Long Paper © 2013 ACEEE DOI: 01.IJRTET.9.1. Int. J. on Recent Trends in Engineering and Technology, Vol. 9, No. 1, July 2013 513 Optimal Control of a Teleoperation System via LMI- based Robust PID Controllers A. Roushandel 1 , A. Alfi 2 , and A. Khosravi 3 1 Babol University of Technology/ Electrical and Computer Engineering Department, Babol, Iran Email: a.roushandel@stu.nit.ac.ir 2 Shahrood University of Technology / Electrical and Robotic Engineering Department, Shahrood, Iran Email: a_alfi@shahroodut.ac.ir 3 Babol University of Technology / Electrical and Computer Engineering Department, Babol, Iran Email: akhosravi@nit.ac.ir Abstract—Since the performance of the teleoperation systems can be considerably degraded by time-delay of communication channels and uncertainty in various parts of such systems, the main objectives of the controller design in loads of different structures of the bilateral teleoperation system are to preserve stability and tracking performance of these systems in spite of aforementioned sources of uncertainty. In this paper, a new robust PID controller will be designed based on H ” control theory by using the Linear Matrix Inequality (LMI) approach. Therefore, the problem of a Robust PID controller design can be regarded as a special case of the output-feedback controller via employing some sorts of changes in control and system parameters. To show the effectiveness of the proposed controller, the robust PID controller is compared with the multiobjective H 2 /H ” one. The main feature of the suggested structure is its ability to control the teleoperation system via using the simplest structure in which two signals will be transmitted to control the teleoperation system. In addition, use of PID controller has more practical applications in industrial units, due to its simplicity in implementation and capability to predict the time responses caused by changes in control parameters. Index Terms— Teleoperation systems, Robust LMI based PID controller, Multiobjective H 2 /H ” , Optimization. I. INTRODUCTION There has been a considerable growth in use of teleoperation systems in different areas such as: telesurgery, telemanipulation, space mission, nuclear power station, under sea research, Etc. A typical bilateral teleoperation system depicted in fig. 1 consists of five important parts: a human operator, a master manipulator, communication channels, a slave manipulator and remote environment. The master manipulator is directly drawn by the human operator and its position/velocity is transmitted to the slave site via communication channel. The slave manipulator has to track the position/velocity of master one and communicate reactions of the remote (task) environment to the master manipulator as the reflected force. Although using the bilateral structure in teleoperation systems increases ability of the human operator to control the teleoperation system, if there was a long distance between slave and master sites, performance and stability of the overall system would be deeply affected. The performance of a teleoperation system is addressed by an index called transparency. Transparency is defined as a match between forces that is exerted by the human operator and one which is reflected from the task environment. In other words, transparency is a capability of a teleoperation system to represent unchanged dynamics of the remote environment to the human operator [1]. Because of the existing uncertainties in dynamics of the system and time-delay in communication channels a compromise is required to be ensured between transparency and stability [2]. Since an increase in stability margins of the system leads to a decrease in transparency, control of a bilateral teleoperation system requires a delicate trade-off between two foregoing requirements namely transparency and robust stability. Until now several control schemes have been applied to bilateral teleoperation systems so next some of these schemes are considered briefly. H ” control theory was used by Sano et al. in 2000 [3]. They used four sensors in order to measure positions and forces of master and slave systems. In 2002, a new control method consisting of smith predictor and wave variables was proposed by Ganjefar et al. [4] for optimizing the performance of the teleportation system against the large variable time-delay. An optimal H 2 procedure was proposed for teleportation systems in 2004 by Boukhnifer and Ferreira [5]. Sunny et al. [6] designed an adaptive position and force stabilizing controller for bilateral teleportation system in 2005. In the same year Ganjefar and Miri [7] used optimization control method in teleportation systems. One year after that, ShaSadeghi et al. [8] proposed a new control structure based on adaptive inverse method while at this year Sirouspoor and Shahdi [9] employed predictor controller for teleportation systems. Tavakoli et al. suggested a modeling and stability analysis for discrete teleoperation system in 2008 [10]. While a sliding mode bilateral control was proposed by Moreau et al. for Master-Slave pneumatic servo systems [11]. Novel adaptive-based methods have been dedicated to various structures of teleoperation systems [12-15] to reduce the destructive effects of time delay and uncertainty on performance of these systems. Additionally, different disturbance based methods have been proposed for bilateral teleoperation systems to preserve the stability and performance [16-18]. 50