Novel Metaheuristic Bio-Inspired Algorithms for Prognostics of Onboard Electromechanical Actuators

Metaheuristic bio-inspired algorithms are a growing class of optimization techniques, particularly suitable for the solution of complex, non-linear and multi-modal problems. In the last few years, new metaheuristic algorithms are continuously being developed, each more suitable for a given class of problems. In this paper, the authors analyze the prognostic Fault Detection and Isolation (FDI) process for an electro-mechanical actuator for flight control system, powered by a BLDC trapezoidal motor. Two algorithms are compared: Genetic Algorithm (GA) and Differential Evolution (DE). A high-fidelity reference model (RM), developed in the Matlab-Simulink environment, simulates the real behavior of the system. Taking advantage of a simpler and faster monitoring model (MM), the algorithms try to estimate the health status of the high fidelity model.