Real-time optimal energy management strategy for hybrid mining trucks based on DP-BPNN fusion

This article aims to solve the contradiction between the optimality and real-time energy management of hybrid mining trucks under outdoor conditions and proposes an innovative energy management strategy based on the fusion of dynamic programming (DP) and a backpropagation neural network (BPNN). For the structural optimisation problem of the BPNN, this paper designs a nested genetic optimisation algorithm, which effectively determines the optimal configuration of the number of hidden layer neurons, network weights and thresholds. On this basis, a multi-BPNN model was constructed to address the problem of identifying multiple working conditions in mining cars. The fuzzy C-means clustering algorithm based on a simulated annealing genetic algorithm was used to classify the working condition data to achieve accurate energy management strategy output. The experimental results show that this strategy considerably improves the fuel economy of hybrid mining trucks while ensuring real-time performance, providing new ideas and methods for the energy management of hybrid mining trucks. Through comparative analysis, the correctness and effectiveness of the proposed plan are verified, which is highly important for promoting the green transformation and sustainable development of the mining industry.