Multi-parameter Driver Intention Recognition Based On Neural Network In Ice And Snow Environment

With the development of communication technology and technology,cars are gradually becoming intelligent.The sensors and communication modules onboard can assist the driver in controlling the vehicle.Accurately obtaining the driver’s intention can make the auxiliary system work better,especially when the driver’s control of the vehicle is reduced in the ice and snow environment,the auxiliary system is particularly important.The driving assistance system can assist the driver in maneuvering the vehicle according to the obtained driver’s intention,avoiding the occurrence of inconsistent vehicle behavior with the driver’s intention,improving the driving experience and reducing the probability of accidents.The driving process of a driver in a car will be affected by many factors,and the corresponding operating habits of different drivers are not the same,and the driving styles of different road conditions will also be different.Therefore,this article will analyze the driving characteristics of rainy and snowy weather in the road environment,obtain the driver’s behavior parameters and vehicle parameters under this condition,and conduct separate researches.Based on this,the following are mainly completed:(1)Establish a sample database.Relying on the semi-physical simulation driving experiment platform based on Carmaker software,select multiple drivers with driving experience(including different ages and genders)to complete the data collection work according to the experimental plan,and then analyze and process the data,and initially screen out the parameters that characterize the intention.And use statistical knowledge to analyze and verify the correlation between parameters and intentions and the validity of intention characteristic parameters.According to the selected intention parameters and the collected data samples,the data processing is completed,and the training data sample database of different driving stages has been established.(2)Based on BP neural network(Back Propagation Neural Network),the driver intention recognition model is established.BP neural network is a multi-layer forward neural network based on error back propagation algorithm,which has good adaptive and generalization capabilities.Combined with the selected driver’s intention characteristic parameters,the input parameters of the neural network are determined,the structure of the hidden layer network is determined through experiments and trials,and the output dimensions of the network are determined according to the characteristics of driving intentions,and finally the network structure of the intention recognition model is completed design.(3)Performance verification and evaluation of the driver’s intention recognition model.Complete sample classification according to the established data sample library,and establish model training data samples and model test data samples under different working conditions.The model is built in matlab software,the model is trained through training data samples,and the model is tested and verified by test data samples.The model is evaluated from the perspective of real-time and accuracy.