Reduced-order Generalized Proportional Integral Observer Based Continuous Dynamic Sliding Mode Control for Magnetic Levi
- PDF / 2,938,856 Bytes
- 10 Pages / 594.77 x 793.026 pts Page_size
- 63 Downloads / 201 Views
ISSN:1598-6446 eISSN:2005-4092 http://www.springer.com/12555
Reduced-order Generalized Proportional Integral Observer Based Continuous Dynamic Sliding Mode Control for Magnetic Levitation System with Time-varying Disturbances Junxiao Wang*, Lei Zhao, and Li Yu Abstract: In order to reduce the influence of time-varying disturbances for magnetic levitation system, we propose a reduced-order generalized proportional integral observer (RGPIO) based continuous dynamic sliding mode control scheme for magnetic levitation system in this paper. Unlike the popular extended state observer (ESO), it could deal with constant or slowing varying disturbances from theoretical point of view, the reduced-order generalized proportional integral observer (RGPIO) is designed to estimate the time-varying disturbances and system states, then the dynamic sliding mode surface is developed and deduce a continuous sliding mode controller (CSMC) for magnetic levitation system. Compared with ESO based continuous sliding mode controller, the proposed method not only ensures the position tracking accuracy, but also obtain better time-varying disturbance reject ability. Simulation and experimental results are also given to verify the effectiveness. Keywords: Continuous sliding mode control, magnetic levitation system, reduced-order generalized proportional integral observer, time-varying disturbances.
1.
INTRODUCTION
Magnetic levitation system has the advantages of no pollution, no contact, no lubrication, long service life, so it has been widely used in many fields such as industry, aviation, medical, etc [1, 2]. The representative applications are maglev train [3], magnetic bearing [4, 5], maglev wind turbine [6], artificial heart [7], etc. However, magnetic levitation systems are always complex nonlinear systems with multiple degrees of freedom, its mathematical model is very complicated, required high real-time performance controller, and often occurs serious chattering during the control process. Therefore, achieving the high control performance of magnetic levitation system is a challenging task. To improve the system performance, many different control methods has been developed, such as the traditional PID control method [8], but its control accuracy is limited. Fuzzy control [9], robust control [10, 11], and neural network control [12,13] are advanced control methods, these methods improve the control performance of the
magnetic levitation system from different aspects. In addition, sliding mode control is a nonlinear control method with simple structure and strong robustness [14–18], it has been widely used in the field of control application during recent years [19–21]. However, the discontinuous switching of the sliding mode control will aggravate the system chattering, so the saturation function based continuous sliding mode control is often used to suppress this phenomenon [22]. Due to time-varying self-inductance and mutual coefficients, nonlinear and time-varying electromagnetic force, the modeling error is avoidable based on
Data Loading...
