Constrained Stability Least Mean Square Algorithm with PI Gains Optimization in Control of UPQC

In this article, an adaptive filtering technique for removing the PQ disturbances from ac mains is proposed based on Constrained Stability Least Mean Square (CS-LMS) algorithm for 4-wire configured Unified Power Quality Controller (UPQC). The CS-LMS algorithm is derived from the minimization of the difference weight vector’s squared Euclidean norm under a stability constraint specified over the posterior error. This algorithm basically minimizes the mean squared error as to act adaptive filter. It CS-LMS algorithm provided active and reactive fundamental components from the distorted voltage and current signals. Apart from CS-LMS algorithm, a new method using Imperialist Competitive Optimizer (ICO) for designing PI controller has been used in UPQC. This overcomes the burden of the manual tuning process of PI-controller gains and the obtained results are compared with hit and trial method. The choice of ICO optimization approaches in PI control tuning over hit and trial method is because of faster tuning, save time and more accurate. The performance of four wire UPQC is evaluated for a variety of PQ disturbances, including low power factor, load balancing, voltage unbalancing, voltage sag mitigation, and neutral current mitigation. The performance results obtained through implementing using the MATLAB/simulink environment and a micro lab box processor, confirming that the UPQC control with CS-LMS algorithms performs satisfactorily.