机械臂时间最优轨迹的样条曲线拟合与智能规划

Spline Curve Fitting and Intelligent Planning of Manipulator Time-Optimal Trajectory

为了解决机械臂关节空间的时间最优轨迹规划问题,提出了改进B样条曲线和优化搜索策略蜂群算法相结合的规划方法。B样条曲线分为近似拟合和插值拟合,近似拟合无法保证经过所有轨迹点,插值拟合存在型值点向控制点反算引起的计算量过大问题,为了兼顾拟合精度和计算量,通过扩展控制点的方法改进了B样条拟合曲线。建立了机械臂关节空间轨迹规划模型,提出了优化搜索策略蜂群算法的求解方法。以蜂群算法为基础,使用高斯伸缩因子调节引领蜂向最优解的学习程度,保证了较大概率向最优靠近,相当概率向外扩展而增加多样性;使用三角变异改进跟随蜂的蜜源搜索策略,综合以上方法提出了优化搜索策略蜂群算法。经仿真验证,改进B样条曲线拟合精度比B样条曲线提高一个数量级以上;优化蜂群算法的规划结果满足约束条件,比传统蜂群算法规划结果减少了9.13%。

To solve the problem of manipulator joint space trajectory time optimal planning,trajectory planning method combin⁃ing improved B-spline curve and optimizing searching strategy bee swarm algorithm is proposed.B-spline curve can be shifted in⁃to approximate fitting and interpolation fitting.Approximate fitting cannot go through all the intermediate points,and interpola⁃tion fitting possesses heavy calculation burden caused by inverse computation of model point to control point.To balance fitting ac⁃curacy and calculation quantity,B-spline curve through expanding control point is improved.Manipulator joint space path plan⁃ning model is built,and solving method by optimizing searching strategy is put forwarded.On the basis of bee swarm algorithm,Gauss stretch factor is used to adjust learning level of leading bee to optimal bee,which ensures leading bee close to the optimal by large probability,and expand to outside by quite probability to improve population diversity.Triangular mutation is used to im⁃prove bee souse searching strategy of following bee.The above method is composed and optimizing searching strategy bee swarm algorithm is raised.Clarified by simulation,fitting accuracy of improved B-spline curve increase by one order of magnitude com⁃pared with B-spline curve.Trajectory planned by optimized algorithm satisfies constraint condition,and decrease by 9.13%com⁃pared with traditional bee swarm algorithm.