An Adaptive Takagi–Sugeno Fuzzy Model-Based Generalized Predictive Controller for Pumped-Storage Unit

Jianzhong Zhou,Xinjie Lai,Chaoshun Li,Yongchuan Zhang,Nan Zhang

doi:10.1109/access.2019.2931575

Abstract

In order to improve the control performance and suppress the “S” characteristics area instability of the pumped-storage unit (PSU), this paper proposes an adaptive Takagi-Sugeno fuzzy model-based generalized predictive controller (ATS-GPC) for the PSU. First, the T-S fuzzy model is used to obtain the controlled autoregressive integrated moving average (CARIMA) model, in which the fuzzy C-means (FCM) clustering algorithm is used for the identification of antecedent parameters and the least square method (LSM) is used to obtain the consequent parameters. Meanwhile, the T-S fuzzy model can be online adjusted according to the real-time tracking error feedback to decrease the influence of the initial offline trained fuzzy model. Then, the generalized predictive controller is designed for the PSU based on the CARIMA. Finally, some numerical simulation experiments including the start-up process, frequency disturbance process, frequency tracking experiments, and robustness analyses have been conducted to verify the proposed method. The experiments results have shown that the proposed ATS-GPC can significantly improve the control performance of the PSU and effectively suppress the unstable operation in “S” characteristics area. In addition, the strong robustness of the proposed controller is verified.

Highlights

With the rapid development of human society, such problem as the traditional energy shortage, environment pollution have become more and more severe, so realizing the transition from fossil energy system to clean and renewable energy system is an effective approach to climate and environment change [1]–[3]
In order to improve the control performance and decreasing the error of the initial offline trained fuzzy model, the adaptive T–S fuzzy model composed of the offline model training and the online consequent parameters adjustment is integrated with the generalized predictive control (GPC) to control a pumped-storage unit (PSU)
The T-S fuzzy model identification is applied to approximate the dynamic characteristics of the established accurate pump-turbine governing system (PTGS)

Summary

INTRODUCTION

With the rapid development of human society, such problem as the traditional energy shortage, environment pollution have become more and more severe, so realizing the transition from fossil energy system to clean and renewable energy system is an effective approach to climate and environment change [1]–[3]. Xu et al designed an adaptively fast fuzzy fractional-order PID control method for PSU using improved gravitational search algorithm (GSA) and obtained better control effect at low and medium water head [19]. Xu et al designed an adaptive condition predictivefuzzy PID controller, which combines the model predictive control (MPC) and fuzzy logic control theory to study the optimal control of PSU under no-load start-up condition at low head area [26]. Li et al designed of a fractional-order PID controller for a PSU using a gravitational search algorithm based on the Cauchy and Gaussian mutation, which effectively improves the control performance of the PSU at different water head [18]. The experiments results have shown that the proposed ATS-GPC can significantly improve the control performance under different water head and avoid the PSU trapping into the ‘‘S’’ characteristics area.

MODELING OF PUMP TURBINE GOVERNING SYSTEM

MODELING OF PUMP-TURBINE

MODEL VALIDATION

EXPERIMENTS AND DISCUSSION

CONCLUSION