苍南售气系统用气用户表具每日用气数据预测

工商业用户表具受压力、温度、用气量、环境影响，容易产生安全问题。采集工商业用户表具的每日用气数据、运用大数据构建其正常用气行为特征，得到正常用气的温度、压力值，对其安全用气行为进行预测，如发现异常，可以及时提示预警。1、获取用户表具的每日用气数据，包括温度、压力、标况、工况、每日用气量等维度特征，运用大数据分析用气的安全范围，得到用户每台表具的用气量的区间、温度区间、压力区间 2通过用户每台表具用气时间结合GBDT算法进行建模，形成表具画像3、通过DeepFM深度学习算法，进行表具用气气量、正常范围的温度、压力的预测，并结合表具每日用气数据，给与预警提示。

Gas meters for industrial and commercial users are susceptible to impacts from pressure, temperature, gas consumption volume and environmental factors, posing high risks of safety issues. By collecting daily gas consumption data of these meters and leveraging big data to construct normal gas consumption behavior profiles, we can derive the normal temperature and pressure ranges for gas usage, and predict safe gas consumption behaviors to deliver timely early warnings when abnormalities are detected. The specific workflow includes three parts:
1. Collect daily gas consumption data of each user's meters, covering dimensional features including temperature, pressure, standard condition, working condition and daily gas consumption volume. Analyze the safe gas usage range via big data technologies to obtain the gas consumption interval, temperature interval and pressure interval for each individual meter of the user.
2. Establish predictive models by combining the gas usage timing data of each user's meter with the GBDT (Gradient Boosting Decision Tree) algorithm, to generate meter profiles for each device.
3. Adopt the DeepFM deep learning algorithm to predict the gas consumption volume, normal temperature range and normal pressure range of the meters. Combine the prediction results with the daily gas consumption data of the meters to issue targeted early warning prompts.

搜集汇总

数据集介绍

特点

该数据集包含工商业用户表具的每日用气数据，用于预测安全用气行为并提供异常预警。数据规模为1000条，每日更新，包含18个字段，如温度、压力、工况累计等。