Test Case Generation Method Based On Analysis Of CVIS Mixed Traffic Scenario Elements

Facing a modern transportation system composed of multiple elements of people,vehicles and roads,traditional traffic management methods have led to increasingly serious traffic congestion.Cooperative vehicle infrastructure system(CVIS)has become the cutting-edge technology and inevitable development trend in the field of intelligent transportation today.Since heterogeneous vehicles of different intelligence levels are integrated into the traffic to form a new mixed traffic environment,traditional CVIS simulation and testing technology can no longer solve the new traffic problems caused by the phenomenon of mixed traffic flow,and typical application scenarios of CVIS cannot fully covers the functional characteristics of mixed traffic.It is of great significance to test the functions of CVIS mixed traffic scenarios and establish scenario test cases.However,the traffic scene elements are numerous and mutually constrained,and the real scenes are complex and changeable and difficult to reproduce.According to the test requirements of CVIS mixed traffic scenarios,this paper studies test case generation method based on analysis of CVIS mixed traffic scenario elements.The specific research content includes the following aspects:(1)Defined the CVIS mixed traffic scene formed by the mixture of heterogeneous traffic subjects in the CVIS environment,and constructed the static and dynamic element library of CVIS mixed traffic scene.Based on the scene element library,the feature analysis and test case description of CVIS mixed traffic scene function are carried out,and the test requirements of CVIS mixed traffic scene are clarified.(2)The analysis method of CVIS mixed traffic scene elements is studied.Firstly,the operation mechanism of mixed traffic flow is analyzed to realize the analytical expression of the mathematical expectation of heterogeneous vehicle proportions.The basic flow-density diagram relationship is established for the mixed traffic flow.Secondly,it analyzes the elements of CVIS mixed traffic scene,constructs a hierarchical model of the scene elements and analyzes their importance,and proposes a scene complexity index to describe the scene complexity.(3)The test case generation method based on the analysis of scene elements is studied.Firstly,the Combinatorial Testing(CT)method is used to generate test cases based on the simulation data,and the discrete test cases are generated based on the element importance index.Secondly,the Monte Carlo method is used based on the real data.The probability density distribution of the scene parameters is estimated and sampled,and the scene importance index is used as the clustering weight to generate randomized primitive scene test cases and risk scene test cases.(4)Analyze the test case generation method based on the analysis of scene elements proposed in the paper.Finally build a simulation environment based on the SUMO software,and inject the generated CVIS mixed traffic typical scenario test cases to verify the effectiveness of the method proposed in this paper.The effectiveness of the method proposed in this paper is simulated and the results are analyzed.The research results show that the analysis method of CVIS mixed traffic scenario elements proposed in this paper can effectively guide the generation of test cases of different complexity with mixed traffic characteristics to meet test requirements.Compared with other CT algorithms,the test cases generation by the CT method based on simulation data can have more complex characteristics and generate macro scenarios that are difficult to achieve in real traffic.The Monte Carlo method based on real data can effectively extract primitive scene test cases and risk scene test cases,and generate micro scenes that conform to the probability distribution of real scene parameters.This paper finally realizes the test of different functional requirements of CVIS mixed traffic scenarios.