To address the multi-objective optimization requirements for motion efficiency, stability, and energy consumption during robotic arm operations, a multi-objective trajectory planning method based on the Improved Multi-Objective Whale Algorithm (IMOWOA) is proposed.A six-degree-of-freedom UR16e robotic arm was adopted as the research subject.A multi-objective fitness function was established. The objectives were to minimize motion time, reduce impact, and optimize energy consumption.Constraints on joint position, velocity, acceleration, and jerk were incorporated to ensure solution feasibility.Seven B-spline curves were employed to generate trajectories in the joint space. Smooth continuity of position, velocity, acceleration, and jerk curves was ensured.The IMOWOA was further enhanced for multi-objective trajectory planning. A variable-limit and grouping mechanism was designed. The mechanism was intended to balance local exploration and global search. It was also intended to improve convergence rate and Pareto solution quality.Different convergence factors were applied to distinct grouping categories. An adaptive sine function was introduced during position updates. A Sine-Tent-Cosine hybrid mapping was used to generate perturbation parameter h. Population exploration capacity was enhanced. Convergence was accelerated. Premature homogenization was mitigated. A Cauchy variation strategy based on differential evolution algorithms was implemented to improve search efficiency. The performance of IMOWOA was evaluated through comparisons with seven multi-objective optimization algorithms, including MOWOA. The experimental results show that superior performance is achieved on key metrics. Optimal values are obtained for Hypervolume (HV) and Inverse Generation Distance (IGD). Representative solutions were selected from the obtained Pareto solution set to generate joint trajectories. The resulting motion curves for each joint are smooth and free of abrupt changes. The effectiveness and practicality of IMOWOA in multi-objective trajectory planning for robotic arms are thus demonstrated.