The invention discloses an elevator abnormal energy consumption recognition and alarm method which comprises the following steps: inputting a training data set into an LSTM neural network for training and testing, and outputting a trained prediction model and a prediction error; according to the method, elevator energy consumption abnormal data are preliminarily screened through prediction errors, on this basis, 0 data are recognized, k-means clustering and a CART algorithm are adopted to perceive energy consumption abnormal data, sudden change data are detected in a typical scene analysis mode, a decision making process in an early warning strategy is provided in combination with manual rules, and alarm decision making is conducted on the abnormal data. According to the method, 0 data, energy consumption abnormal data and abrupt change data can be effectively identified, and an alarm decision is set for the corresponding abnormal data, so that the false alarm rate and the missing report rate are lower than those of a traditional alarm strategy.
本发明公开一种电梯异常能耗的识别与报警方法,包括如下步骤:将训练数据集输入LSTM神经网络进行训练和测试,输出训练好的预测模型和预测误差;通过预测误差对电梯能耗异常数据进行初步筛选,在此基础上识别0数据,采用k‑means聚类和CART算法感知用能异常数据,采用典型场景分析的方式检测突变数据,并结合人工规则提出了预警策略中的决策流程,对异常数据进行报警决策。本发明能够有效识别0数据、用能异常数据和突变数据,并对对应异常数据设置报警决策,使得误报率和漏报率相较于传统报警策略更低。
Identification and alarm method for abnormal energy consumption of elevator
一种电梯异常能耗的识别与报警方法
2023-04-25
Patent
Electronic Resource
Chinese
IPC: | B66B Aufzüge , ELEVATORS |
European Patent Office | 2023
|Elevator running state detection device and abnormal state alarm method
European Patent Office | 2023
|ELEVATOR ALARM DEVICE AND ELEVATOR ALARM NOTIFICATION METHOD
European Patent Office | 2016
|