Fuzzy immune PID neural network control method based on boiler steam pressure system Third pacific-asia conference on circuits,communications and system,

Third pacific-asia conference on circuits ,communications and system, p.p. 1-5, July 2011

Abstract Introduction Theory on fuzzy neural network (FNN) Immune PID control Fuzzy neural network immune PID control Conclusions References

Steam pressure is a key point to keep the steam pressure constant in various operation situations. Considering steam pressure with the time delay and uncertainties, the sliding mode predictive control was used to design the controller. The predictive control was used to deal with time delay, the sliding mode control was used to deal with the uncertainties. And the predictive control can reduce the chattering phenomenon of sliding mode. This simulation results show that the proposed algorithm can largely imporved the system response performance compared to the single generalized predictive control

Boiler system is a complex industrial process, it has high nonlinearity, large delay, strong coupling and load disturbance. Boiler steam pressure power plant control system is a key link in the system, which directly affect the turbine speed. In engineering, the current steam pressure control system is mainly dominated by traditional PID control. In theory, land-based power plant boiler, the use of intelligent control strategy has been the boiler combustion system has been extensively studied [1-4]. Control of the ship boiler control system study is to PID [5], intelligent control has lagged behind, only a few fuzzy or neural network PID parameter calibration literature. However, due to the complexity of fuzzy rules of precise formulation, making the control precision is often not very satisfactory

Neural network parallel computing, distributed storage, fault-tolerant capability, with adaptive learning function and a series of advantages. But generally speaking, the expression of neural network is not suitable for rule-based knowledge, and therefore the neural network training, because it is not already some experience to good use the knowledge, often the initial weights can only be taken as zero, or the random number , which increases the training time or network requirements into a non-local extremum, which is insufficient neural network.

Intelligent behavior of biological information systems for science and engineering fields to provide a reference for a variety of theoretical and technical methods. Biological immune principle combined with conventional PID control from the immune PID control, can be mutually reinforcing in order to further improve the control performance. PID control is a reference biological immune system, immune mechanism and design of a nonlinear control method.

The system analyzes the dynamic characteristics of the boiler steam pressure, based on the model through the mechanism of a ship boiler steam pressure control system of intelligent control design, because we are fully Use of the advantages of the fuzzy neural network, combined with immune algorithm gives an adaptive, self-learning algorithm for PID controller design, simulation shows that the control algorithm has good control quality.