Aiming at the problems of low efficiency and high cost of multi-target point path planning for robotic arms in complex transmission environments, the improved artificial potential field-informed rapidly-exploring random trees star (IAPF-IRRT*) algorithm is proposed to enhance path planning performance. Firstly, a cuboid repulsive field model is introduced to improve the spherical repulsive field model in traditional artificial potential field, and a repulsive field is established for complex obstacles in the transmission environment. Then, ellipsoidal regions are adopted with uniformly distributed positions to enhance the ellipsoidal regions in the IAPF-IRRT* algorithm. Local redundancies are avoided in sampling points in complex transmission environments and search efficiency is improved. Finally, the redundant nodes in the path are optimized by using a triangular optimization method and the path is smoothed by using cubic spline interpolation. Validation is conducted on three sets of obstacle maps of different complexity: simple 3D, complex 3D, and complex transmission environments. The results show that the time efficiency of IAPF-IRRT* algorithms compared with the standard RRT and RRT* algorithms is improved by 44.8%~83.8%, 68.3%~95.2%, and 26.5%~71.8%, respectively. The cost of the path is reduced by 15.5%~35.0%, 14.1%~35.3%, and 31.5%~43.5%, respectively. The number of nodes in the path is reduced by 75.6%~78.8%, 75.0%~78.0%, and 70.4%~72.0%, respectively.