根据不同功能测试点在芯片代码中的逻辑深度与相应测试向量覆盖到的测试点多少的关系,对不同测试点设置了相应的权重,提出一种基于自适应遗传算法的激励向量生成方法。实验结果表明,该方法能减少编写约束文件时间,较快自动搜索有针对性的测试激励,提高芯片功能验证的可靠性和仿真效率。 According to the relationship between the logic depth of different functional test points in the chip code and the test points covered by the corresponding test vectors, corresponding weights are set for different test points, and a generation method of the excitation vector based on the adaptive genetic algorithm is proposed. The experimental results show that this method can reduce the time for writing constraint files, search for targeted test stimuli quickly, and improve the reliability and simulation efficiency of chip functional verification.