Local random search algorithm with worst priority for selective assembly

The invention discloses a worst-priority local random search algorithm for selective assembly, which comprises the following steps of: firstly, reading size information of parts from a database, storing the size information in a three-dimensional matrix, storing assembly information among the parts in a matching matrix, then establishing an assembly matrix, and matching the assembly matrix with the matching matrix; after a simulation assembly matrix is established according to the assembly matrix and the matching matrix, a matching scheme search program is entered, a random initial selection scheme is generated, searching in a limited range is carried out according to the initial selection scheme, and a matched optimal result is obtained. According to the method, the optimal matching scheme is approximately obtained by adopting the worst-first matching direction and through a local random search and multiple iteration mode, the assembly qualification rate can be improved on the premise that the machining prec