ANFIS parallel hybrid modeling method for optical encoder calibration

A novel hybrid modeling method, combining the knowledge-based model and adaptive-network-based fuzzy inference system (ANFIS) model, is proposed in this paper. The modeling process is presented step by step. Firstly, a simulation based on a single-input nonlinear function is carried out to verify its feasibility and effectiveness, in comparison with results in the previous publication. Secondly, a calibration experiment based on a 16-bit absolute type encoder is presented. Significant improvements regarding the measurement accuracy of the encoder are achieved by employing the proposed approach with respect to original data and traditional algorithms.