数控机床节能策略TL-DBN模型设计及优化

Design and Optimization of TL-DBN Model for Energy Saving Strategy of NC Machine Tools

机床运行控制性能受到实际车间环境与生产工艺条件的综合影响,如何提高节能效果是保证机械制造成本的一个关键。为了提高数控机床多台设备协调工作的能力,进而实现节能的效果,设计了一种利用迁移学习方法来实现CNC机床的节能控制过程。建立了以RF实现驱动功能的离散CNC机床能耗调控模型,由此减小等待阶段所消耗的能量。研究结果表明:相比较其他算法,TL-DBN模型决策误差达到3.2%的最低值,迁移学习过程能够优化节能决策的效果,能够使机床达到理想的节能效果。该研究的控制方案为实现生产过程管控提供指导,为后续的控制优化奠定一定的理论基础。

The operation control performance of machine tools is affected by the actual workshop environment and production process conditions.How to improve the energy saving effect is a key to ensure the cost of machinery manufacturing.In order to improve the coordination ability of NC machine tools and realize the effect of energy saving,a transfer learning method was designed to realize the energy saving control process of CNC machine tools.The energy consumption control model of discrete CNC machine tool with RF driving function was established to reduce the energy consumed in the waiting stage.The results show that,compared with other algorithms,the decision error of TL-DBN model reaches the lowest value of 3.2%.The transfer learning process can optimize the effect of energy saving decision,and can make the machine tool achieve the ideal effect of energy saving.The control scheme of this study provides guidance for the realization of production process control and lays a certain theoretical foundation for the subsequent control optimization.