Research On Prediction And Path Guidance Of Effective Parking Space In Parking Lot

As the important links of intelligent transportation construction,prediction of effective parking space and guidance of internal paths of parking lot are playing a vital role.The prediction of effective parking spaces refers to the application of scientific methods to predict the usage of parking spaces in the parking lot,which is conducive to travelers to more rationally arrange travel plans,improve travel efficiency and parking lot utilization.Path guidance inside the parking lot means that after the driver reaches the parking lot,it will choose the optimal parking space for the driver and conducts path guidance,which is helpful to shorten the parking time and improve the parking efficiency.This paper uses grey wavelet neural network model to predict parking spaces in parking lots,further analyzes the conflict mechanism in parking lots,and proposes parking path optimization methods.The main research work is as follows:First of all,analyze the relevant theories of parking space prediction and prediction value correction,respectively explore the parking behavior guided by the parking guidance system,and provide a theoretical basis for free parking space prediction and path guidance in the parking lot.Secondly,in view of the characteristics of time volatility,spatial non-uniformity,and resource limitations of changes in effective parking spaces,a gray wavelet neural network prediction model for effective parking spaces is proposed.The initial time series of effective parking spaces are predicted by the gray prediction model,which is used as the input value of the wavelet neural network prediction model,so that the wavelet neural network is predicted,and the predicted result of the effective parking space is obtained.The Markov chain correction model is used to modify the prediction results,transform the single prediction results into prediction intervals,which improves the accuracy of the prediction model.And through an example: compared with the traditional BP neural network prediction model,the gray wavelet neural network effective parking space prediction model has reduced volatility by 15%,stability by 26%,and the fit of the predicted value by 13% And proved the feasibility of the model.Finally,by analyzing the structural characteristics of the parking lot and the internal parking situation,aiming at the scenario where there is no parking induction in the parking lot after the traveler reaches the parking lot according to the predicted information.Based on the improvement of Dijkstra algorithm,considering the situation that the front car is congested due to parking,and the back car needs to bypass or wait for the parking,a route induction optimization model based on traffic conflict in the parking lot is proposed to alleviate the parking conflict caused by parking conflict Parking problem.The simulation results show that when thenumber of cars on the parking path in the parking lot does not reach the maximum limit,parking waiting can be completely avoided,and when the number of cars exceeds the maximum limit,the waiting time for parking can be reduced.And it can reduce the parking time of the car by an average of 17% compared with the traditional improved Dijkstra algorithm,thereby reducing parking waiting,parking queues,and improving the parking efficiency of drivers in large parking lots.The purpose of this study is to use the predicted information of parking spaces to provide travel guidance information for travelers,and after the travelers arrive at the parking lot,transform the "repetitive patrol behavior" into "one-time parking process under the intelligent parking system" To further improve urban travel efficiency and service levels.