The invention discloses an elevator idle stop scheduling method based on a long and short term memory network. The method comprises the following steps that (1), an elevator controller records a floor number initiated by an elevator taking demand and a recorded value change condition of a weighing module when an elevator runs; (2), an LSTM algorithm in the elevator controller takes the recorded data as input for training, and after convergence, the LSTM predicts the initiation time of the elevator taking demand of each floor on the same day according to the data of a past period of time; and (3), after the elevator completes operation and is in an idle state, the elevator stops on the floor where the request is expected to be initiated earliest according to the elevator taking requirement predicted through the LSTM algorithm. The method can effectively solve the problem of invalid stop scheduling of an elevator in the prior art.
一种基于长短时记忆网络的电梯空闲停靠调度方法,包括以下步骤;1)电梯控制器记录电梯运行时乘梯需求发起的楼层号和称重模块的记录值变化情况;2)电梯控制器中LSTM算法以记录的数据作为输入进行训练,待收敛后,LSTM将以过去一段时间的数据预测当天各楼层乘梯需求的发起时间;3)电梯完成作业处于空闲状态后,将根据LSTM算法预测的乘梯需求,停靠在预计最早发起请求的楼层。本发明能够有效克服现有技术中电梯无效停靠调度的问题。
Elevator idle stop scheduling method based on long and short term memory network
一种基于长短时记忆网络的电梯空闲停靠调度方法
2021-09-24
Patent
Electronic Resource
Chinese
Elevator idle state stop floor analysis system based on big data
European Patent Office | 2021
|