姿态能力最优的2PUR-PSR并联机构尺度综合

Dimensional synthesis of a 2PUR-PSR parallel manipulator with optimal orientational capability

以两转动一移动2PUR-PSR并联机构为研究对象,应用智能算法求解机构尺度参数优化设计问题。为提高机构的姿态能力,将其动平台输出轴设置为偏置形式。采用方位角和倾摆角描述动平台轴线姿态,建立机构位置反解数学模型。以旋量理论为工具,建立机构运动/力传递性能评价指标,给出优质传递姿态工作空间和优质传递姿态能力(Good Transmission Orientational Capability,GTOC)定义和计算方法。以GTOC最大为目标,建立机构尺度参数约束优化设计模型,并应用混合社会认知优化(Hybrid Social Cognitive Optimization,HSCO)算法求解该问题。给出机构尺度优化设计实例,结果表明,模型和算法可行有效。还比较了社会认知优化(Social Cognitive Optimization,SCO)、粒子群优化(Particle Swarm Optimization,PSO)、差分进化(Differential Evolution,DE)和HSCO算法求解该机构综合实例的优化性能,数值试验显示,HSCO算法的计算效率优于SCO,PSO和DE算法。

In this article,with the 2PUR-PSR parallel mechanism equipped with two-rotational-and-one-translational degrees of freedom as the research object,the intelligence algorithm is adopted for optimal dimensional parameters of the parallel manipulator.The deviating output axis of the moving platform enhances the orientational capability of the parallel manipulator.The mathematic model of inverse displacement solutions of the parallel manipulator is set up by means of the orientational depiction of the output axis with the azimuth and tilt angles.With the screw theory as a mathematic tool,the evaluation indices for motion/force transmission performance of the parallel manipulator are worked out;the definitions and calculation methods of Good Transmission Orientational Workspace and Good Transmission Orientational Capability(GTOC)are identified.The constrained optimization model is set up to formulate the dimensional parameters by means of the maximized GTOC,and the Hybrid Social Cognitive Optimization(HSCO)algorithm is adopted to solve this problem.An example of dimensional optimization design is given to suggest that the presented model and algorithm are feasible and effective.In addition,a comparative test is carried out on computational performance when Social Cognitive Optimization(SCO),Particle Swarm Optimization(PSO),Differential Evolution(DE),and HSCO algorithms are adopted.The numerical results indicate that HSCO outperforms SCO,PSO and DE in terms of computational efficiency.