Construction of high-precision driving cycle based on Metropolis-Hastings sampling and genetic algorithm

摘要

Due to the increasing complexity of the current traffic environment, the existing construction methods of working conditions have large errors and poor representativeness. Therefore, it is difficult to reflect the fuel consumption and emission of vehicles in the actual road environment. To solve the problem, combine the advantages of the Genetic Algorithm (GA) and Metropolis-Hastings Sampling (MHS) methods, a high-precision construction method-based on stepwise regression characteristic parameter selection for fuel consumption-oriented driving cycles is proposed. Firstly, to improve the construction efficiency and effect of driving cycles, a fuel consumption-oriented characteristic parameter selection method based on stepwise regression is proposed; Then, to overcome the disadvantage of the traditional MC method just taking finite discrete values, MHS is applied to the random transition of the velocity-acceleration state; Finally, the driving cycle construction method is design. According to experimental verification and algorithm comparison, the research results show that the maximum and average relative deviations of characteristic parameters of the driving cycle obtained from the MHS-GA method are 0.074% and 0.48%, respectively, and the relative deviation of fuel consumption is 1.01%. The driving cycle based on the MHS-GA method is more representative than existing methods.