Development of Robust Control Schemes with New Estimation Algorithms for Shunt Active Power Filter

Abstract

The widespread use of power electronics in industrial, commercial and even residential electrical equipments causes deterioration of the quality of the electric power supply with distortion of the supply voltage. This has led to the development of more stringent requirements regarding harmonic current generation, as are found in standards such as IEEE-519. Power Quality is generally meant to measure of an ideal power supply system. Shunt active power filter (SAPF) is a viable solution for Power Quality enhancement, in order to comply with the standard recommendations. The dynamic performance of SAPF is mainly dependent on how quickly and how accurately the harmonic components are extracted from the load current. Therefore, a fast and accurate estimation algorithm for the detection of reference current signal along with an effective current control technique is needed in order for a SAPF to perform the harmonic elimination successfully. Several control strategies of SAPF have been proposed and implemented. But, still there is a lot of scope on designing new estimation algorithms to achieve fast and accurate generation of reference current signal in SAPF. Further, there is a need of development of efficient robust control algorithms that can be robust in face parametric uncertainties in the power system yielding improvement in power quality more effectively in terms of tracking error reduction and efficient current harmonics mitigation.
The work described in the thesis involves development of a number of new current control techniques along with new reference current generation schemes in SAPF. Two current control techniques namely a hysteresis current control (HCC) and sliding mode control (SMC) implemented with a new reference current generation scheme are proposed. This reference generation approach involves a Proportional Integral (PI) controller loop and exploits the estimation of the in phase fundamental components of distorted point of common coupling (PCC) voltages by using Kalman Filter (KF) algorithm. The KF-HCC based SAPF is found to be very simple in realization and performs well even under grid perturbations. But the slow convergence rate of KF leads towards an ineffective reference generation and hence harmonics cancellation is not perfect. Therefore, a SMC based SAPF is implemented with a faster reference scheme based on the proposed Robust Extended Complex Kalman Filter (RECKF) algorithm and the efficacy of this RECKF-SMC is compared with other variants of Kalman Filter such as KF, Extended Kalman Filter (EKF) and Extended Complex Kalman Filter (ECKF) employing simulations as well as real-time simulations using an Opal-RT Real-Time digital Simulator. The RECKF-SMC based SAPF is found to be more effective as compared to the KF-HCC, KF-SMC, EKF-SMC and ECKF-SMC.
Subsequently, predictive control techniques namely Dead Beat Control (DBC) and Model Predictive Control (MPC) are proposed in SAPF along with an improved reference current generation scheme based on the proposed RECKF. This reference scheme is devoid of PI controller loop and can self-regulate the dc-link voltage. Both RECKF-DBC and RECKF-MPC approaches use a model of the SAPF system to predict its future behavior and select the most appropriate control action based on an optimality criterion. However, RECKF-DBC is more sensitive to load uncertainties. Also, a better compensation performance of RECKF-MPC is observed from the simulation as well as real-time simulation results. Moreover, to study the efficacy of this RECKF-MPC over PI-MPC, a comparative assessment has been performed using both steady state as well as transient state conditions. From the simulation and real-time simulation results, it is observed that the proposed RECKF-MPC outperforms PI-MPC. The thesis also proposed an optimal Linear Quadratic Regulator (LQR) with an advanced reference current generation strategy based on RECKF. This RECKF-LQR based SAPF has better tracking and disturbance rejection capability and hence RECKF-LQR is found to be more efficient as compared to RECKF-SMC, RECKF-DBC and RECKF-MPC approaches.
Subsequently, two robust control approaches namely Linear Quadratic Gaussian (LQG) servo control and H∞ control are proposed in SAPF with highly improved reference generation schemes based on RECKF. These control strategies are designed with the purpose of achieving stability, high disturbance rejection and high level of harmonics cancellation. From simulation results, they are not only found to be robust against different load parameters, but also satisfactory THD results have been achieved in SAPF. A prototype experimental set up has been developed in the Laboratory with a dSPACE-1104 computing platform to verify their robustness. From both the simulation and experimentation, it is observed that the proposed RECKF-H∞ control approach to design a SAPF is found to be more robust as compared to the RECKF-LQG servo control approach in face parametric uncertainties due to load perturbations yielding improvement in power quality in terms of tracking error reduction and efficient current harmonics mitigation. Further, there is no involvement of any voltage sensor in this realization of RECKF-H∞ based SAPF resulting a more reliable and inexpensive SAPF system. Therefore, superiority of proposed RECKF-H∞ is proved amongst all the proposed control strategies of SAPF.