International Journal of Control, Automation and Systems 15(X) (2017) 1-10 http://dx.doi.org/10.1007/s12555-015-0307-5 ISSN:1598-6446 eISSN:2005-4092 http://www.springer.com/12555 Active Fault-tolerant Control Against Actuator Fault and Performance Analysis of the Effect of Time Delay Due to Fault Diagnosis Qikun Shen*, Bin Jiang, and Peng Shi Abstract: This paper discusses the problem of fault-tolerant control against actuator fault, derives the time spent at each steps in fault diagnosis which is called as the time delay due to fault diagnosis and quantitatively analyzes its effect on the faulty system’s performance. A fault diagnosis algorithm is first proposed. The proposed fault tolerant controller is designed to guarantees that all signals in the closed-loop system are semi-globally uniformly ultimately bounded, where the controller singularity is avoided without projection algorithm. What’s more, the analytical expression of the time delay is derived strictly. Further, the quantitative analysis of system performance which is degraded by the time delay is developed, and the conditions that the magnitudes of the faults should be satisfied such that the faulty system controlled by the normal controller remains bounded even stable during the time delay are derived. In addition, the corresponding solution to the adverse effect of the time delay is proposed. Finally, an experimental test shows that the proposed control algorithm has a very reliable efficiency. Keywords: Fault-tolerant control, fault diagnosis, fault estimation, performance analysis. 1. INTRODUCTION In modern control mechanisms, various systems com- ponents such as actuators, sensors and processors may un- dergo abrupt failures during plant operation. To improve system reliability and to guarantee system stability in all situations, many effective fault-tolerant control (FTC) ap- proaches have been reported in literature [1–7]. Passive FTC approaches use feedback control laws that are robust with respect to possible system faults [8–13]. On the other hand, active FTC uses a fault detection and isolation (FDI) module and accommodation techniques [14–26]. In [18], asymptotic estimation error lim t →∞ e x (t )= e x (∞) was con- sidered as a fault indicator, by which the faulty system can be distinguished from the normal system. In [24], a general active FTC framework was proposed for nonlin- ear systems with actuator faults. However, this control framework works under the assumption that system states are bounded in all cases. Furthermore, the fault estimation algorithm was not given. However, e x (∞) is not available in practice, and e x (∞) can not practically be considered as a fault indicator. A motivation of this work is thus to Manuscript received August 10, 2015; revised December 5, 2015 and March 1, 2016; accepted March 16, 2016. Recommended by Associate Editor Guang-Hong Yang under the direction of Editor Duk-Sun Shim. This work was supported in part by the National Natural Science Foundation of China ( 61473250, 61490703, 61573112), the Australian Research Council (DP140102180, LP140100471) , the Natural Science Foundation of Jiangsu Higer Educatuion Institution (14KJB120013), and a Project Funded by the Priority Academic Program Development (PAPD) of Jiangsu Higher Education Institutions. Qikun Shen is with the College of Information Engineering, Yangzhou University, Yangzhou 225127, China (e-mail: qkshen@yzu.edu.cn). Bin Jiang is with the College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China; and also with Jiangsu Key Laboratory of Internet of Things and Control Technologies (Nanjing Univ. of Aeronautics and Astronautics) (e-mail: binjiang@nuaa.edu.cn). Peng Shi is with College of Automation, Harbin Engineering University, Harbin 150001, China; and also with the College of Engineering and Science, Victoria University, Melbourne VIC 8001, Australia (e-mail: peng.shi@adelaide.edu.au). * Corresponding author. provide a fault indicator with an associated decision algo- rithm which is efficient in practical application. Generally speaking, an active FTC is designed, based on an open-loop system modeled as a function of fault parameters under the assumption that they are immedi- ately identified by a FDI model [27]. In fact, there is always some level of time delay, which is called as the time delay due to fault diagnosis (FD) in this paper, to detect, isolate and estimate the faults occurred in the con- trolled systems. When a fault occurs, the faulty system works under the nominal control until the fault is detected, isolated and fault accommodation is performed, which may cause severe loss of performance and stability. To our best knowledge, there are few papers considering the time delay’s adverse effect on the stability of the system. [21–23, 28–31] tried to investigate the problem. However, the results in [21–23, 27–29] were obtained under some restrictive conditions. Furthermore, the analytical expres- sion of the time delay due to FD did not be given explicitly in [21–23, 27–29], which motivates this paper, again. In this paper, we investigate the problem of fault toler- ant control for a class of uncertain systems with actuator c ⃝ICROS, KIEE and Springer 2017