The Fuzzy Control Rules Are Automatically Generated Software Research And Implementation

For the issues of time, location, experience and decision-making restrictions in industrial control automation, the software generated automatically under fuzzy control rules is designed. This software is to study and design the automatic generation of control rules in the fuzzy controller based on the theory of data mining, which contains two theories of data mining and fuzzy control.To analysis the massive industrial data, extracting unknown industrial state reaction mode and rules, data clustering and bulk incremental clustering the complex industrial process historical data including normal, abnormal status and fault status of various historical data, to automatically generate the fuzzy control rules and update them, and to achieve self-learning and self-improvement of the control system of the rule base, so that the system could continually acquire new knowledge from new instances and further expand the knowledge base while running the process.For the fuzzy control rules getting from data mining, we have to guarantee the optimization of rules and the good performance of the system. Therefore, using phase plane method to analysis the stability of the fuzzy control system, so as to ensure that the system is running at steady state; the relative Euclidean distance is used to calculate the intensity of the interference between rules, and to amend the rules and determine the completeness of the rules; using genetic algorithm to optimize fuzzy control rules, so that the system will get the optimal rules under steady operation.By the study of overall design and module design, this dissertation expounds the design of particular functional modules of data preprocessing, data mining, automatic generation of fuzzy control rules and performance determination and optimization, describes the key technology to realize functional modules, and ultimately accomplish parameter adaptive control, operation optimization control. This Expert System has laid a theoretical foundation for the intelligent control in the areas of inexperienced applications.