06-02-2010, 08:39 AM
[attachment=1760]
CAN-based Synchronized Motion Control for Induction Motors
Abstract:
A control area network (CAN) based multi-motor synchronized motion control system with an advanced synchronized control strategy is proposed. The strategy is to incorporate the adjacent cross-coupling control strategy into the sliding mode control architecture. As illustrated by the four-induction-motor-based experimental results, the multi-motor synchronized motion control system, via the CAN bus, has been successfully implemented. With the employment of the advanced synchronized motion control strategy, the synchronization performance can be significantly improved.
Submitted by:
Jun Ren* Chun-Wen Li De-Zong Zhao
Department of Automation, Tsinghua University, Beijing 100084, PRC
Introduction
Because of the quick evolution of manufacturing processes, the demand for flexible automation systems is on the rise. To meet these requirements, distributed motion control architecture based on intelligent drives and fieldbus communication tends more and more to replace the traditional solutions. Many types of manufacturing equipment, such as printing machines, computer numerical control (CNC) machine tools, robots, and assembly lines, require operations of high speed and high precision as well as accurately coordinated motions among multiple motors. Therefore, many researches concerning synchronized motion control have been proposed in recent years. Lorenz and Schmidt[1] presented three approaches for process automation, namely, synchronization master-slave approach, master-slave approach, and relative dynamic stiffness approach. Cross coupling technique was initially proposed by Koren[2] for manufacturing systems and extended by Tomizuka et al.[3] Yeh and Hsu[4] proposed a new integrated control structure for multi-axis motion systems. Sun et al.[5, 6] combined adjacent cross-coupling control with adaptive control, and applied the control scheme to robot synchronization. Other approaches, such as relative coupling control[7], predictive control[8] , optimal synchronization control[9] and sliding-mode[10], were also applied to the synchronized motion system. However, the above mentioned synchronization control methods exhibit many problems[11,12], such as large numbers of wiring, complexity of electric circuits, noise and maintenance, which reduce reliability while increasing cost. Especially, many references mentioned above did not provide a possible way to extend their arrangements to more than two systems. Because of the rapid development of networked techniques[13,14] in practical applications, the integration of the control network and the multi-motor motion system becomes a promising prospect in modern Manuscript received June 23, 2008; revised August 11, 2008 This work was supported by National Natural Science Foundation of China (No. 69774011) *Corresponding author. E-mail address: renj03[at]mails.tsinghua.edu.cn industry[15-17]. In the networked control system, the wiring can be organized systematically by using a shared data network instead of hardwired connections, which can provide the control system with much more modularity, remotecontrol capability, and ease in diagnosis. In this paper, total sliding mode control is adopted in adjacent cross-coupling control structure to implement speed synchronization of multi induction motors. The speed synchronization strategy is to stabilize synchronization errors between each motor and its two adjacent motors to zero. A distributed control area network (CAN) bus synchronized motion control system is designfed to simplify the control structure of the system. Finally, simulation results of a fourinduction- motor experimental system demonstrate that the motion accuracy has been significantly improved and that flexibility and maintainability has become accessible to distribute the control structure.