Research and Application Based on Rough Sets and MEA
|School||Taiyuan University of Technology|
|Course||Control Theory and Control Engineering|
|Keywords||rough sets information quantity mind evolutionary algorithm fuzzy modeling|
This paper deeply analyzes basic ideas and pith of Rough Set theory, introduces its combination with other soft computing methods such as evolution algorithms, fuzzy sets, neural networks, and then reviews its current application in Intelligent Control systems.Rough Set theory method analyzes limited-dimension discrete datasheet, so we must discretize data in database. But for an original database with large continuous attributes values, how to discretize data reasonably is a noticeable problem.Aiming at this condition, by using the characteristics of rough set which can explore knowledge from data and mind evolutionary algorithm which can find the best point globally, this paper combines the two methods for the first time, and used mind evolutionary algorithm to search reasonable split points. Considering the whole condition attributes values, this paper proposes a global discretization approach to avoid creating unreasonable discrete split points that appeared in discretizing eachcontinuous attribute independently. By using the method in this paper, the input and output data of cement stove were analyzed, and the fuzzy model of cement stove is established. The experiment shows that the method is simple and rapid and can elide complex handwork reduction.Applying this method into the complex controlled object with large time delay and characteristic of time varying, a fuzzy controller based on rough set and mind evolutionary algorithm is presented in this paper. By studying input and output characters of the controlled object, which is under the control of "experts", rough set and the mind evolutionary algorithm together provide an efficient method to convert expert knowledge from data to fuzzy rules. This method will largely enhance the MIQ of the intelligent system.