Improved Multi-Kernel Extreme Learning Machine Optimized by Sparrow Search Algorithm for Photovoltaic Array Fault Diagnosis

Photovoltaic arrays are exposed to complex and harsh environments for a long time, and are prone to multi-type and multi-degree faults. In order to improve the accuracy and efficiency of fault diagnosis, this paper proposes an improved multi-kernel extreme learning machine (IMKELM) diagnosis strategy based on sparrow search algorithm (SSA) optimization (SSA-IMKELM). Firstly, for 15 typical faults, a 9-dimensional feature vector containing key parameters such as current, voltage and fill factor is constructed to quantify fault characterization. Secondly, the global parameter optimization problem of the traditional multi-kernel extreme learning machine is innovatively decomposed into three sub-models. The hierarchical optimization strategy is used to reduce the complexity of multi-parameter coupling: the pre-order sub-model parameters are fixed, and the current sub-model is optimized by SSA to solve the problem of high-dimensional parameter space search step by step. Finally, a multi-condition fault data set is generated based on the simulation platform to verify the diagnostic performance and compare it with other algorithm models. The results show that the proposed method has high recognition rate, and can effectively judge the working state of photovoltaic array.