Performance optimization of fuel cell hybrid power robot based on power demand prediction and model evaluation

In order to improve the stability, real-time performance and economy of the proton exchange membrane fuel cell (PEMFC) hybrid welding robot system, the system energy optimization was studied by comprehensive performance evaluation and random forest prediction method. On the basis of rule partition, the optimal control strategy was designed based on entropy weight method and cloud model comprehensive performance evaluation method; The random forest prediction method was put into the energy management system, and the model parameters with the least mean square error were determined by particle swarm optimization, and the load power of the robot is predicted. Finally, the evaluation results are applied to the predicted power to further optimize and improve the performance of the hybrid power welding robot system. The experimental results show that the stability of fuel cell power output based on the optimization strategy in this paper is improved by 11.26%, and the hydrogen consumption is reduced by 3.24%. The experimental results show that the energy optimization strategy can not only ensure the high precision and real-time performance of the welding robot system, but also improve the stability and energy economy of the hybrid welding robot system, and reduce the energy consumption.