706 IEEE/CAA JOURNAL OF AUTOMATICA SINICA, VOL. 5, NO. 3, MAY 2018 Neural-Network-Based Terminal Sliding Mode Control for Frequency Stabilization of Renewable Power Systems Dianwei Qian and Guoliang Fan Abstract—This paper addresses a terminal sliding mode con- trol (T-SMC) method for load frequency control (LFC) in renewable power systems with generation rate constraints (GRC). A two-area interconnected power system with wind turbines is taken into account for simulation studies. The terminal sliding mode controllers are assigned in each area to achieve the LFC goal. The increasing complexity of the nonlinear power system aggravates the effects of system uncertainties. Radial basis function neural networks (RBF NNs) are designed to approximate the entire uncertainties. The terminal sliding mode controllers and the RBF NNs work in parallel to solve the LFC problem for the renewable power system. Some simulation results illustrate the feasibility and validity of the presented scheme. Index Terms—Generation rate constraint (GRC), load fre- quency control (LFC), radial basis function neural networks (RBF NNs), renewable power system, terminal sliding mode control (T-SMC). I. I NTRODUCTION W ITH the rapid development of economy, electrical power demand has become continuously stronger year by year. A vast amount of fossil fuels utilized in power generation results in energy crisis and environmental deterio- ration around the world [1]. One possible solution is to adopt clean and renewable energy instead of fossil fuel for power generation. Wind power is now the fastest growing energy source around the world because of its zero emission [2]. The percentage of wind generation in power systems increases with years. Wind energy has become one of the central research themes in energy science. Manuscript received November 2, 2015; accepted February 28, 2016. This work was supported by National Natural Science Foundation of China (60904008, 61273336), the Fundamental Research Funds for the Central Universities (2018MS025), and the National Basic Research Program of China (973 Program) (B1320133020). Recommended by Associate Editor Chengdong Li. (Corresponding author: Dianwei Qian.) Citation: D. W. Qian and G. L. Fan, “Neural-network-based terminal sliding mode control for frequency stabilization of renewable power systems,” IEEE/CAA J. of Autom. Sinica, vol. 5, no. 3, pp. 706-717, May 2018. D. W. Qian is with the School of Control and Computer Engineering, North China Electric Power University, Beijing 102206, China (e-mail: dianwei.qian @ncepu.edu.cn). G. L. Fan is with the Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China (e-mail: guoliang.fan@ia.ac.cn). 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/JAS.2018.7511078 Concerning real applications, a constant supply of electricity in some remote areas cannot be guaranteed by power grids. In these areas, wind energy may be inexhaustible and convenient. Therefore, wind power has been paid more and more attention and some control problems rise up in power systems with wind turbines [3]. An advocacy of wind power is due to its sustainable and renewable status. However, wind power af- fected by climate changes is intermittent. Its intermittence also has impressive effects on operation and control of renewable power systems. Consider a power system with wind turbines. The load in the power system is random and the power output of wind power is fluctuating. The power-output fluctuation and the load change would pose a reliability supply challenge. The challenge is displayed by power imbalance and frequency deviation in the power system [4]. Consequently, frequency control strategies must be adopted to overcome the challenge. Load frequency control (LFC) is one of the most profitable auxiliary services to guarantee the stable operation of power systems with the objective of preserving the balance between power generation and power consumption [5]. Recently, the LFC problem of renewable power systems has been paid more and more attention [6], [7]. In order to attack the generation intermittency of renewable power systems, some LFC methods have been investigated, such as fuzzy control [8], [9], predictive control [10], [11], and adaptive control [12], [13]. Invented by Utkin, the sliding mode control (SMC) is recognized as a powerful design tool [14]. On the sliding-mode stage, an SMC system is completely insensitive to parametric uncertainties and external disturbances under certain matching conditions, which exhibits better performance than the con- ventional robust control methods [15]. This property inspires some researches on SMC for the LFC problem [16]− [22]. However, the SMC-based LFC methods in pervious works [16]−[22] do not consider the complexity and challenge of renewable-energy sources. The method of terminal sliding mode control (T-SMC) [23], [24] guarantee the convergence of the SMC system within finite time. The T-SMC method can be considered for the LFC problem of renewable power systems. Power systems are inherently nonlinear [25]. The two main nonlinear factors are the governor dead band (GDB) and the