Nonlinear Error Compensation of Capacitive Angular Encoder

Angular encoder are widely used in industrial fields. In addition to the true angle, the output of the encoder is superimposed with nonlinear errors caused by installation, processing, demodulation circuits, etc. The methods such as least square method (LSM) and back propagation neural network (BP-ANN) cannot effectively identify and compensate the nonlinear errors. Aiming to solve the problem, the method of support vector machine (SVM) is proposed to achieve the nonlinear error compensation. Furthermore, the adopt particle swarm optimization (IPSO) is applied to determine the parameters of SVM for further improve the precision of compensation model. The encoder is applied to verify the validity of the algorithm and experimental results show that the IPSO-SVM method can compensate the nonlinear and reduce form 0.08° to 0.0004°, and it is an effective method for compensation the nonlinear error of capacitive angular encoder.