Research And Application Of Electric Vehicle Driving Behavior Energy Consumption Scoring Algorithm

At present,the intensification of environmental pollution and the emergency of oil resources have sounded an alarm to us.In the face of the environment and energy crisis,the traditional automobile industry must take action to transform and evolve.New energy vehicles use electric energy as the power source,with zero emission and no pollution.Electrical energy is known to be secondary energy,which can effectively promote energy evolution.Due to the limitation of battery technology concerning its capacity,volume and safety,the electric energy consumption and max driving mileage of new energy vehicles have become the primary indicators of performance,which has been considered as core competitiveness of EV Companies to contend for market share.Currently,the innovation in the battery tech and citywide construction of battery charging/s witching station are two main approaches to relieve the battery life anxiety,which is effective however costly and time-consuming.This paper adopted a new perspective to help optimize the energy consumption and increase the max driving mileage,which is achieved by analyzing and modifying driver’s behavior.By harnessing the power of bigdata and software-based solution,this approach sees a wide application scenarios and values with low cost.we use data mining technology,cloud computing concept to mine and analyze the IoV Data,and carry out the Driving Behavior Scoring Model based on Energy Consumption.By visualizing the performance score and driving suggestions produced by the model,drivers will be directly exposed to and guided by exemplary driving habits that lead to a lower energy COSt.The main work of this research paper is as follows.Firstly,the IoV Data acquisition and storage strategy of this project is elaborated.Based on TBOX data acquisition technology,the training data set of the model was obtained by three sub-steps:data source parsing and storing,relevance analysis of key factors,and key factor calculation according to the inventory vehicles and data universality.Secondly,an in-depth study of energy consumption evaluation factors combined with common driving experience was conducted,and the influencing factors of driving energy consumption were divided into four first-level indexes,namely,activity,driving speed,driving smoothness and energy consumption.It was then subdivided into vehicle identification number,acquisition time stamp,current speed,steering wheel Angle,accelerator pedal state,brake pedal state,total accumulated mileage,driving motor speed,vehicle lateral acceleration,vehicle longitudinal acceleration and battery residual signal.Furthermore,the popular scoring models in the current industry were scrupulously compared,optimized and redesigned.Firstly,a comprehensive matching weight scoring model was proposed which combines classical entropy weight generation and hierarchical analysis.Then,the score card algorithm model of the credit industry was innovatively introduced into the field of big data of the Internet of vehicles,and the energy consumption score was studied by using the score card algorithm based on logistic regression.Real data samples were used to analyze the above two algorithms and compare the results.Finally,the driving behavior energy consumption analysis system was designed.A series of modules such as system development platform,algorithm model training,bigdata processing,front-end and back-end query were carefully developed in detail.The system has been proven effective in terms of functional objectives,and is currently integrated into a certain brand of vehicle product verified by positive market feedback.