The inverse kinematics solution for 6DOF industrial robots has the problems of large computational complexity, poor versatility and singularity. An inverse kinematics solution for industrial robots was proposed based on mind evolutionary algorithm (MEA) optimization BP neural network to address these problems. Several sets of joint angle values were randomly generated within the working range of the robot. Then, the position of the robot end link was obtained by the forward kinematics equation, the end link position was used as model inputs, and the joint angle was taken as the output of model. The model parameters were determined by the training sample data, and the model was used to solve the inverse kinematics of the robot. Compared with the traditional BP and RBF neural network solving methods, the results show that the method has higher precision and strong generalization ability.