Research On Distribution Route Optimization Of Electric Logistics Vehicles Considering Mileage Prediction

At present,the carbon emission problem in the urban transportation field has become more and more prominent,and the pursuit of low-carbon and environmentally friendly logistics and distribution methods has become the development trend of the whole society.Due to the features of low consumption and low emissions,electric vehicles(EVs)have developed rapidly in the field of urban distribution.However,due to the problem of "mileage anxiety",this limits the promotion and application of EVs,which will be affected by factors such as battery capacity limitation,power consumption,charging station location and charging time during the distribution process,making them different from traditional vehicles.The path planning is quite different,and the problem is more complicated.Therefore,it is of great significance to predict the mileage based on the actual operation data of EVs,and to study the optimization of the distribution path of EVs on this basis.Based on this,the main research of this paper is as follows.First,this paper analyzes the driving features and charging features of EVs,which lays the foundation for subsequent research.Next,the influencing factors of the vehicle mileage were analyzed,the correlation analysis and partial correlation analysis between the mileage and the influencing factors were carried out,the main influencing factors were determined,and the multiple linear regression model of SOC and the lowest temperature of the battery cell was established to predict the mileage.Secondly,a mathematical model is constructed with the goal of minimizing the total cost,and with the constraints of path,load,time,remaining mileage and number of vehicles.Based on the features of the model,designing an improved adaptive genetic algorithm,and using MATLAB software to program and solve the model.Finally,based on the empirical data of J logistics enterprise in Beijing,two scales of cases are designed for verification,and the algorithm performance before and after improvement is compared.The influence of the vehicle mileage,charging strategy and penalty coefficient changes on the path optimization results is analyzed.The results show that the method for predicting the mileage of EVs in this paper is feasible.The optimal path obtained by using the algorithm designed in this paper takes into account customer satisfaction,and reduces the total distribution cost of the enterprise.Through sensitivity analysis,it is found that: a)The longer the driving range,the lower the number of charging times and the total cost,but when the driving range increases to a certain extent,the total cost cannot be reduced.The optimal driving range in this paper is 204km;b)The less the charging strategy,the less the charging cost,but frequent charging will lead to an increase in the driving cost.The optimal charging strategy in this paper is 25%;c)The larger the penalty coefficient,the more significant the cost advantage of additional vehicles.Enterprises need to balance the relationship between customer satisfaction and total cost.37 figures,29 tables,and 69 references.