The test sequence optimization problem is a key problem in the fault diagnosis process; for the test sequence optimization problem of the multi-valued attribute system, the adaptive differential evolution algorithm is used, combined with the characteristics of the multi-valued attribute system, analysed the role of mutation operators in algorithms, and designed the individual coding strategy and two different diagnosis methods, proposed a differential evolution algorithm that integrates Gaussian, Cauchy mutation operators and multi-difference strategies; Through experimental comparison and analysis, the results show that the algorithm can not only be well applied to multi-valued attribute systems, but also when dealing with the test sequence optimization problem of binary attribute systems, compared with the existing algorithms, the number of test points obtained by this algorithm is less. It is expected that the test cost is lower, and it can be used to solve the problem of diagnosis strategy in multi-valued attribute systems.