基于PSO-SCE的下肢外骨骼机器人步态多目标优化

针对一些传统下肢外骨骼机器人步态优化算法中的寻优精度低和收敛速度慢等问题,提出一种下肢外骨骼机器人步态优化算法,该算法采用粒子洗牌策略和复合形信息互通技术,保留了粒子群优化(Particle Swarm Optimization,PSO)算法与复合形交叉进化(Shuffled Complex Evolution,SCE)算法的优点,通过建立以机器人步态零力矩点(Zero Moment Point,ZMP)稳定裕度和每步驱动能耗为参数的多目标优化函数进行寻优,并在SolidWorks、ADAMS和MATLAB软件中进行联合对比测试,仿真表明:与PSO算法和SCE算法相比,采用PSO-SCE算法得到的机器人步态ZMP稳定裕度增大,驱动能耗平均值分别减小了18.4%和13.5%,机器人能够以较小的能量消耗实现稳定平滑行走。

Multi-objective optimization of gait of lower limb exoskeleton robot based on PSO-SCE algorithm

Aiming at the problems of low optimization accuracy and slow convergence speed in some traditional lower limb exoskeleton robot gait optimization algorithms,an optimization algorithm for lower limb exoskeleton robot gait was proposed,which used particle shuffling strategy and complex information communication technology,and retained the advantages of Particle Swarm Optimization(PSO)algorithm and Shuffled Complex Evolution(SCE)algorithm.By establishing a multi-objective optimization function with robot gait Zero Moment Point(ZMP)stability margin and energy consumption per step as parameters,optimization and joint comparative tests were conducted in SolidWorks,ADAMS and MATLAB software.The experiment shows that,compared with the PSO algorithm and SCE algorithm,the robot gait ZMP stability margin obtained by the PSO-SCE algorithm is increased,and the average driving energy consumption is respectively decreased by 18.4%and 13.5%,the robot can achieve stable and smooth walking with less energy consumption.