Empirical and sensor knowledge-extraction for fuzzy logic motor control design

This paper presents a methodology for human and sensor data knowledge-extraction to assist in the design of a Fuzzy Logic Controller (FLC) when no parameterized model of the motor is available, thus it relays mainly on linguistic motor throughput description as its main data source. Proposed design methodology achieves acceptable control objective with two design stages; first, human empirical knowledge is used to specify FLC architecture and its initial parameters, employing experts' linguistic descriptions to construct controller rule base and knowledge base in accordance with cognitive map theory; Mamdani Fuzzy Inference Engine model (FIE) enables the designer to directly use empirical knowledge to create appropriate FLC by using linguistic terms to specify FLC structures. On second design stage, sensor data is use to fine-tune FLC parameters, as FLC parameters to motor control throughput relations is known by observation. The main objective of this paper is to develop a strategy of a FLC implementation capable of self-tuning, based on cognitive map theory and linguistic descriptions.