New Learning Algorithm for High-Quality Velocity Measurement and Control When Using Low-Cost Optical Encoders

Abstract

A new compensation method that can greatly reduce the slit errors (i.e., transition location errors) due to nonidealities in optical incremental encoders caused by manufacturing limitations in the code wheel, optical components, and analog circuitry is presented. An M/T-type constant sample-time digital tachometer (CSDT), which involves pulse-count and high-frequency timer measurement, effectively time-stamps the encoder transitions. Using CSDT-based data, encoder compensation techniques that improve velocity measurement accuracy are presented. These do not require precise knowledge of the shaft velocity, thereby eliminating the need for high-accuracy reference equipment. During the initial learning stage (possibly performed in situ ), slit errors are calculated through pseudoinverse-based solutions of simple approximate linear equations or an iterative method that requires very little memory storage. Subsequent operation of the motion system utilizes adjusted slit positions for more accurate velocity calculation. The performance improvement in velocity measurement is experimentally demonstrated using motor drive systems, each of which includes a field-programmable gate array (FPGA) and a digital signal processor (DSP). Results from open-loop velocity measurement and closed-loop servo control applications are given, with the latter highlighting the resultant reduction in high-frequency motor torque. Slit error reductions in the range of 60%-86% are obtained (typically approximately 80%) with a similar improvement in velocity measurement error.