The application of Fuzzy-Neural network on control strategy of Hybrid Vehicles

This paper focus on the control strategy of Hybrid Vehicles. In order to increase fuel economy and decrease emitted pollution of hybrid vehicles, firstly a Fuzzy Logic Controller (FLC) is considered in this paper. However, FLC is mainly based on the process knowledge and intuition. In comparison, the Adaptive Neural-Fuzzy Inference System (ANFIS) is a modeling method which primarily based on data. So secondly, this paper presents that the membership functions and rules of FLC could be optimized once ANFIS is trained by actual driving cycle data collected from software ADVISOR. Then the FLC controller block in ADVISOR is rewritten by the optimized membership functions according to the ANFIS training. Some simulation results are compared and discussed: the optimized FLC exhibits better performance in terms of fuel consumption and pollutants emission.