Parameter Tuning Method Of Active Disturbance Rejection Control In Pneumatic Position Servo System

Pneumatic system has been widely used in machinery manufacture,food processing,printing and packaging,medical and chemical industries due to its advantages such as non-pollution,high power-to-weight ratio,safety and reliability,explosion-proof and anti-electromagnetic interference,simple structure,and low cost.Pneumatic system usually consists of components such as proportional valve and cylinder,and uses air as its working medium.Due to the compressibility of air.the non-linearity of the flow through the proportional valve,and the complex frictional force in the cylinder,the high-precision control of the pneumatic system is a challenging task,which limits the application of the pneumatic system in the field of high-precision position servo systems.The precise mathematical model of the pneumatic position servo system is difficult to obtain because of the influence of the operating state and the uncertain interference.Therefore,it is difficult to achieve desired performance for control method based on the precise mathematical model of the controlled object.The active disturbance rejection controller(ADRC)does not rely on the precise mathematical model of the controlled object,which make it gain extensive attention.For the control problem of pneumatic position servo system with unknown precise mathematical model,the linear and nonlinear active disturbance rejection controllers are used for the position control.However,active disturbance rejection controller has many parameters,how to tune these parameters is a challenging work without matured theoretical guidance.To deal with such problem,swarm intelligent algorithms,including genetic algorithm,particle swarm optimization and differential evolution,are proposed to optimize the controller's parameters online.In order to get the balance of the multi-objectives in optimizing process of active disturbance rejection controller parameters,a fitness function based on Pareto rank is proposed,Pareto rank based multi-objective genetic algorithm,Pareto rank based particle swarm optimization algorithm and Pareto rank based differential evolution algorithm are used to optimize the parameters of the controller,respectively.Since the optimization result of the multi-objective optimization algorithm based on Pareto rank is a set of equal-rank non-inferior solutions,the Euclidean distance between one individual objective functions and the minimum reachable value of each objective function is introduced as the index to obtain the global optimal solution from the set of equal-rank non-inferior solutions.The control experiments of the pneumatic system are carried out using the linear and nonlinear active disturbance rejection controller with the optimized parameters.The experimental results show that,compared with the active disturbance rejection control ler with parameters tuned using traditional trial and error tuning method,the tracking errors of the controllers using parameters optimized by the swarm optimization algorithm are all reduced,and the controf performance is improved.With the same population size and the number of evolution iterations,the active disturbance rejection controller optimized by the genetic algorithm has the largest tracking error for the three references,the controller optimized by the particle swarm optimization algorithm has less error,and the controller optimized by the differential evolution algorithm has the smallest tracking error for the given reference signals.Therefore,the differential evolution algorithm has the best comprehensive optimization performance.The optimized ADRC are compared with five control methods in the existing references.Experimental results show that the optimized active disturbance rejection controller has a small tracking error,high control accuracy,and less control energy consumption and obtain a good compenhensive performance.