Locomotion speed is a key performance index of legged robots. However, methods to analyze and improve the locomotion speed capability are seldom developed, especially for six-legged robots. This paper develops a method to analyze and improve the omnidirectional walking speed and the turning speed of six-legged robots. The models of the inverse kinematics and the influence coefficients are built. Making use of the only-position-related property of the influence coefficients, a general optimization model of the locomotion trajectory is established. A two-step optimization method is introduced to solve the optimization problem. Based on the optimization, a comprehensive speed capability analysis is conducted on both omnidirectional walking and turning of the six-parallel-legged robot. The results clearly show the relationships among the speed capability, the walking direction and the duty cycle. The two-step optimization method improves the speed capability by 12.4%–13.2% for turning and 18.5%–20.5% for omnidirectional walking. Finally, the costs of the speed improvement are analyzed, including the stability, the energy consumption and the calculation time.

运动速度是有腿机器人的关键性能指标。但是，很少开发分析和提高运动速度能力的方法，特别是对于六足机器人。本文提出了一种分析和提高六足机器人全向步行速度和转弯速度的方法。建立了逆运动学模型和影响系数。利用影响系数的仅与位置有关的性质，建立了运动轨迹的一般优化模型。引入了两步优化方法来解决优化问题。基于此优化，对六平行腿机器人的全向步行和转弯进行了全面的速度能力分析。结果清楚地表明了速度能力，步行方向和占空比之间的关系。两步优化方法将转弯的速度能力提高了12.4％–13.2％，对全向步行的速度能力提高了18.5％–20.5％。最后，分析了提高速度的成本，包括稳定性，能耗和计算时间。