英语翻译目前,由于PID结构简单,可通过调节比例积分和微分取得基本满意的控制性能,广泛应用在电厂的各种控制过程中.电厂主

Currently, the PID structure is simple, but by adjusting the proportion integral and differential obtain basic satisfactory control performance, widely used in power plant all kinds of control process。 Power plant main steam temperature controlled object is a big inertia, time-delayed, nonlinear and object changes system, conventional steam temperature control system for cascade PID control or guide differential control, when the unit before the stable operation, the general will main steam temperature control in allowing range。 But when running condition when great changes have taken place, but it was difficult to ensure that the control quality。 Therefore this paper studies based on BP neural network PID control, using neural network self-learning, nonlinear and not rely on the model of characteristics to realize PID parameters online auto-tuning, make full use of the advantages of PID and neural network。 This place USES a multilayer feedforward neural network, and adopts back propagation algorithm, based on real-time output control requirements for Ki, Kd Kp mohan, PID controller, ordinal as the real-time parameters, instead of the traditional PID parameters depend on experience neatly and engineering setting, in order to achieve the temperature system of time-delayed Lord good control。 For such a system in MATLAB simulation research, the simulation results show that based on the BP neural network auto-tuning PID control has fine self-adaptive capacity and the ability to learn, to change the system ChiYanHe object large can get a good control effect。