The Research And Application Of Fuzzy Control Based On Granular Computing

Fuzzy control is the application of fuzzy theory in the field of control, and it is a primary method of intelligent control. Generally speaking, fuzzy control imitates the operator's control procedure, including the control experience and knowledge, and describes control law with language. Because fuzzy control presents knowledge in the form of language, its mechanism accords with people's direct description and logic towards the process control. Fuzzy control has strong robustness, and can be used for non-linear, time-varying, time-delay system's control. Now, fuzzy control is developing towards self-adaptation, self-organization, and self-learning, allowing control parameters and rules can adjust, modify, and improve automatically, then the control performance could be improved constantly. To combine neural network, genetic algorithm, chaos theory, or other soft computing methods with fuzzy control has become fuzzy control's trend.Granular computing is a new concept and computing paradigm of information processing. It abandons the traditional model which always tries to find the exact solution. Granular computing uses good approximate solution to research the uncertain, fuzzy and a mass of real problems, changes the traditional computing concept, and makes the information processing more scientific, more reasonable, more economic and easier to operate.This paper researches the relationship between input and output of fuzzy controller with the help of granular computing. It uses the rough set theory to establish the attribute reduction algorithm based on granular matrix, reduces dimensions of fuzzy controller's input space, solves the fuzzy controller rule's exponential explosion problem, simplifies the fuzzy controller's design and lays a good foundation of subsequent rule extraction. I use quotient space theory to combine fuzzy control and PID control. Through adopting fuzzy control on coarse granularity and PID control on fine granularity, the whole control system's accuracy and speed could be improved. This paper also applies fuzzy control into the drainage system of city highway tunnel. Water level and water level's change rate are input parameters of the fuzzy control system, and they should be taken into account together. The number of water pump which should be turned on is output parameter of the system. I use Matlab to simulate the fuzzy control system and get the surfview of the fuzzy controller.