Robust load states estimation against mechanical parameters variation of a two-mass system using acceleration-aided dynamic Kalman filter

This paper presents a robust load states estimation method of a linear motor driven table system using an acceleration-aided dynamic Kalman filter. The system is composed of a table and a load connected via an elastic beam attached on the table, therefore the system is modeled as a two-mass system. In order to improve the positioning performance at load- side, acquisition of the load position is important. However it might be difficult to place position sensors on the load from the viewpoint of cost saving and available space. In this research, a small-size and low-cost MEMS accelerometer is installed on the load in order to measure load acceleration. Load acceleration and table displacement information are utilized to the acceleration- aided dynamic Kalman filter in order to estimate the load position. Moreover, to achieve robust estimation capability against mechanical parameter variations of the system, the extended Kalman filter is configured in order to estimate mechanical parameters together with the load states. The effectiveness of the proposed estimation method is verified by numerical simulations.