A bp neural network pid control algorithm based on variant particle swarm optimization

Abstract

The invention discloses a BP neural network PID control algorithm optimized by variation particle swarm optimization, which comprises the following steps: step 1, initializing the BP neural network; step 2, initializing the variation particle swarm optimization algorithm; step 3, determining the input signal of the BP neural network Step 4: Calculating the output adjustable parameters of the PID control system; Step 5: Updating the individual optimal value of the particle and the global optimal value; Step 6: Performing a mutation operation; Step 7: Judging whether to terminate the iteration; Step 8: Determining the output of the PID control system Tunable parameters. The present invention does not need to establish an accurate mathematical model, can automatically identify the parameters of the controlled process, automatically adjust the control parameters, and adapt to the parameter changes of the controlled process, and also overcomes the slow execution speed of the traditional BP neural network backpropagation error correction method, which is extremely slow. It is easy to fall into local minimum defects, has stronger robustness and better control effect.

Description

technical field [0001] The invention belongs to the technical field of intelligent control, in particular to a BP neural network PID control algorithm optimized by variant particle swarm optimization. Background technique [0002] PID control is proportional-integral-derivative control. It is a control strategy based on classical control theory. It is the most widely used in industrial process control, with the longest history and the strongest vitality. Among them, more than 90% of the control systems are PID control systems. It adopts a method based on mathematical models. Because of its simple algorithm, good robustness, high reliability, and good control effect, it is widely used in industrial control processes. For traditional PID control systems, before it is put into operation, it must first Set three parameters: proportional coefficient k p , integral coefficient k i , differential coefficient k d , in order to get the best control effect, if the control system p...

AI Technical Summary

Problems solved by technology

However, in the process of its practical application, the BP neural network PID control system designed by BP neural network has low learning efficiency, slow convergence speed, weak global search ability and easy to fall into local optimum due to the backpropagation learning algorithm. The effect is not ideal, which limits the application of neural network in PID controller

[0004] Paper: Xu Shengnan et al. Neural Network Adaptive Control Algorithm Based on Particle Swarm Optimization [J]. China Mechanical Engineering, 2012, 23(22), 2732-2738. Disclosed a neural network adaptive control algorithm based on particle swarm optimization , overcome the defects that the neural network is easy to fall into local minimum and slow convergence

Although the neural network adaptive control algorithm of particle swarm optimization is simple and easy to operate, and overcomes the BP algorithm to a certain extent, it is easy to fall into the local optimal value, but in the later stage of algorithm optimization, there are also situations such as reduced optimization speed and premature maturity. , leading to its weak global search ability, therefore, its control effect is not good and leads to relatively large errors in individual regions

Images