<p>This paper proposes a novel motion planning framework for the efficient trajectory generation of the complex boom system in intelligent concrete pumping trucks. First, the framework is based on energy-optimal principles and utilizes a gradient projection approach to achieve mapping between the Cartesian and joint spaces. An artificial neural network (ANN)-based joint deformation compensation model is designed to obtain low-energy and high-precision inverse kinematics solutions. Second, to further enhance the safety and efficiency of boom motion planning, the framework incorporates a direct connection dimensionality reduction (DCDR) strategy into a rapidly-exploring random tree star (RRT*) algorithm guided by the artificial potential field (APF). The acquired paths are smoothed via B-spline curve optimization. Finally, simulation and experimental results verify that the proposed framework can effectively resolve critical challenges in pumping truck boom motion planning, providing a systematic solution for intelligent pumping truck automation.</p>