关键词: 大语言模型；车辆测试；自动化测试用例生成；自动化测试脚本编写

Abstract: To enhance the depth of analysis and generation efficiency of vehicle test cases and scripts, and to promote the development of automated testing, this paper focuses on the application and practice of artificial intelligence (AI) technology in vehicle automated testing, mainly covering the generation of automated test cases and the writing of automated test scripts. Specifically, the self-generation method of test cases based on large language models dynamically builds test scenarios with higher coverage by analyzing requirement documents and historical test data. The automated test scripts based on large language models combine reinforcement learning and code generation techniques to achieve the automatic writing and optimization of test scripts, reducing the error rate of manual coding while improving the efficiency of script maintenance. Practical results show that the automated generation system of vehicle test cases and test scripts based on large language models proposed in this paper can effectively increase the coverage of test cases by 18%, improve the efficiency of automated script writing by 50%, and significantly shorten the testing cycle.

Key words: large language model; vehicle testing; automated test case generation; automated test script writing