This paper proposes an Improved Dung Beetle Optimization algorithm (IDBO) for solving the trajectory optimization problem of a 6-DOF industrial robotic arm under constraints, aimed at achieving optimal time, minimal energy consumption, and minimal impact. A 5-order NURBS curve is employed to interpolate the path points of the robotic arm, and the IDBO is used for multi-objective optimization of this trajectory. To solve the issue of uneven distribution of the initial solution of the Dung Beetle Optimization (DBO) algorithm, which affects convergence speed and search capabilities of the algorithm, a tent chaotic mapping is introduced to increase population diversity. A spiral exploration strategy and an adaptive mixed strategy combining random reverse learning strategy and domain centroid reverse learning strategy are proposed. Simulations on the KUKA KR6 R700–2 robotic arm shows the IDBO algorithm outperforms others like PSO, GWO, and WOA, achieving faster convergence speed and higher solution accuracy. The IDBO not only reduces operation time by 57.44% but also optimizes performance indicators, such as time, energy, and impact, demonstrating its significance for six-degree-of-freedom robot trajectory planning.