Kabushiki Kaisha Toshiba

Tsunekazu Endo

Naoki Imasaki

Junichi Kiji

Disclosed is a fuzzy control system which calculates a membership function concerning a control value of a to-be-controlled object through a fuzzy inference operation in accordance with an input value from the to-be-controlled object and a control rule, and determines the control value from the membership function to control the to-be-controlled object. This fuzzy control system has a plurality of internal states respectively expressed by a plurality of membership functions in addition to the input value and the control value, determines the state of the to-be-controlled object using a rule for evaluating the input value, alters the internal state by fuzzy inference, stores a decision result as a new internal state, and determines the control value of the to-be-controlled object through fuzzy inference from the new internal state and the input value.

Fuzzy control system

The basic state feedback gain is derived and selected by the on-line identification based on the on-line input/output data of a controlled object, and at the same time, the sensitivity of the state feedback gain is directly and sequentially calculated based on the on-line input/output data of the controlled object. The state feedback gain is finely adjusted when the state feedback gain is not optimum. Further, the sensitivity of the state feedback gain of the controlled object is derived by the feedback gain sensitivity calculating unit according to the on-line input/output data of the controlled object. When the sensitivity is determined to be zero based on the criterion by the controller, the state feedback gain is determined to be optimum. When the sensitivity is determined to be near zero, the state feedback gain is finely adjusted by use of the sensitivity. When the sensitivity is determined to be greatly deviated from zero, an initialization instruction is supplied to the on-line identifier to initialize the covariance matrix in the on-line identification. Thus, a satisfactory response speed and precision can be obtained even when the present invention is applied to a controlled object dynamic characteristic of which is significantly changed.

Adaptive control system and method thereof

A fuzzy control method of a controlled object uses control rules, having a condition part, each condition part having both a first proposition characterized by a membership function whose variable is a time and a second proposition with respect to the state of the controlled object. According to the method, inputting information with respect to the state of the controlled object is input to a fuzzy control execution circuit. The degree of confidence of the first proposition characterized by the membership function is calculated according to the information for each of the control rules by using a particular equation. Then, the degree of confidence of the satisfaction of a condition part of each of the control rules is calculated according to the degree of confidence of each first proposition. A suitable control rule is obtained according to the degree of confidence of the satisfaction of each condition part so as to control the controlled object according to the control rule.

Growing community of inventors

Yokohama, Japan

Tsunekazu Endo