Identification and Adaptive Position Control of Uncertain Two-Mass Systems with Backlash Hard Nonlinearity

This paper presents a novel method for identification and direct and indirect adaptive position control of uncertain two-mass mechanical systems with unknown backlash. The identification process consists of two cascade blocks. The first block is a linear-in-parameter model, and the second block is an iterative algorithm for estimating physical parameters of the shaft. By merging the identification method with an indirect adaptive control strategy, simultaneous identification of the physical parameters and adaptive position control of the load will be applicable. Additionally, a direct adaptive controller based on model reference adaptive control theory is designed to control the position of the load with all unknown physical parameters. The simulation results on a sample two-mass system depict the effectiveness of the identification, direct and indirect adaptive control algorithms.