PREDICTIVE ANALYSIS SYSTEM AND METHOD FOR ANALYZING AND DETECTING MACHINE SENSOR FAILURES

    公开(公告)号:EP3410308A1

    公开(公告)日:2018-12-05

    申请号:EP18170205.1

    申请日:2018-04-30

    申请人: Deere & Company

    发明人: WU, Yifu

    摘要: A system for performing predictive analysis and diagnostics is disclosed. The system includes a plurality of sensors communicatively coupled to a vehicle electronics unit. The plurality of sensors are configured to generate at least one first signal indicative of a first sensed condition and at least one second signal indicative of a second sensed condition. A remote central processing system is coupled to the vehicle electronics unit. The remote central processing system comprises a remote processor and a remote data storage device, wherein the remote central processing system is configured to receive each of the at least one first and second signals. A predictive diagnostic unit is arranged in the remote data storage device and comprises machine readable instructions that, when executed by the remote processor, causes the system to partition the second signal into a predetermined number of successive time intervals; generate a similarity value based on a comparative analysis between the partitioned second signal and a stored first signal; and determine an estimated degree of failure of a machine component based in part on a computed average of the similarity value.

    SENSOR DATA FUSION FOR PROGNOSTICS AND HEALTH MONITORING
    10.
    发明公开
    SENSOR DATA FUSION FOR PROGNOSTICS AND HEALTH MONITORING 审中-公开
    传感器数据融合预测和健康监测

    公开(公告)号:EP3234870A1

    公开(公告)日:2017-10-25

    申请号:EP15823465.8

    申请日:2015-12-18

    IPC分类号: G06N3/04 G05B23/02

    摘要: A method includes fusing multi-modal sensor data from a plurality of sensors having different modalities. At least one region of interest is detected in the multi-modal sensor data. One or more patches of interest are detected in the multi-modal sensor data based on detecting the at least one region of interest. A model that uses a deep convolutional neural network is applied to the one or more patches of interest. Post-processing of a result of applying the model is performed to produce a post-processing result for the one or more patches of interest. A perception indication of the post-processing result is output.

    摘要翻译: 一种方法包括将来自多个预测和健康监测(PHM)传感器的时间序列数据转换为频域数据。 频域数据的一个或多个部分被标记为指示一个或多个目标模式以形成标记的目标数据。 包含深度神经网络的模型被应用于标记的目标数据。 应用该模型的结果被分类为与一个或多个目标模式相关联的一个或多个离散化PHM训练指标。 输出一个或多个离散化的PHM训练指示符。