A Comparative Study of Design of Active Fault-Tolerant Control System for Air–Fuel Ratio Control of Internal Combustion Engine Using Particle Swarm Optimization, Genetic Algorithm, and Nonlinear Regression-Based Observer Model

Abstract

In this article, three distinct strategies for designing an Active Fault-Tolerant Control System (AFTCS) for Air-Fuel Ratio (AFR) control of an Internal Combustion (IC) engine in a process plant to avoid engine shutdown, are presented. The proposed AFTCS employs a Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and a Nonlinear Regression (NLR)-based observer model in the Fault Detection and Isolation (FDI) unit for analytical redundancy. A comparison between these three proposed techniques is carried out to determine the least expensive and most accurate approach. The results show that the nonlinear regression produces highly accurate results by consuming very low computational power, and its response time is also very low as compared to GA and PSO. The results obtained show that NLR requires 99.6% and 93.1% less computational time for throttle and MAP estimation, respectively, by reducing the estimation error to as low as 0.01. The simulation of the proposed system is carried out in the MATLAB/Simulink environment. The results prove the superior fault tolerance performance for sensor faults of the AFR control system, especially for the Manifold Absolute Pressure (MAP) sensor in terms of less oscillatory response as compared to that reported in existing literature.