Application of Genetically Optimized ADRC Based on Backpropagation Neural Network

Abstract — The mathematical representation of a two-link robot is a highly nonlinear dynamic equation. It is kind of difficult to design a controller to achieve the angle setting of a robot. Inheriting the merit of independent of system model and overcoming the shortcomings of classic PID controller, Active Disturbance Rejection Controller (ADRC) possesses good performance with relatively simple algorithm. To realize a two-link robot control, the ADRC based on Backpropagation neural network (BP) is researched and the Adaptive Genetic Algorithm (AGA) is used to exercise the initial weight of BP and optimize parameters of the controller. Simulation results prove the effectiveness of the algorithm.