilst the existing ones deal only with some specific cases [31,37].…”

“…The existing artificial intelligence tools for condition monitoring depend heavily on the simulation model while field tests of solar power systems are hardly possible due to the uncontrollable environmental variables and the high cost testing [30,35,37]. To overcome the shortcomings, micro-identical solar power stations are developed to make a comparative study possible that are used extensively in the verification of a novel supervised learning-based monitoring scheme.…”

“…However, artificial intelligence strategies in the literature present some drawbacks. First, these artificial intelligence strategies are developed on simulation models with SIMSCAPE [28] and MATLAB [29,30] that are far cry from the real case scenarios. In fact, these strategies are off-line approaches that have not been examined through real-world experiments on a systematic level [21,[31][32][33].…”

Fault detection and diagnosis of GCPV systems using bayesian neural network

“…Thus, the competencies of frequent maintenance, detection as well as diagnosis of faults get to be increasingly significant in the goal to warrant an excellent power production, efficiency, reliability, quality and safety in global PV systems. Where real time fault identification and localization have been a crucial interest by a large number of international scientists proved and discussed in different articles [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20]. The Table 1 below illustrates all these references talking about Artificial Intelligence in fault detection and diagnosis of PV systems with their advantages and limitations.…”

Clustering-Based Self-Imputation of Unlabeled Fault Data in a Fleet of Photovoltaic Generation Systems

“…A kernel extreme learning machine was investigated owing to its fast learning speed and good generalization [40]. Particle swarm optimization-back-propagation (PSO-BP) has been shown to improve the convergence and prediction accuracy of fault diagnosis systems [41].…”