Fuzzy test data generation method based on multi-population genetic algorithm

The invention relates to a fuzzy test data generation method based on a multi-population genetic algorithm. The method comprises the following steps: generating fuzzy test data based on vulnerability library information, a fuzzy test tool AFL and a network message to construct a plurality of initial populations for population initialization and coding; according to the code coverage rate, designing a fitness function for calculating the individual fitness value of each individual in the encoded population, and according to the individual fitness value, dividing each individual into an elite population or a common population; performing individual crossover and individual mutation operation on individuals in the elite population to obtain new individuals; performing interpopulation crossover operation and population and individual mutation operation on the common population to obtain a new population; and obtaining a final population as the generated fuzzy test data until a genetic algebra or a termination con