224 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 29, NO. 1, FEBRUARY 2014 Adaptive Control Using Constrained RLS and Dynamic Pole-Shift Technique for TCSCs Dipendra Rai, Member, IEEE, Ramakrishna Gokaraju, Member, IEEE, and Sherif O. Faried, Senior Member, IEEE Abstract—In this paper, an adaptive pole-shift control technique for a FACTS device, namely Thyristor Controlled Series Capac- itor (TCSC), is presented. Adaptive pole-shift techniques have been successfully implemented for power system stabilizer applications in the past, but one of the difficulties in extending such a technique for transmission line control devices has been its inability to handle large disturbance conditions such as three-phase faults. In recent literature, random walk technique has been suggested during the system identification process, to overcome this problem. This paper presents a simple parameter constrained RLS identification proce- dure to track the large disturbance conditions. The effectiveness of the proposed methodology is demonstrated using (i) a three-area six-machine power system with a TCSC, and (ii) a IEEE 12 bus power system configuration with a TCSC. Index Terms—Adaptive control, flexible ac transmission systems (FACTS). I. INTRODUCTION I N COMPLEX interconnected systems, lightly damped in- terarea modes of oscillations may get excited during dis- turbances leading to an unstable system operation [1]. Flexible ac transmission system (FACTS) placed in transmission lines have been used as a mean to damp such oscillations [2], [3]. The phase lead-lag type of controllers is commonly used for the FACTS devices to improve the damping performance. How- ever proper design of the phase lead-lag controllers for FACTS could be a cumbersome task and the difficulties in tuning will be briefly discussed. One of the presently practiced procedures is to use linearized representation of the FACTS device, and designing the lead-lag controller based on the frequency-response characteristics of the linearized system [4]. Such linearization procedures some- times hide the unforseen interactions between the different com- ponents of the system. The performance of such fixed param- eter-based lead-lag controllers is generally good for one or two operating conditions but it has been found that the lead-lag con- trollers have to be retuned again using analytical tools when the system configuration undergoes significant changes (i.e., when new transmission lines, new types of generation, and power system components are added to the power system). Manuscript received December 13, 2012; revised May 20, 2013; accepted July 18, 2013. Date of publication September 16, 2013; date of current version January 21, 2014. Paper no. TPWRD-01358-2012. D. Rai is with BC Hydro, Vancouver, BC V4N 4X8 Canada. R. Gokaraju and S. O. Faried are with the University of Saskatchewan, Saska- toon, SK S7N 5A9 Canada (e-mail: (e-mail: rama.krishna@usask.ca; sherif. faried@usask.ca). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TPWRD.2013.2279539 Another commonly used procedure for tuning the lead-lag control parameters is using transient simulation programs, which requires a large number of repeated runs of the program. The Monte-Carlo search is one such approach, in which the control parameters are varied in a random manner [5]. This pro- cedure requires a enormously large number of electromagnetic simulation runs. One recent innovation is using a nonlinear optimization pro- cedure (such as simplex optimization) to reduce the number of electromagnetic simulations runs to find the optimum set of pa- rameters by directing the search to the most promising regions in the search space [6]. However, the controllers optimized offline using such a procedure cannot account for unforeseen signifi- cant changes in the system. Adaptive control techniques do not need such tuning and are able to achieve optimal operation for large disturbance con- ditions, a wide range of operating scenarios, and significant changes in the system [7]–[9]. The adaptive algorithm works on an estimated plant model at every sampling instant. The adap- tive controllers also track the system changes and are able to self-optimize. The model estimator tracks the changes in the power system rapidly and smoothly for uniform control action [10]. In adaptive control, methods based on least-squares fil- ters, such as recursive least squares (RLS) and Kalman filters, are most commonly used for system identification because of their simplicity and numerical stability[11]. However, during large disturbances, parameter identification using least-squares procedures is a real challenge. The parameters identified during such conditions have large and rapid fluctuations [12], and this results in undesirable control output (bang-bang-type control). The adaptive pole-shift control for FACTS devices have been studied in recent literature [13]–[15]. It was reported in [13] that the start of the estimation process gave poor system response to initial transients while using the recursive least square (RLS) estimator. Similarly, [15] reported that the variable forgetting factor-based RLS results in large variations in estimated param- eters during transients leading to wide variations in the control output and poor controller performance. For better parameter tracking, Sadikovic et al. [14], [16] proposed the use of a regularized constant tracing algorithm [10] to keep the correlation matrix symmetrical during the Kalman filter identification procedure. The main aim behind it is to ensure that the covariance matrix stays bounded. Plant dynamics is approximated by a 12th- order autoregressive external input (ARX) model. The authors use a 12th-order au- toregressive external input (ARX) plant model and sensitivity approach to modify the pole-shift factor every sampling instant. They use the derivative of control output [17] with respect to 0885-8977 © 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.