Abstract: In recent years, with the improvement of industrial production requirements, the seven joint manipulator, which is the representative of intelligent and automatic technology, has received extensive attention from scholars at home and abroad. One of the most important key technologies in the research field of the seven joint manipulator is the obstacle avoidance and trajectory control technology of the manipulator. In the field of obstacle avoidance of robot arm, although the traditional gradient projection method based on obstacle avoidance index can achieve the purpose of obstacle avoidance, there will be problems such as the decline of end tracking accuracy and the inability to accurately control the trajectory of joint linkage. In view of the above deficiencies, this paper proposes a new manipulator trajectory control algorithm based on double interpolation based on the analytical solution of the seven-joint manipulator. This method uses the given position vector of the wrist joint center point of the manipulator to deduce the position. The vector interpolation operation equation in the trajectory planner TP, and the rotation angle of the seventh joint's self-motion is obtained according to the position vector of the wrist center point, and the rotation angle value interpolated in each interpolation cycle is added to the analytical solution. The calculated inverse solution joint vector is on the seventh joint angle value, so as to achieve the purpose of joint trajectory control. In addition, a set of joint angle values can be calculated by the general gradient projection method as a reference for analytical solution selection, so as to avoid the effect of singular points. The method based on the modified seven joint manipulator configuration has been tested on the Linux CNC real-time control platform and the matlab simulation platform, and the end accuracy of the manipulator is taken as the measurement index to verify the effectiveness of the joint trajectory control of the manipulator based on the method and the superiority of the end accuracy control compared with the gradient projection method.