Improved scheme of membership function optimisation for fuzzy air‐fuel ratio control of GDI engines

Abstract

This paper researches an improved scheme of Membership Function Optimisation (MFO) for fuzzy Air-fuel Ratio (AFR) control of Gasoline Direct Injection (GDI) engines based on Correspondence Analysis (CA). This PI-like Fuzzy Knowledge-Based Controller (FKBC) optimised by the proposed scheme can further optimise AFR control performance while maximising conversion efficiency of the Three-Way Catalyst (TWC) to eliminate the exhaust emissions in real-time. Different from the conventional experience-based Membership Function Design (MFD) method for an FKBC, the proposed MFO scheme uses CA approach and can visualise the relationship between engine step gain scenarios and designed MF patterns to precisely determine its scalar parameters for AFR regulation of GDI engines. Within this context: 1) Specialised MFs for self-adaptive AFR control system of a GDI engine are designed with weight distribution. 2) Based on designed scalar parameters, the CA model with taxonomic dimensions is built for acquiring a customised MF to counter transient scenario changes more effectively. 3) The engine controller with the proposed scheme is real-time validated in a production V6 GDI engine, and its advantage in terms of engine transient control performance is further demonstrated by comparing with a benchmark controller designed based on experience.

One-Step Ahead Control Using Online Interpolated Transfer Function for Supplementary Control of Air-Fuel Ratio in Thermal Power Plants

“…On the other hand, the supplementar control is positive during the second step change, from 1500 s to 2300 s, in order to increa the amount of combustion air of the existing multi-loop control Figure 8. Figures [11][12][13] show the comparison of total air demand, fuel demand, and ma steam pressure of conventional multi-loop control and that of proposed adaptive contro respectively. Because the amplitude of supplementary control is not so large, total air d mands look similar in Figure 11.…”

“…From Figures 11-13, withou affecting the existing power plant operation, the proposed adaptive supplementary con trol can effectively update the AFR control performance. Figures [11][12][13] show the comparison of total air demand, fuel demand, and mai steam pressure of conventional multi-loop control and that of proposed adaptive contro respectively. Because the amplitude of supplementary control is not so large, total air de mands look similar in Figure 11.…”

“…However, it is hard to find advanced research to update the AFR control performance of standard thermal power plants. For AFR control of vehicle engines, there have been many studies with advanced control techniques, such as supervisory control [10], feedback control with Kalman filter estimation [11], generalized predictive control [12], and fuzzy logic control [13]. Fuzzy inference with multiple neural networks (NN) is also applied for the AFR control in the heating furnace of a steelmaking plant [14].…”

Fuzzy-tree-constructed data-efficient modelling methodology for volumetric efficiency of dedicated hybrid engines

“…Luján et al develop an adaptive learning algorithm to increase hidden layer weight update speed for NN-driven volumetric efficiency model [20], wherein this algorithm performs higher learning speed, reduced computational resources and lower network complexities. As a representative of heuristic approaches, the fuzzy inference system is widely used in engine modelling and control due to its excellent self-interpretability and robustness [21][22][23]. As a hybrid of the first two methods, the adaptive neuro-fuzzy inference system has the potential to capture the benefits of both in a single framework [24].…”

Dual-loop online intelligent programming for driver-oriented predict energy management of plug-in hybrid electric vehicles

“…Firstly, a PHEV under research is analysed and its optimal problem is formulated. Subsequently, inspired by the authors' past research [30], interval fuzzy and deep fuzzy prediction methods are developed and compared, in terms of the performance of velocity prediction; meanwhile, a finite-state MC is exploited to learn transition probabilities from the vehicle speed to acceleration. The chaos-enhanced accelerated particle swarm optimization (CASPO) algorithm is harnessed to realize the predictive optimal control for increasing fuel economy and maintaining battery charge level.…”