Research On Genetic Algorithm Optimization-based Neural Network PID Control Strategy Of Electro-hydraultc Position Servo System

PID controller is widely used in various industrial control circumstance for its simplestructure、easy to implement、good control effect and strong robustness features．However,parameters of the traditional PID controller once set will not be changed, especially whenit comes to the system with characteristics of nonlinear complex、parameters timechanging、uncertainties,etc. So it is often difficult to obtain satisfied control performance.Luckily, with the development of intelligent control theory, the combination ofconventional PID controller and intelligent control technology, forming variety ofintelligent PID controllers, strengthening their own merits and complementing each other,provides a new way to solve the complex、dynamic and uncertain systems. It’s true of itscharacteristics: that not relying on accurate mathematical system model and betteradaptability to changes, making intelligent PID controllers become effective，practicableand having very broad prospects.As the most popular training algorithm for forward neural network learning, BPlearning algorithm modify BP neural network’ additional coefficient according to steepestdescent method based on non-linear programming methods, to make learning error reachthe minimum with the least-squares sense. But falling into local minimum、the sensitivityof the initial value of additional coefficient and slow convergence are its seriousdrawbacks, and that is the accuracy and good real-time characteristic of algorithm itselfremaining to solve. The key of BP neural network PID parameter tuning is to fuse BPneural network learning on the basis of conventional PID, so it is inevitable that there aresome common shortcomings of BP neural networks. Thus, when the network structure,learning rate factor, the activation function are defined reasonably, the research of therandomness selection of initial value and non-precision of the controlled plant’s Jacobianinformation (the sensitivity between system output and control input) has a veryimportant significance.Therefore, a new intelligent PID control strategy based on Genetic Algorithm(GA)off-line global convergence initial value of BP neural network、and BP neural networkself-learning on-line is proposed in this paper, and it solves the systemic defects immersedin a local minimum and accelerate the network’s convergence during the training of BPneural network effectively. This method can automatically adjust PID parameters toobtain satisfied performance, and was applied in the electro-hydraulic position servo system. Simulation analysis based on MATLAB shows that the intelligent PID controllerpossessing better dynamic response characteristics and robust performance than traditionalones.