A Boustrophedon‐Optimized Neural Network for Autonomous Path Planning in Large‐Scale Photovoltaic Farms

ABSTRACT To address autonomous mowing requirements in photovoltaic power plants, this study addresses critical path‐planning challenges in conventional methods—specifically, low coverage rates and elevated path redundancy in cellular‐decomposition‐based Boustrophedon algorithms, coupled with excessive turning maneuvers and elevation changes in bio‐inspired neural network (BINN) approaches. We propose a hybrid algorithm integrating the reciprocating motion principle of Boustrophedon patterns with BINN optimization. Simulation experiments demonstrate that the proposed algorithm achieves a coverage rate exceeding 94%, reduces path repetition to below 3%, and decreases turning frequency and elevation changes compared with conventional methods. Field tests confirm complete area coverage in practical applications, demonstrating enhanced operational safety and energy efficiency through optimized energy consumption.