The Research Of Multi Tank Water Control Strategy Based On PID Neural Network Algorithm

In the modern control field, the controlled variable is increasingly complex, we often meet with unknown mathematical model is hard to establish and system often parametervariation in the control process. And such objects are mostly mul-tivariable, strong coupling, and some also have characteristics of a large delay. Multi tank is a kind of typical complex system with nonlinear, time de-lay and strong coupling, it can simulate the industrial well. Due to the multi tank characteristic, this article discusses and studies on the four tank water level con-trol algorithm in detail, in this paper, PID neural network control algorithm of water level is given. Therefore, the research on the Liquid level control of multi tank is very important.The PID neural network control method, which is used, can improve the ef-fect of the traditional PID control.The neuron is used to determine the process of traditional PID control parameters, the method based on PID control algorithm on a plurality of neural network to realize the increase of the controlled object, we determine the adjustment of PID parameters needed by using neural network and eventually its learning function. The use of the network can be given on the basis of the controlled system setting value and returns the value of the input and output variables of repetitive learning and repetitive control parameters, so as to increase our controlled system PID controller structure using parameters. Through this kind of control experiments on multi-tank water liquid level, the experimental results were sorted and analyzed, to draw conclusions from the data, reliability, stability, fast is better than the traditional PID control system, that this control algorithm is designed in this paper multi-tank water control effect, good gained control system control on this, provides a good solution for solving simi-lar complicated control system, incident, inertia.