基于改进BP神经网络的施工升降机平层控制系统研究

Research on Control System of the Construction Elevator Leveling Based on Improved BP Neural Network

摘要: 针对传统施工升降机平层过程需要人工判断、人工操作、精度较差、智能化不高的问题，提出了将改进的BP神经网络应用到施工升降机中的平层控制系统。该系统基于DSP主控制器，首次使用时通过自学习，记录相应楼层信息，利用优化的BP神经算法对设计的网络进行训练和仿真，在升降机的运行过程中，升降机在与设定楼层高差为L时自动减速，当到达设定楼层时自动精确停靠，实现自动平层过程。

Abstract: Aiming at the problems of traditional construction elevator leveling process, such as manual judgment, manual operation, poor accuracy and low intelligence, an optimization algorithm of improved BP (Back Propagation) neural network to the leveling control system of construction elevator is put forward. The system is based on DSP master controller. When it is first used, it records the corresponding floor information through self-learning, and uses the optimized BP neural algorithm to train and simulate the designed network. During the operation of the elevator, the elevator will automatically slow down when the height difference between the elevator and the set floor is L, and when it reaches the set floor, it will stop automatically and accurately to realize the automatic leveling process.