Research Of Fuzzy Controller Based On Association Rules With Temporal Constraints

The fuzzy control rules generate from domain expert knowledge, so control systemexists problem which is the bottleneck of knowledge acquisition and lacks self learningability in the static knowledge, and it can not only adapt system but also track the changeof the control system. The method of data mining is introduced to the intelligent controlsystem, the accumulated historical data and current data is using, the potential patternsand rules is extracting. The connection between the fuzzy control and of data mining isestablished to provides a way of obtaining fuzzy control rules.The date changes over time in the actual, compared with the traditional associationrules, and the data may be out of date, deleting and adding in the datebase, and thetemporal constraint association rules reflect knowledge which is hidden in the date and isrelated to time.With the changing of time the data generation will change in intelligent control field,and the data changes can make system change which affect the performance of thesystem. The paper integrates temporal constraint association rule and fuzzy control rules,and gives a method to solve the defects existing in traditional fuzzy control rules, and thefuzzy control rules can be updated, so that they can better adapt to the reality in theprocess of production, to achieve better control effect. In this paper, the paper putforward a kind algorithm which is related to incremental updating, for the system to addan optimized update mechanism. At last, the simulation results prove that the fuzzycontroller based on the temporal constraint association rules has obvious advantages inthe timetopeak, the maximum overshoot and number of oscillation compared to atraditional fuzzy controller, and proved that using the method has its rationality andfeasibility.