Addressing to online adaptive controller malfunction in fault tolerant control

Abstract

A complete Fault Tolerant Control solution calls for a nonlinear adaptive controller with universal approximation capability and guaranteed stability. To fulfill this requirement we propose the use of Neural Networks trained online under a Globalized Dual Heuristic Programming architecture supervised by B decision logic capable o f identifying controller malfunctions in early stages and providing new avenues with greater probability of convergence using information from a Dynamic Model Bank. The classification and distinction of controller malfunctions and of the faults in the system is achieved through three independent quality indexes. Proof-ofthe-concept simulations of nonlinear plants demonstrate the approach legitimacy. Index Terms -Fault tolerant control, confroller malfunction, fault detection and diagnosis, adaptive critic, multiple models 0-7803-8359-

Linear parameter‐varying‐based fault‐tolerant controller design for a class of over‐actuated non‐linear systems with applications to electric vehicles

“…Considering the physical limitations of the actuators and transient responses, the control inputs should be restricted by constraining all the eigenvalues of the closed-loop system into a disk. Hence we introduce the following lemma as, Lemma 2 Chilali and Gahinet [40]: Consider the system in (18), all the poles of the closed-loop system lie inside a disk centered at (−q, 0) with radius r if there exist positive symmetric matrix P 4 and general matrix H satisfying…”

“…Several passive FT controllers have been proposed so far. For example, a robust control-based FT controller for a double inverted pendulum system were reported in [14], adaptive control-based passive FT controllers are proposed in [7,[15][16][17][18], and sliding mode control-based FT controllers are given in [2,19,20]. Despite the various passive FT controllers constructed by the aforementioned methods, several limitations of the existing passive FT controllers for over-actuated systems deserve further investigations.…”