摘要
为了减小冗余机械臂的工作时间和运动冲击,提出了基于增广lagrange-多学习行为粒子群算法的轨迹优化方法。介绍了7自由度冗余机械臂的构型,以减小工作时间和运动冲击为目标建立了约束优化模型。使用增广拉格朗日乘子法将约束优化问题转化为无约束优化问题。在粒子群算法中引入了3种新型的粒子学习行为,并依据学习行为价值确定粒子选择各学习行为的概率,既保证了粒子多样性也保证了收敛的快速性。经实验验证,多学习行为粒子群算法优化的轨迹在时间和冲击方面好于传统粒子群算法优化轨迹,且改进粒子群算法优化轨迹平滑,运动参数在约束范围内,以上结果验证了增广lagrange-多学习行为粒子群算法在机械臂轨迹优化方面的有效性和优越性。
In order to reduce the working time and motion jerk of redundant manipulator,a trajectory optimization method based on Augmented Lagrange-multiple learning behavior particle swarm optimization algorithm is proposed. The configuration of 7-DOF redundant manipulator is introduced,and the constraint optimization model is established to reduce the working time and motion jerk. The augmented Lagrange multiplier method is used to transform the constrained optimization problem into an unconstrained optimization problem. Three new types of particle learning behaviors are introduced into PSO,and the probability of each learning behavior is determined according to the value of learning behavior,which ensures the diversity of particles and the fast convergence. The experimental results show that the trajectory optimized by multi learning behavior particle swarm optimization algorithm is better than that of traditional particle swarm optimization algorithm in time and impact,and the trajectory optimized by improved particle swarm optimization algorithm is smooth,and the motion parameters are within the constraint range.The above results verify the effectiveness and superiority of augmented Lagrange multi learning behavior particle swarm optimization algorithm in manipulator trajectory optimization.
作者
吴国强
WU Guo-qiang(Huzhou Vocational&Technical College,Zhejiang Huzhou 313000,China)
出处
《机械设计与制造》
北大核心
2023年第1期268-272,277,共6页
Machinery Design & Manufacture
基金
浙江省公益技术研究计划项目(LGN18E050002)。
关键词
冗余机械臂
轨迹优化
增广拉格朗日乘子
新型学习行为
粒子群算法
Redundant Manipulator
Trajectory Optimization
Augmented Lagrange Multiplier
New Learning Behavior
Particle Swarm Algorithm
