Summary
The mobile welding robot based on a hybrid power drive system is a system with complex control characteristics such as nonlinearity, strong coupling, and disturbance of working environment. These characteristics have different effects on the control performance. In order to improve control characteristics of the mobile welding robot, it is necessary to study an effective evaluation method to explore the influence of these parameters on the control performance of the mobile welding robot. In the process of seam tracking, the operation indexes of welding robot system are determined by comprehensive evaluation. On the premise of ensuring tracking accuracy and stability, the energy distribution of hybrid power system is optimized to improve results of continuous power supply time. Therefore, a comprehensive evaluation of the welding robot is established, which is combined information entropy and Kernel Principal Component Analysis (KPCA). The evaluation parameters are divided into the first‐level evaluation index and the second‐level evaluation index. The evaluation model is established by using information entropy, and the data are linearized and dimensionally reduced by using the KPCA. The results show that the method can well deal with and analyze the nonlinear relationship between various indexes. At the same time, the numerical relationship between performance indexes can be established. On this basis, a comprehensive performance evaluation theory based on information entropy, and nuclear principal components is established. Finally, the comprehensive evaluation results are brought into the energy management optimization strategy. The experimental results show that this method not only improves the accuracy of weld tracking, but also ensures the stability of the robot and reduces energy consumption.