The invention designs an electric vehicle charging process fault early warning method based on an adaptive deep belief network. The method comprises the steps that firstly historical data of each physical quantity in the electric vehicle charging process are collected, and a data set is built; secondly, the data set is divided into a normal charging data set and a fault charging data set, and the normal charging data set and the fault charging data set are preprocessed; then, a Nesterov acceleration adaptive moment estimation algorithm is used to optimize a training process of a deep belief network, the adaptive deep belief network is constructed, and a grey wolf algorithm is used to determine a network structure of the adaptive deep belief network; the normal charging data set is used to train the adaptive deep belief network to obtain an electric vehicle normal charging model, and the fault charging data set and a Pearson coefficient are used to test the early warning performance of the model; and finally, real-time data is input into the normal charging model to predict and output, Pearson coefficients of a predicted value and an actual value are calculated, and if the value of the Pearson coefficient exceeds an expected value, fault early warning is performed, and charging of the electric vehicle is cut off to prevent fire accidents of the electric vehicle.

    本发明设计一种基于自适应深度置信网络的电动汽车充电过程故障预警方法,首先收集电动汽车充电过程各物理量的历史数据,建立数据集;其次将数据集划分为正常充电数据集和故障充电数据集,并对其进行预处理;然后用Nesterov加速自适应矩估计算法优化深度置信网络的训练过程,构建自适应深度置信网络,并采用灰狼算法确定其网络结构;接着再使用正常充电数据集训练自适应深度置信网络,得到电动汽车正常充电模型,并使用故障充电数据集和皮尔逊系数,测试模型预警性能;最后将实时数据输入正常充电模型中预测输出,计算预测值与实际值的皮尔逊系数,若皮尔逊系数的值超过期望值,则进行故障预警,并切断电动汽车的充电,防止其发生起火事故。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Electric vehicle charging process fault early warning method based on adaptive deep belief network


    Weitere Titelangaben:

    基于自适应深度置信网络的电动汽车充电过程故障预警方法


    Beteiligte:
    YANG QING (Autor:in) / GAO DEXIN (Autor:in) / WANG YI (Autor:in)

    Erscheinungsdatum :

    2021-10-01


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

    IPC:    B60L PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES , Antrieb von elektrisch angetriebenen Fahrzeugen / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen



    Electric vehicle AC charging state monitoring and fault early warning method based on deep learning

    YANG QING / GAO DEXIN / WANG YI et al. | Europäisches Patentamt | 2022

    Freier Zugriff

    Electric vehicle high-power DC charging state monitoring and fault early warning method based on deep learning

    GAO DEXIN / WANG YI / YANG QING et al. | Europäisches Patentamt | 2022

    Freier Zugriff

    Electric vehicle charging pile fault early warning system and method based on big data

    ZHENG SHUFEN / ZHENG JIANZHOU / ZHENG YONGJIN | Europäisches Patentamt | 2024

    Freier Zugriff

    Electric vehicle charging equipment fault analysis and early warning method and system

    ZHANG YUANXING / LI BIN / LI TAOYONG et al. | Europäisches Patentamt | 2021

    Freier Zugriff

    Electric vehicle charging early warning method and device

    CAI QUANLI | Europäisches Patentamt | 2022

    Freier Zugriff