Online Recognition of Driving Cycle of Hybrid Electric Vehicle

In order to realize the online driving condition recognition of vehicles, a method of online vehicle driving condition recognition of hybrid electric vehicle (HEV) based on cluster analysis was proposed. K-Means algorithm was used to classify all driving conditions in the document library of vehicle simulation software ADVISER, and the driving characteristic parameter intervals of different driving conditions were obtained. In the process of vehicle running, a fixed time interval was divided, the real-time characteristic parameters of the vehicle were calculated periodically, and the current driving condition was obtained by matching with the characteristic parameter interval. The online working recognition module was built in MATLAB/Simulink, and the simulation model of a series hybrid electric vehicle was embedded to simulate and verify. The results show that adding online identification module does not affect the normal running of the vehicle simulation model, the module of random condition online identification effect is good, recognition accuracy reaches more than 86%. Compared with the traditional charge depleting control strategy, the control optimization strategy of HEV based on this method can reduce the equivalent fuel consumption by 10.74%, and improves the fuel economy.