<p>Multi task optimization algorithm (MTOA) uses evolutionary algorithm to solve multiple optimization problems at the same time. Multi factor optimization (MFO) is one of the most successful algorithms in MTOA. In the research of multi task evolution optimization, in fact, the knowledge exchange between tasks is beneficial only under the reasonable knowledge transfer, so exploring how to make the knowledge transfer more reasonable is the focus of our research. Moreover, the existing algorithms generally have problems such as premature convergence due to insufficient search diversity and excessive computational complexity in large-scale populations. In order to solve the above problems, a crossover operator with Lévy flight distribution parameters is designed. This algorithm can alleviate the negative transfer of knowledge through balanced exploration while effectively improving the diversity of descendants. In addition, improving the operation efficiency of the algorithm is also an inevitable difficulty and pain point on our research road. We improved the calculation speed of multi factor evolutionary algorithm and updated the population search strategy of non dominated sorting in the algorithm. In this paper, the algorithm is called “fast multifactor evolutionary algorithm with Lévy flight distribution” (FMFEA-LF). In order to verify the effectiveness of the algorithm, we have carried out a large number of experiments on the classical multi-task and multi-objective optimization benchmark problem, and compared the results with some classical MFO. Experimental results show that the proposed algorithm has good performance.</p>