Research On Control Strategy Of Idling Start-stop Based On Multi-information Fusion

Idling start-stop technology is a technology for saving energy and reducing emission during the driving process of vehicles.However,the energy saving effect of idle start-stop system can not reach the ideal state because it does not consider the complex and changeable characteristics of actual traffic roads.Frequent and short shutdown of the starting engine not only fails to achieve the purpose of saving fuel,but also it increases the emission of fuel consumption and affects the comfort of driving.On the other hand,the frequent stop and start of the engine wears out the starting motor and the related firmware of the engine,it reduces the service performance of the battery and it increases the deterioration of exhaust emissions at the start moment.Therefore,the study of the idle start-stop control strategy based on multi-information fusion,it can improve the performance of the idle start-stop system,reduce the invalid idle stop and frequent start-stop,which is of great significance to energy saving and emission reduction.The main research contents of this paper are as follows:(1)Construction of vehicle driving cycleAccording to the vehicle driving data collected,the data is preprocessed and divided into kinematics segments,and the characteristic parameters of each kinematics segment are extracted.Principal component analysis is used to reduce the dimension of the characteristic parameters.Secondly,the kinematics fragments are divided into four categories by K-means clustering,and the congestion degree of each category is analyzed.Finally,on the basis of the correlation coefficient method,the representative kinematic sections were selected in descending order and time to construct four driving cycle classified based on the degree of traffic congestion,which are respectively severe congestion,congestion,relatively smooth and smooth driving cycle.The four types of driving cycle constructed here are used as the basis for the subsequent vehicle speed prediction and idle start-stop control strategy research.(2)Traffic control signal acquisitionThe length of stopping time of a vehicle is one of the conditions that determines whether the vehicle should use idle start-stop,it can be determined by obtaining traffic control signals.Firstly,the collected traffic signal light image is converted from RGB color space to HSI color space.Based on the HSI image,the red region containing the signal light information is segmented,and the red signal light and red light countdown digital shapes are identified based on the region descriptor.In order to identify the red light countdown digital,a standard digital template library is established,and features are extracted from the shape of the countdown digital to carry out template matching,and identify the red light countdown digital.Consider the red light countdown at here as one of the information conditions of the idle start-stop control strategy.(3)Research on forecasting algorithm of driving cycleThe four types of driving cycle constructed above are used as the research basis,and the feasibility of driving cycle prediction is judged through the correlation analysis of the speed value of driving cycle.Put forward by using the Markov method and BP neural network method set up speed prediction models for four types of driving cycle respectively,and the error curve and RMSE value analysis of prediction results and analysis prove that BP neural network has better prediction effect.Therefore,the prediction results of BP neural network are used to further calculate the length of idle speed.Finally,the predicted length of idle time is taken as one of the information conditions of the start-stop control strategy for idle speed.(4)Research on control strategy of idling start-stop based on multi-information fusionThe original idle start-stop strategy usually only considers the driver’s intention and the vehicle’s own information condition,but does not consider the influence of the traffic light signal and the length of the stop time on the performance of start-stop system.In this paper,the red light countdown time and the predicted stop time are added as the start-stop judgment conditions to carry out the idle start-stop control with multi-information fusion.With the use of start-stop control strategy in this paper,the number of invalid idle speeds can be theoretically reduced by about 40%,and the invalid idle speeds can be reduced by 16%,6.9%,8.6% and 3% under severe congestion,congestion,relatively unobstructed and unobstructed cycle.According to the fuel analysis,the fuel consumption is reduced by 25% and 4.6% under severe congestion and unobstructed cycle under no strategy,and respectively reduced by 5% and 1.6%under the original strategy.To sum up,the idle start-stop control strategy based on multi-information fusion proposed in this paper can effectively reduce the invalid idle speed,fuel consumption and harmful gas emission.