Jihua Wang, Wei Zhang

 Chongqing Colleage of International Business and Economics

Abstract：

Due to the high viscosity of the sauce-flavored liquor, industrial robots cannot meet the needs of retorting, and industrial robots can be replaced by the study of trajectory tracking of retort robot. In the retorting process of soy sauce-type baijiu, in order to improve the accuracy of the trajectory tracking of the retorting robot and reduce the energy consumption of the retorting robot, an adaptive optimal torque control is proposed to compensate the uncertainty part by the RBF neural network. First, an ideal nominal model is built, and the uncertainty part of the model is compensated by an RBF network, where the weights of the neural network are given by an adaptive law. The RUN optimization algorithm is then used to seek the minimum motor initial output, i.e., the optimal PD control coefficients and proves its stability by means of Lyapunov theory. Finally, the simulation results show that this control method can perform better trajectory tracking and can complete the distillation of soy sauce-type baijiu.

Key Words：

the retort robot; optimal torque; adaptive control; trajectory tracking