Aiming at the problem of assembly sequence planning,an assembly relation model considering geometric feasibility,stability,convergence and redirection of assembly sequence and a mathematical model of assembly sequence optimization based on fitness function were established.An improved cuckoo search algorithm was proposed to solve the assembly sequence planning problem,a three-layer coding scheme based on part number,assembly direction and assembly tool was designed by using random key and minimum position rule.A mixed population initialization strategy combining initialization strategy based on minimum assembly cost and random initialization strategy was designed to improve population quality.The method of population evolution and search was improved,and the population was divided into three subgroups,and the methods of adaptive step flight,standard step flight,crossover and mutation were used to update the population,which improved the convergence speed and solution accuracy of the algorithm.Finally,the effectiveness and superiority of the proposed improved cuckoo algorithm in solving the assembly sequence planning problem were verified by an example application and comparison with other algorithms.