Font Size: a A A

The Research And Application Of Fuzzy Neural Network In Multi Water Tank

Posted on:2017-12-16Degree:MasterType:Thesis
Country:ChinaCandidate:S H GaoFull Text:PDF
GTID:2348330482486469Subject:Control engineering
Abstract/Summary:PDF Full Text Request
Multi tank liquid level system can be often seen in the process of industrial production. So it is an important part of the national production. PID algorithm is commonly used into the liquid level control system. However, with the expansion of the scale of system level and application scope of the promotion, the structure of the system more and more complex. And PID algorithm often cannot meet the requirements of the level control system. At the same time, with the theoretical knowledge of the intelligent algorithm constantly improve, the intelligent control algorithm in practical application is more and more widely. Fuzzy control, neural network and fuzzy neural network, which combines the advantages of both, are typical intelligent algorithm. They all have broad application space in the complex and uncertain systems.This paper firstly makes a theoretical analysis of the static and dynamic characteristics of single and double water tank. And this paper uses the experimental method to establish the model of liquid level system. Then, combining with the dynamic and static characteristic of the level system, the principle of adjusting PID parameter and the working principle of fuzzy controller, the fuzzy rules are designed to adjust PID parameters. Then choosing the fuzzy reasoning model, this paper design fuzzy PID controller. According to the principle of fuzzy neural network, fuzzy neural network PID controller is designed in two tank system. As the characteristics of slow convergence, which as the fuzzy neural network training too many parameters, combining with fuzzy rules of adjusting the PID parameters, this paper reduce the number of the parameters of the fuzzy neural network training, to accelerate the convergence speed. Paper simulates and analyses the controlling effect of two tank water level control system with PID controller, PID controller and fuzzy neural network PID controller. For the four tank system with multi variable strong coupling, this paper designs a fuzzy decoupling controller, basing on the idea of decoupling and applying the ability of universal approximation of fuzzy reasoning. This paper simulates the decoupling effect of fuzzy decoupling controller.Last, in the experimental device of CS4000, this paper design the level control system with fuzzy neural network PID, analysis of the experimental results of fuzzy neural PID control system.
Keywords/Search Tags:liquid-level system, fuzzy control, fuzzy neural network, fuzzy decoupling
PDF Full Text Request
Related items