摘要
本文以某冷库运行数据为支撑,对海量数据进行了预处理,获取了冷库在不同时间段内的单次开门持续时长、上中下旬库门总开启时长、总开启次数以及冷风机开启情况、冷冻库内温度等数据,并对这些数据进行探索性分析发现其中的规律,明确了冷库开门的行为特征以及关联温度的变化情况。在这一基础上,利用数据挖掘技术提取了冷库开门行为的典型模式,以及各种开门行为模式与库内温度的关系。最后,采用MLP神经算法模型构建了冷库能耗预测模型,为冷库能耗的准确预测提供了支持。
Based on the operation data of a cold storage,this paper preprocesses massive data,obtains data such as the duration of a single door opening,the total opening time of the door in the first,middle,and last ten-day periods,the total opening times,and the opening condition of the cold fan,as well as the temperature in the cold storage,etc.,in different time periods.Exploratory analysis is conducted to discover the rules among these data,and the behavioral characteristics of the cold storage door opening and the associated temperature changes are clarified.On this basis,based on data mining methods,typical patterns of cold storage door opening behavior were extracted,as well as the relationship between various typical patterns of door opening behavior and temperature in the warehouse.Finally,the MLP neural algorithm model is used to construct a cold storage energy consumption prediction model,which provides support for accurate prediction of cold storage energy consumption.
作者
王雅博
马晓蕾
Wang Yabo;Ma Xiaolei(Tianjin Key Laboratory of Refrigeration Technology,Tianjin University of Commerce,Tianjin,300134,China)
出处
《冷藏技术》
2023年第4期17-23,共7页
Journal of Refrigeration Technology
