| Study purpose: Dynamic Matrix Control (DMC) which is often used in control is a way of prediction control. Fuzzy prediction control express manual control strategy by fuzzy logic, so that it can intelligently deal with prediction error.Study way: Convert a serial of fuzzy rules into a querying control table which is stored in computer to be used by on-line control. Study result: The error matrix of the system improved is not rulelessly seted, but instead of its function by fuzzy controller which realize fuzzy control. Study conclusion: exert manual control experience and strategy into system. Prediction control's essence is that using system prediction information to optimize some capability in limted time zone.Using roll optimizing strategy to conquer system uncertainity and roll optimizing is the kernel. So it is a control way which base on model and optimizing but different from traditional control ways.We need to compare traditional optimizing control , prediction control with traditional feedback control and point out that prediction control take the cost of using part optimizing instead of global optimizing ,take account of producing reality by feedback and optimizing.So prediction control is between optimizing control and modelless PID control .My work in my paper can be divided into two parts: arithmetic improvement and arithmetic implement . arithmetic improvemen include looking up lots of datum and choosing correct control way to make up for prediction control arithmetic's defaults;design fuzzy subset and control rule tables;design entire fuzzy prediction controller by locating table way;take place of error correct matrix with fuzzy prediction controller and improve entire system. arithmetic implement part include choosing correct developing way;design entire program implement steps; software coding , implement , testing and debugging ;change parameters to observe control result; change control rule to observe control result and analyse control result. |
PDF Full Text Request