Research On Parameter Identification And Trajectory Control Of Flexible Joint Robot

In recent years,China's industrial automation technology has developed rapidly,and the automated production line with robot as the core has gained a lot of recognition.But the traditional industrial robot joints and connecting rods are made of rigid materials,and its own weight will influence the performance of the robot.In contrast,flexible joint robot has been widely used in industrial production with its advantages of light weight and low energy consumption.On the other hand,man-machine safety robot with the characteristics of joint flexibility has gradually become a research hotspot with the advent of the era of artificial intelligence.Therefore,this paper studies the dynamics parameter identification and controller design of flexible joint robot,and the main research work is as follows:First of all,the dynamic analysis of flexible joint robot was performed.On the basis of the simplified model of flexible joint robot presented by Spong,the dynamic model of flexible joint robot is established by using Lagrange equation,this lays a foundation for the design of the parameter identification and controller.Secondly,according to the problem of parameter identification of flexible joint robot,the corresponding identification equations are deduced according to the prior model to determine the parameters that need to be identified.Considering the traditional least square method for nonlinear system identification effect is poorer,intelligent algorithm identification strategy is used,an adaptive particle swarm optimization genetic algorithm(APSO-GA)based on particle swarm algorithm and genetic algorithm is proposed,it avoids the problem that the traditional particle swarm algorithm is easy to get into the local optimal solution and further improves the identification precision.Simulation results show that compared with traditional artificial bee colony algorithm,genetic algorithm and particle swarm algorithm,APSO-GA has higher accuracy.Moreover,according to the flexible joint robot trajectory tracking control problem,a fixed-time sliding mode disturbance observer(FTSMDO)is designed,so as to realize accurate estimates of the interference in the fixed-time.Combined with dynamic method,to avoid the traditional inversion control problem of "differential explosion".A fixed-time controller is designed for each subsystem,so that the flexible joint robot can complete the trajectory tracking task in a fixed time when the initial state is unknown.The simulation verifies the effectiveness of the controller.Finally,in view of the flexible joint robot working in the presence of parameter uncertainties and actuators saturation constraint problem,a kind of anti-windup adaptive neural network sliding mode controller(AWANNSMC)is designed.Firstly,the feedback linearization method is adopted to linearize the flexible joint robot model through appropriate state transformation.The adaptive neural network sliding mode controller is designed for the linear-treated model,and the RBF neural network is used to approximate the unknown model online and compensate in the controller.According to saturation constraint problem,adding saturated in the controller,makes the flexible joint robot in model parameters are completely unknown and under the condition of limited input will still be able to complete well trajectory tracking control.Simulation results demonstrate the effectiveness of the proposed controller.