Enhanced intelligent proportional-integral-like fuzzy knowledge-based controller using chaos-enhanced accelerated particle swarm optimization algorithm for transient calibration of air-fuel ratio control system

摘要

The self-adaptive and highly robust proportional-integral-like fuzzy knowledge-based controller has been developed to regulate air-fuel ratio for gasoline direct injection engines, in order to improve the transient response behaviour and reduce the effort to be spent on calibration of parameter settings. However, even though the proportional-integral-like fuzzy knowledge-based controller can automatically correct the initially calibrated proportional and integral parameters, a more appropriate selection of controller parameter settings will lead to better transient performance. Thus, this article proposes an enhanced intelligent proportional-integral-like fuzzy knowledge-based controller using chaos-enhanced accelerated particle swarm optimization algorithm to automatically define the most optimal parameter settings. An alternative time-domain objective function is applied for the transient calibration programme without the need for prior selection of the search-domain. The real-time transient performance of the enhanced controller is investigated on the air-fuel ratio control system of a gasoline direct injection engine. The experimental results show that the enhanced proportional-integral-like fuzzy knowledge-based controller based on chaos-enhanced accelerated particle swarm optimization is able to damp out the oscillations with less settling time (up to 75% reduction) and less integral of absolute error (up to 64.07% reduction) compared with the conventional self-adaptive proportional-integral-like fuzzy knowledge-based controller. Repeatability tests indicate that the chaos-enhanced accelerated particle swarm optimization algorithm-based proportional-integral-like fuzzy knowledge-based controller is also able to reduce the mean value of objective function by up to 10.61% reduction and the standard deviation of the objective function by up to 28.29% reduction, compared with the conventional accelerated particle swarm optimization algorithm-based proportional-integral-like fuzzy knowledge-based controller.