IJCAS

The traditional coal mining machine uses a single-motor system, which will terminate when encountering a hard road header surface because of power limitations. The same problem exists in large radar servo systems and other applications of heavy industrial. To address this issue, this paper develops the multi-motor driving servo system for the coal mining machine, and designs the adaptive optimal torques for the cut-off gear and the multi-motor system. Firstly, the multi-motor driving system for the coal mining machine is modeled. The optimal performance functions of the cut-off gear and the driving motors are presented, and the Nash equilibrium among the optimal torques is defined. Then, based on the given performance functions, the adaptive optimal torques are found by approximate dynamic programming (ADP) technique, which can find the saddle point and optimize the coal mining machine performance. Moreover, the neural network (NN) weight convergence in the ADP structure is investigated. The stability of the multi-motor driven system with the proposed torques is proved. Finally, taking the coal mining machine as an example, the effectiveness of the performance optimization strategies of cut-off gear and multi-driving motors is verified.

Keywords: Approximate dynamic programming, coal mining machine, large inertia systems, multi-motor driving system, optimal control.