State Estimation and Control of Three-Phase Two-Level Converters via Sliding Mode

Abstract

From the simulation results in Chap. 2, it can be seen that the second-order sliding mode observer (SMO) obtains the best disturbance observation, and offers the best compensation for the sliding mode controller. However, it has similar shortcoming as the linear observer in Chap. 2, i.e., if it needs to accelerate the observation speed, the observer parameters need to be increased, which leads to severer current THD level in steady state. Therefore, following the method in Chap. 3, an improved SMO is proposed in this chapter, which has following characteristics: Comparing with the SMO in Chap. 2, the improved SMO has one more tuning parameter, which is used to improve the system’s transient response while not influence the steady state performance, while the original two parameters are used to maintain system’s steady state performance. In this way, the transient time of the system after it is disturbed can be shortened and the grid current THD level can be kept low. Comparing with the improved LDO in Chap. 3, this improved nonlinear SMO has faster converging speed and stronger robustness against system uncertainties. In this chapter, a sliding mode control strategy based on the improved SMO is proposed to regulate the dc-link voltage of three-phase two-level power converter. In the voltage regulation loop, the improved SMO compensates the adopted sliding mode control. The SMO has similar characteristics as the LDO in Chap. 3, in that it has parameters to tune the transient process and the performance of steady state. Comparing with linear observer, the SMO converges faster and is more robust. The SMO is used to estimates the dc-link load that is abruptly connected and causes voltage fluctuation. The estimated value is used to compensate the controller in the voltage regulation loop, thus the voltage regulation loop obtains stronger robustness against external disturbances, so that the disturbing impact from the load can be greatly reduced. This chapter first proves the stability of voltage regulation loop under SMO-SMC control, via Lyapunov method. Then the effectiveness and advantage of the proposed control strategy is verified via simulation. Afterwards, experiments are carried out on a \(5\,\mathrm{kW}\) grid-connected power converter. The experiment results show that, comparing with conventional PI control, the proposed SMO-SMC control strategy significantly improves control performance of the voltage regulation loop, and results in stronger robustness against the operating point variation caused by the changes in external load and dc-link capacitance.