设备支路电流异常检测数据

通过分析设备支路电流测点值的变化趋势，构造测点数据变化特征，根据该设备支路电流特征值与正常设备支路电流特征值的偏差幅度、偏差标准差及持续周期，对设备支路的电流异常类型进行区分，（计算公式详见邮件-数据知识产权算法补充-设备支路电流异常检测数据 2024/10/12 (周六) 13:36）供电站管理参考。通过收集电站设备历史支路电流值得变化特征，计算设备支路的电流差异情况（如：偏差幅度、偏差标准差和持续周期等），识别差异是否具有可消缺性。具体步骤：1、数据来源：企业自身的生成数据，包括：各时刻支路电流、电压值等；2、数据清洗，对于异常的生产数据，或完整度不满足需求的数据，采用一些统计方法进行甄别和修复；3、数据应用，对设备应用算法；4、数据成果，标记设备异常类型和异常导致的电量损失，并将算法结果提供业务参考。

By analyzing the variation trends of current measurement values of equipment branch circuits, we construct the variation features of the measurement data, and distinguish the current anomaly types of equipment branch circuits based on the deviation magnitude, deviation standard deviation and duration period between the current feature values of the target equipment branch and those of normal equipment branches. (For the calculation formulas, please refer to the email - Data Intellectual Property Algorithm Supplement - Equipment Branch Circuit Current Anomaly Detection Data, 2024/10/12 (Saturday) 13:36) for power station management reference.

By collecting the historical variation features of branch currents of power station equipment, we calculate the current difference conditions of equipment branches (e.g., deviation magnitude, deviation standard deviation and duration period, etc.), and identify whether the differences are remediable.

The specific steps are as follows:
1. Data Source: Self-generated internal data of the enterprise, including branch current and voltage values at each time stamp, etc.;
2. Data Cleaning: For abnormal production data or data with insufficient completeness to meet the requirements, statistical methods are adopted for identification and repair;
3. Data Application: Apply the algorithm to the equipment;
4. Data Outcomes: Mark the equipment anomaly types and the power loss caused by the anomalies, and provide the algorithm results as a business reference.