Research On The Optimal Path Of Spatial-temporal Correlation Of Traffic Flow Based On Improved NSGA-Ⅲ Algorithm

In recent years,with the continuous increase in the number of cars and the increasing complexity of urban road networks,urban traffic congestion has become increasingly severe.This leads to a decrease in the operational efficiency of the transportation network,longer travel times for people,and rising travel costs.Intelligent transportation systems can effectively alleviate urban road traffic congestion and improve the operational efficiency of the transportation network.Path planning,as an important component of intelligent transportation systems,can provide references for people’s travel and reliable technical support for road planning by relevant departments.Currently,although most urban path planning algorithms use multiple objective functions to find the optimal path,they still suffer from poor convergence and the risk of getting stuck in local optima,and are not suitable for high-dimensional objective spaces.In addition,traditional path planning methods only rely on single road length to select the optimal path,ignoring the impact of spatial-temporal correlations on vehicle speed.To address these issues,the main research contents of this paper are as follows:1.This paper proposes an adaptive-scheduling-based NSGA-Ⅲ algorithm(AS-NSGA-Ⅲ)to address the low population diversity,high computational complexity,and slow convergence speed issues of the NSGA-Ⅲ multi-objective optimization algorithm in high-dimensional objective spaces.Firstly,in the reference point generation,the algorithm dynamically adjusts the number of partitions to control the number of reference points,which avoids the problem of excessive computational complexity caused by too many reference points,as well as the issue of insufficient reference points leading to a loss of population filtering function.Secondly,in the population evolution stage,the algorithm adapts the number of populations dynamically by designing crossover and mutation operators to increase the diversity of solutions.Experimental results demonstrate that the proposed AS-NSGA-Ⅲ algorithm improves the convergence and distribution of the Pareto solution set.Traditional shortest path search algorithms only calculate the shortest path between two roads in a city based on the length of a single road,leading to unreliable optimal path planning.This paper fully considers the influence of spatial-temporal correlation on vehicle speed and proposes an optimal path planning method based on spatial-temporal correlation.Firstly,the k-shell algorithm is used to hierarchically divide the topological graph of the road network based on the node’s k_svalue,so that the algorithm searches towards nodes with larger k-shell values.Secondly,based on the congestion index,the level of congestion of each road is divided and assigned a corresponding speed value.Finally,both are introduced into the Floyd algorithm.The results show that the improved Floyd algorithm proposed in this paper can find optimal paths with longer distances but shorter actual travel times,which better aligns with the actual road conditions.Moreover,it also improves the efficiency of shortest path search to some extent.