Research On Optimizing Fuzzy Logic Controllers With Genetic Algorithm And Its Application To Ball Mill

Ball coal mill system is a three input and three output coupling system, which have characteristic of time-varying and great changes in operating conditions. Thus it is difficult to establish precise mathematical model and determine the model parameters and then application of the traditional control theory and methods to achieve the desired control. By analyzing the working process of ball coal mill system which takes outlet temperature, vacuum and load as the controlled variable. One of the main focuses is to find effective measure method to deal with the load measure problems and load control problems to solve blocking problem and guarantee long-term automatic operation and then by analyzing and testing to find out lower power consumption operation conditions. The paper puts forward a method to measure the loading of ball mill which uses both the coal density signal with pressure difference signal to represent the loading of ball mill. To improve its control performance, feed forward decoupling reduce the coupling then three single loops was set up. Due to its complicated characteristics fuzzy controller has many advantages when compared with several other advanced control techniques. The two-input single-output structure of the fuzzy controller was designed on the basis of analysis and judgment according to error and error change rate. It introduces that how to design a fuzzy controller step by step, such as build the fuzzy rules, choose the scale factor and so on in detail. The fuzzy PID controller preserves the structure of the conventional one, but the proportional, integral, and derivative gains rate are modified according to their input signals of the error and error change rate, which have certain adaptive capability with PID controller. Then the proposed design of fuzzy controller is optimized by using the GA to modify the parameter and the membership function. It is tested with a typical ball mill model, which demonstrated that these optimized gains make the fuzzy PID controller robust with faster response time and less overshoot than its conventional controller. During the design process, Matlab modeling and programming is used from the beginning to the end, typical model for the simulation objects was simulated, and some laws and skills to find the parameters of how to adjust to complete the design was discovered. Simulation experiment shows better robustness performance of fuzzy controller that has been optimized by GA.Now that the design and simulation study of high-performance control system of a fuzzy controller based on genetic algorithms is applied to ball mill system in power plant, the characteristic curve and the best coal loading region which has lowest power consumption and high safety standard can be get by analyzing and testing. In this way the mill can reduce power consumption and guarantee the safety and economic operation.