Research On Dynamic Path Planning Algorithm Based On Real-time Location Services

With the arrival of traffic big data,it is easier to collect and analyze traffic data.Real-time location,traffic flow and speed data generated by various mobile terminals have become an important source of traffic big data.In order to make effective use of traffic data and provide accurate and timely feedback and prediction of road condition information for path planning service,a dynamic path planning algorithm based on real-time location service was developed.The research methods and results of this paper are as follows:(1)Short-term traffic flow prediction model based on convolutional neural networkA Convolutional Neural Networks(CNN)prediction model is proposed,which is based on the temporal and spatial characteristics of the short-time traffic flow data.The model consists of two convolutional pooling layers and three full connection layers to automatically extract the spatial and temporal characteristics of traffic flows and convert them into a two-dimensional characteristic matrix.CNN learns these characteristics to optimize the prediction model.By comparing the predicted results with the actual data,the effectiveness of the method is evaluated and compared with other models.The results show that CNN outperforms Long Short-Term Memory and Gated recurrent in accuracy.(2)Dynamic path planning algorithm considering short-term traffic flow predictionIn order to adapt to traffic conditions dynamically to avoid traffic congestion and reduce travel time,a dynamic path planning algorithm based on short-term traffic flow prediction information is proposed.Based on the prediction model of short-time traffic flow data and realtime traffic information,the algorithm combines the predicted driving speed with the urban road network weight,and uses the improved dynamic ant colony algorithm(D-ACO)to solve the optimal path.For verifying the reliability and practicability of the D-ACO algorithm,two examples of large-scale emergency evacuation and urban traffic congestion avoidance are carried out in the complex and changeable urban road network environment to test the algorithm.Experiments show that the algorithm can not only find the shortest travel time path in daily travel,but also achieve the shortest evacuation time under the premise of minimizing global congestion and maximizing road utilization.The innovation of this paper is as follows: 1)The CNN prediction model is proposed to predict traffic flow and speed parameters,and the mean square error(MSE)is used to evaluate.2)The traffic message channel in the GIS software to bind the driving speed and the road section in the road network to realize the timely updating of the road weight by the CNN prediction results.3)The D-ACO algorithm based on CNN prediction is proposed to solve the optimal path in dynamic road network environment.The results show that the dynamic path planning algorithm based on real-time location service can more accurately reflect the real-time traffic situation of the given location in the complex road network environment,reduce travel time and avoid traffic congestion to the maximum extent.