Robust fuzzy control method for monotone inertial systems

In this paper, a robust controller based upon Mamdani architecture is presented, which is equipped with an adaptive measure to plant parameter variations and environmental effect. Besides, a compelling stable strategy is configured in it. Its control cycle can be altered automatically with the system output. To adapt the changes of system and external environment parameters, a rule-base self-adjusting method is employed. The adaptive process can be triggered when some state aberrance points are found or when the controller is turned on each time. To identify the trend of the system state, shape vector of a curve and state aberrance points are proposed. The adaptive process begin with a sequence of test input, then system output process is recorded by response function of the system. Next deviation degree between the current response function and the standard output response is calculated. By matching the shape vector and analyzing the deviation degree, the current circumstance is identified so that the rule-base can be automatically modified.