Abstract: The low energy exchange efficiency of magnetostrictive transducer is an important reason for the low accuracy of nondestructive testing of bolt. It is very important to determine the appropriate coil structure parameters to improve the energy exchange efficiency. The nondestructive testing model of bolt is established based on Comsol multi-physical field finite element simulation software. Firstly, the influence law of different coil turns and lifting distance on energy exchange efficiency is determined. Then, the beetle antennae search and particle swarm genetic algorithm are applied to the coil parameter optimization, a single objective optimization design model is proposed, which took improving the energy exchange efficiency of the transducer as the optimization objective and coil turns and coil lifting distance as independent variables, and selects the optimal coil structure parameters. Finally, an experimental platform for nondestructive testing of bolt is built, and the values within the value range of independent variables are verified experimentally. The results show that increasing the number of coil turns and reducing the distance of coil lifting can improve the energy exchange efficiency of magnetostrictive transducer. The optimization parameters of the beetle antennae search and particle swarm genetic algorithm are the same and consistent with the influence law obtained by simulation. Compared with particle swarm genetic algorithm, beetle antennae search has simple principle, fewer parameters and less computation, so it has greater advantages in dealing with low dimensional optimization problems. The values of coil parameters obtained by the experiment are the same as the optimization results of the algorithm, which verifies that the beetle antennae search is reliable and fast for the parameter optimization of the anchor transducer.