龙门式多轴自动锁螺丝设备横梁结构优化

Structural Optimization of Crossbeam in the Gantry Type Multi-Axis Automatic Locking Screw Device

为实现龙门式多轴自动锁螺丝设备横梁的轻量化,结合BP神经网络与粒子群优化算法对其横梁进行结构优化。以横梁质量为目标函数,数个关键尺寸为设计变量,变形量及固有频率为约束条件建立数学模型。利用BP神经网络拟合设计变量与约束变量的映射关系,结合已建立的神经网络模型,应用基于Deb可行性规则改进的粒子群算法,在满足约束要求的条件下,寻求各关键尺寸的最优值。优化结果表明,优化后的横梁质量减少29.61%,实现横梁轻量化。

This study aims to achieve the lightweight of the gantry type multi-axis automatic locking screw device, the algorithm combined with BP neural network and particle swarm optimization is applied to optimize the structure of the croSsbeam. The mathematical model was established with the mass of crossbeam as object function, several critical dimensions as the design variable and the deformation and natural frequency as the constraint conditions. BP neural network was employed to fit the mapping relationship between design variables and constraint variables. Under the conditions of meeting the constraints, particle swarm optimization improved based on Deb feasibility rules combines with neural network model has been established to optimize the distribution of the key dimensions. The result indicates the weight of the optimized crossbeam reduced by 29.61% which meets the goal of the lightweight.