基于深度信念网络的计量装置故障溯源研究

针对电网海量大数据中存在异常的电表数据，提出了一种基于深度信念网络的计量装置故障溯源模型。首先，分析了深度信念网络（DBN）模型基本原理，提出了一种智能电表故障分类DBN结构模型，并给出了计量装置故障溯源建模流程；然后，通过建立离线台账样本库、实时用电曲线故障特征样本库，实现了计量装置故障样本库构建；最后，基于实际电表测试和数据异常识别，完成了计量装置的台账样本库溯源测试，并与已有的贝叶斯、决策树、随机森林、决策树提升等分类算法进行比较，测试结果验证了所提模型和方法的正确性和有效性。

Fault Traceability of Metering Device Based on Deep Belief Network

Aiming at the abnormal electricity meter data in massive big data of power grid， a fault traceability model of metering device based on deep belief network is proposed. Firstly， the basic principle of deep belief network （DBN) model is analyzed. A DBN network structure model for intelligent meter fault classification is proposed. And the process of fault traceability modeling of metering device is given. Then， by establishing off-line account sample bank and real-time power curve fault feature sample bank， the fault sample bank of metering device is constructed. Based on actual meter testing and data anomaly recognition， the traceability test of the accounting sample database of the metering device have been completed. Compared with the existing classification algorithms such as Bayes，Decision Tree，Random Forest and Adaboost， the test results verify the correctness and effectiveness of the model and method proposed in this paper.