WO2018028086A1 - 电子设备健康监测预警系统和方法 - Google Patents

电子设备健康监测预警系统和方法 Download PDF

Info

Publication number
WO2018028086A1
WO2018028086A1 PCT/CN2016/107704 CN2016107704W WO2018028086A1 WO 2018028086 A1 WO2018028086 A1 WO 2018028086A1 CN 2016107704 W CN2016107704 W CN 2016107704W WO 2018028086 A1 WO2018028086 A1 WO 2018028086A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
sensor
control device
prediction
embedded control
Prior art date
Application number
PCT/CN2016/107704
Other languages
English (en)
French (fr)
Inventor
陈义强
黄云
雷登云
陆裕东
恩云飞
何春华
王力纬
Original Assignee
工业和信息化部电子第五研究所
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 工业和信息化部电子第五研究所 filed Critical 工业和信息化部电子第五研究所
Priority to US15/556,702 priority Critical patent/US10458823B2/en
Publication of WO2018028086A1 publication Critical patent/WO2018028086A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
    • G01D21/00—Measuring or testing not otherwise provided for
    • G01D21/02—Measuring two or more variables by means not covered by a single other subclass
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
    • G01D3/00—Indicating or recording apparatus with provision for the special purposes referred to in the subgroups
    • G01D3/08—Indicating or recording apparatus with provision for the special purposes referred to in the subgroups with provision for safeguarding the apparatus, e.g. against abnormal operation, against breakdown
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/28—Testing of electronic circuits, e.g. by signal tracer
    • G01R31/2801—Testing of printed circuits, backplanes, motherboards, hybrid circuits or carriers for multichip packages [MCP]
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/28—Testing of electronic circuits, e.g. by signal tracer
    • G01R31/2801—Testing of printed circuits, backplanes, motherboards, hybrid circuits or carriers for multichip packages [MCP]
    • G01R31/281—Specific types of tests or tests for a specific type of fault, e.g. thermal mapping, shorts testing
    • G01R31/2817—Environmental-, stress-, or burn-in tests

Definitions

  • the present invention relates to the field of fault prediction technology, and in particular, to an electronic device health monitoring and early warning system and method.
  • PHM Prognostics and Health Management
  • PHM Prognostics and Health Management
  • the PHM system enables the transition from traditional sensor-based diagnostics to intelligent system-based predictions, greatly facilitating the process of state maintenance to replace post-repair and preventive maintenance.
  • the PHM system can determine its current status and the possibility of failure in a timely period of time in a timely and accurate manner, and make recommendations for assistance in the use and maintenance activities.
  • the traditional failure warning method is that when the critical failure mechanism of the integrated circuit occurs and fails, the early warning circuit will output an alarm signal.
  • the reference data is stored in the reference device, and the stress device inputs the parameters from the input pin of the integrated circuit for testing, and the test data is obtained from the output pin of the integrated circuit, and the test data and the reference data are compared by the comparison circuit. And realize the early warning output.
  • the traditional failure warning method requires parameter input and data acquisition for the pins of the integrated circuit. It is only applicable to the failure warning of CMOS integrated circuits, and it is impossible to predict the failure of electronic equipment.
  • An electronic device health monitoring and early warning system includes a sensor and an embedded control device, the sensor is disposed in an electronic device, and the embedded control device is connected to the sensor,
  • the sensor is configured to perform physical parameter monitoring on a host electronic system circuit board of the electronic device, and send the obtained sensor data to the embedded control device; the sensor data Including at least one of current data, vibration data, temperature data, and voltage data;
  • the embedded control device is configured to perform feature extraction on the sensor data to obtain feature data, and perform real-time analysis and prediction according to the feature data, and obtain a prediction result and display the same.
  • An electronic device health monitoring and early warning method includes the following steps:
  • the sensor performs physical parameter monitoring on the host electronic system circuit board of the electronic device, acquires sensor data, and sends the data to the embedded control device;
  • the sensor data includes at least one of current data, vibration data, temperature data, and voltage data;
  • the embedded control device performs feature extraction on the sensor data to obtain feature data
  • the embedded control device performs real-time analysis and prediction based on the feature data, and obtains a prediction result and displays it.
  • the physical parameter monitoring of the host electronic system circuit board of the electronic device is performed by the sensor, the sensor data is acquired and sent to the embedded control device, and the sensor data includes current data, vibration data, temperature data and voltage. At least one of the data.
  • the embedded control device extracts the feature data from the sensor data, and performs real-time analysis and prediction based on the feature data, and obtains the prediction result and displays it, providing the user with real-time health monitoring and real-time prediction information of the host electronic system circuit board. It can monitor the performance degradation process of electronic equipment in real time, predict the degradation trend of electronic equipment performance, and realize the fault prediction and health management functions of electronic equipment.
  • FIG. 1 is a structural diagram of an electronic device health monitoring and early warning system in an embodiment
  • FIG. 2 is a structural diagram of an embedded control device in an embodiment
  • FIG. 3 is a flow chart of an electronic device health monitoring and warning method in an embodiment
  • FIG. 4 is a flow chart of an embedded control device performing real-time analysis and prediction based on feature data in an embodiment to obtain a prediction result.
  • an electronic device health monitoring and early warning system includes a sensor 110 and an embedded control device 120 , and the sensor 110 is disposed in the electronic device, and the embedded control Device 120 is coupled to sensor 110.
  • the sensor 110 is configured to perform physical parameter monitoring on the host electronic system circuit board of the electronic device, and send the acquired sensor data to the embedded control device 120.
  • the number and type of sensors 110 are not unique, and may be one or more, or one or more types. Depending on the type of sensor, the type of sensor data may also be different.
  • the sensor data may specifically include at least one of current data, vibration data, temperature data, and voltage data.
  • the layout of the sensor 110 can be designed for the host electronic system board to monitor the required physical parameters.
  • the sensor 110 includes a current sensor, a vibration sensor, a temperature sensor, and a voltage sensor connected to the embedded control device 120, and respectively monitors the host electronic system circuit board, and the corresponding sensor data includes current data, vibration data, Temperature data and voltage data.
  • the four physical parameters of the host electronic system circuit board are collected and sent to the embedded control device 120 for health prediction, which ensures that the prediction result is more in line with the actual situation and improves the prediction accuracy.
  • the specific types of sensors 110 include, but are not limited to, the above four, and may include other embeddable sensors.
  • the embedded control device 120 is configured to perform feature extraction on the sensor data to obtain feature data, and perform real-time analysis and prediction based on the feature data to obtain a prediction result and display the same.
  • the feature data is obtained by feature extraction of the sensor data, and the physical state of the host electronic system circuit board is characterized for use as a subsequent health prediction.
  • the specific manner in which the embedded control device 120 performs feature extraction on the sensor data is not unique.
  • the feature data of the same type of sensor data collected at different positions of the host electronic system circuit board may be extracted at the same time to obtain feature data.
  • the sensor data of the same type collected by the host electronic circuit board at the same position is extracted to obtain feature data.
  • the type of feature data is also not unique, and may specifically include a mean or a mean square error.
  • the temperature monitoring of the different positions of the host electronic system circuit board may be simultaneously performed by using a plurality of temperature sensors, and the average value is calculated according to the plurality of temperature data collected at the same time. Data; or temperature monitoring of the same position of the host electronic system circuit board by the temperature sensor, and obtaining a plurality of temperature data in each acquisition period to obtain an average value as the characteristic data.
  • the specific type of the embedded control device 120 is not unique, and may specifically be a SoPC embedded device. Or SoC embedded device.
  • the embedded control device 120 is a SoPC embedded device, and integrates a system design such as a processor, a memory, an I/O (input/output) port, and other functional modules required by the user into one device, and constructs one into one device.
  • SoPC embedded devices are flexible in design, scalable, scalable, scalable, and have system programmability for both hardware and software.
  • the embedded control device 120 includes an FPGA (Field-Programmable Gate Array) logic device 121, an embedded processor 122, and a display 123.
  • the FPGA logic device 121 is connected.
  • the sensor 110, the embedded processor 122 is coupled to the FPGA logic device 121 and the display 123.
  • the embedded processor 122 acquires the sensor data output by the sensor 110 through the FPGA logic device 121, performs feature extraction on the sensor data to obtain feature data, and performs real-time analysis and prediction according to the feature data to obtain a prediction result, and sends the prediction result to the display 123. display.
  • embedded control device 120 also includes a memory 124 coupled to embedded processor 122 and an application programming interface 125.
  • the memory 124 is used to store data such as sensor data and prediction results
  • the application programming interface 125 is used to provide a program access interface for application development and access, thereby improving operational convenience.
  • FPGA logic device 121 can be an FPGA logic device with a soft core or a hard core.
  • the embedded processor 122 has a kernel layer, a service layer, and an application layer.
  • the kernel layer includes an operating system kernel and a sensor driver.
  • the operating system kernel is an operating system kernel for the SoPC
  • the sensor driver is used to drive the sensor in combination with the operating system kernel.
  • the service layer includes a data-driven prediction algorithm model and an application programming interface.
  • the data-driven prediction algorithm model is used to acquire sensor data according to the interface provided by the operating system kernel, perform feature extraction on the sensor data, and perform real-time analysis on the feature data. prediction.
  • the application layer includes a fault prediction and health management APP (Application) and access to the APP.
  • the fault prediction and health management APP is used to send the prediction result to the display 123 for display, and provides the user with real-time health monitoring and real-time prediction information of the host electronic system circuit board.
  • Staff can design operating system kernels, sensor drivers, data-driven predictive models, and applications for fault prediction and health management applications by accessing the application programming interfaces provided by the APP connection service layer.
  • the sensor 110 acquires sensor data in real time, and the obtained sensor data is passed through the FPGA logic After the sensor driver and the operating system kernel are transferred in the kernel layer, the feature extraction is performed in the data-driven prediction model in the service layer, and the prediction data is used to predict the feature data in real time, and the fault prediction in the application layer is performed.
  • the health management APP displays the predicted results obtained, and provides users with real-time health monitoring and real-time prediction information of the host electronic system board.
  • the embedded control device 120 performs real-time analysis and prediction based on the feature data, and the specific manner of obtaining the prediction result is not unique, and any algorithm with a prediction function can be used for analysis and prediction.
  • any algorithm with a prediction function can be used for analysis and prediction.
  • extended Kalman filtering, unscented Kalman filtering, and particle filtering algorithms can be used for real-time analysis and prediction. These algorithms are based on Kalman filtering.
  • filtering means Given N measured output data y 1 , y 2 , ..., y N , the system state x N+p after p step is predicted. Among them, filtering means:
  • the one-step prediction and the two-step prediction are:
  • a and C are preset matrices
  • w k is the mean value of 0
  • the variance is Q uncorrelated process noise
  • v k is the mean value of 0
  • the variance is R irrelevant measurement noise
  • w k , v k are irrelevant.
  • the embedded control device 120 uses the Kalman prediction algorithm to perform real-time analysis and prediction based on the feature data, and specifically includes the following steps:
  • Kalman filter filtering is performed on the feature data to obtain filtered data. Specifically:
  • K k P' k C T (CP' k C T +R) -1 ,
  • a and C are preset matrices
  • a T and C T respectively represent transpose of matrices A and C, respectively
  • Q and R are variances of uncorrelated process noise and uncorrelated measurement noise, respectively
  • y k represents feature data.
  • the Kalman prediction is performed based on the filtered data to obtain a prediction result. Specifically:
  • N is the number of feature data
  • p is the number of prediction steps.
  • the electronic device health monitoring and early warning system performs physical parameter monitoring on the host electronic system circuit board of the electronic device through the sensor 110, acquires sensor data and sends the data to the embedded control device 120, and the sensor data includes current data, vibration data, temperature data, and voltage. At least one of the data.
  • the embedded control device 120 performs feature extraction on the sensor data to obtain feature data, and performs real-time analysis and prediction based on the feature data, and obtains prediction results and displays, and provides real-time health monitoring and real-time prediction information of the host electronic system circuit board for the user. It can monitor the performance degradation process of electronic equipment in real time, predict the degradation trend of electronic equipment performance, and realize the fault prediction and health management functions of electronic equipment.
  • an electronic device health monitoring and early warning method is implemented based on the above-described electronic device health monitoring and early warning system. As shown in FIG. 3, the method includes the following steps:
  • Step S110 The sensor performs physical parameter monitoring on the host electronic system circuit board of the electronic device, acquires sensor data, and sends the data to the embedded control device.
  • the sensors are disposed in the electronic device, and the number and type of the sensors are not unique, and may be one or more, or one or more types. Depending on the type of sensor, the type of sensor data may also be different.
  • the sensor data may specifically include at least one of current data, vibration data, temperature data, and voltage data.
  • the layout of the sensor can be designed for the host electronic system board to monitor the required physical parameters.
  • the sensor includes a current sensor, a vibration sensor, a temperature sensor, and a voltage sensor, and respectively monitors the host electronic system circuit board, and the corresponding sensor data includes current data, vibration data, temperature data, and voltage data.
  • the four physical parameters of the host electronic system board are collected and sent to the embedded control device for health prediction to ensure that the prediction results are more in line with the actual situation and improve the prediction accuracy.
  • the specific types of sensors include, but are not limited to, the above four types, and may include other embeddable sensors.
  • Step S120 The embedded control device performs feature extraction on the sensor data to obtain feature data.
  • the feature data is obtained by feature extraction of the sensor data, and the physical state of the host electronic system circuit board is characterized for use as a subsequent health prediction.
  • the specific method for feature extraction of sensor data is not unique. It may be to extract feature data of the same type of sensor data collected at different positions of the host electronic system circuit board at the same time to obtain feature data; or it may be a host electronic system circuit board. Multiple sensor data of the same type collected at the same location is extracted to obtain feature data.
  • the type of feature data is also not unique, and may specifically include a mean or a mean square error.
  • the specific type of embedded control device is not unique, and may specifically be a SoPC embedded device or an SoC embedded device.
  • Step S130 The embedded control device performs real-time analysis and prediction based on the feature data, and obtains the predicted result and displays it.
  • the embedded control device performs real-time analysis and prediction based on the feature data, and the specific method for obtaining the prediction result is not unique.
  • the Kalman filter, the unscented Kalman filter and the particle filter algorithm can be used for real-time analysis and prediction. Both are based on Kalman filtering.
  • the Kalman prediction algorithm is used in step S130 to perform real-time analysis and prediction according to the feature data, and specifically includes steps S132 and S134.
  • Step S132 performing Kalman filter filtering on the feature data to obtain filtered data. Specifically:
  • K k P' k C T (CP' k C T +R) -1 ,
  • a and C are preset matrices
  • a T and C T respectively represent transpose of matrices A and C, respectively
  • Q and R are variances of uncorrelated process noise and uncorrelated measurement noise, respectively
  • y k represents feature data.
  • Step S134 Performing a Kalman prediction according to the filtered data to obtain a prediction result. Specifically:
  • N is the number of feature data
  • p is the number of prediction steps.
  • the above-mentioned electronic device health monitoring and early warning method performs physical parameter monitoring on the host electronic system circuit board of the electronic device through the sensor, acquires sensor data and sends the data to the embedded control device, and the sensor data includes current data, vibration data, temperature data and voltage data. At least one of them.
  • the embedded control device extracts the feature data from the sensor data, and performs real-time analysis and prediction based on the feature data, and obtains the prediction result and displays it, providing the user with real-time health monitoring and real-time prediction information of the host electronic system circuit board. It can monitor the performance degradation process of electronic equipment in real time, predict the degradation trend of electronic equipment performance, and realize the fault prediction and health management functions of electronic equipment.

Abstract

一种电子设备健康监测预警系统和方法，通过传感器（110）对电子设备的宿主电子系统电路板进行物理参数监测，获取传感器数据并发送至嵌入式控制装置（120），传感器数据包括电流数据、振动数据、温度数据和电压数据中的至少一种。嵌入式控制装置（120）对传感器数据进行特征提取得到特征数据，并根据特征数据进行实时分析与预测，得到预测结果并显示，为用户提供宿主电子系统电路板实时健康监测及实时预测信息。可实时监测电子设备的性能退化过程，预测电子设备性能退化趋势，实现了对电子设备的故障预测与健康管理功能。

Description

电子设备健康监测预警系统和方法 技术领域
本发明涉及故障预测技术领域，特别是涉及一种电子设备健康监测预警系统和方法。
背景技术
故障预测与健康管理(Prognostics and Health Management，PHM)技术利用尽可能少的传感器采集系统各种数据信息，采用智能推理算法来评估系统自身健康状态，在系统故障发生前对其故障进行预测，并根据可利用资源信息提供维修保障措施以实现系统视情维修。PHM系统可实现由传统的基于传感器诊断转向基于智能系统预测，极大地促进了状态维修取代事后维修和预防性维修的进程。PHM系统能及时、准确地确定其当前状态以及在未来一段时间内发生故障的可能性，并对使用、维修活动做出辅助决策建议。
传统的失效预警方法是当集成电路关键失效机理发生而失效时，预警电路将输出报警信号。具体在参考器件中存储参考数据，并由应力器件从集成电路的输入管脚输入参数进行测试，并从集成电路的输出管脚进行检测得到测试数据，通过比较电路对测试数据和参考数据进行比较而实现预警输出。传统的失效预警方法需要对集成电路的管脚进行参数输入和数据采集，只适用于CMOS集成电路的失效预警，无法对电子设备进行故障预测。
发明内容
基于此，有必要针对上述问题，提供一种适用于电子设备的电子设备健康监测预警系统和方法。
一种电子设备健康监测预警系统，包括传感器和嵌入式控制装置，所述传感器设置于电子设备内，所述嵌入式控制装置连接所述传感器，
所述传感器用于对所述电子设备的宿主电子系统电路板进行物理参数监测，并将获取得到的传感器数据发送至所述嵌入式控制装置；所述传感器数据 包括电流数据、振动数据、温度数据和电压数据中的至少一种；
所述嵌入式控制装置用于对所述传感器数据进行特征提取得到特征数据，并根据所述特征数据进行实时分析与预测，得到预测结果并显示。
一种电子设备健康监测预警方法，包括以下步骤：
传感器对电子设备的宿主电子系统电路板进行物理参数监测，获取传感器数据并发送至嵌入式控制装置；所述传感器数据包括电流数据、振动数据、温度数据和电压数据中的至少一种；
所述嵌入式控制装置对所述传感器数据进行特征提取得到特征数据；
所述嵌入式控制装置根据所述特征数据进行实时分析与预测，得到预测结果并显示。
上述电子设备健康监测预警系统和方法，通过传感器对电子设备的宿主电子系统电路板进行物理参数监测，获取传感器数据并发送至嵌入式控制装置，传感器数据包括电流数据、振动数据、温度数据和电压数据中的至少一种。嵌入式控制装置对传感器数据进行特征提取得到特征数据，并根据特征数据进行实时分析与预测，得到预测结果并显示，为用户提供宿主电子系统电路板实时健康监测及实时预测信息。可实时监测电子设备的性能退化过程，预测电子设备性能退化趋势，实现了对电子设备的故障预测与健康管理功能。
附图说明
图1为一实施例中电子设备健康监测预警系统的结构图；
图2为一实施例中嵌入式控制装置的结构图；
图3为一实施例中电子设备健康监测预警方法的流程图；
图4为一实施例中嵌入式控制装置根据特征数据进行实时分析与预测，得到预测结果的流程图。
具体实施方式
在一个实施例中，一种电子设备健康监测预警系统，如图1所示，包括传感器110和嵌入式控制装置120，传感器110设置于电子设备内，嵌入式控制 装置120连接传感器110。
传感器110用于对电子设备的宿主电子系统电路板进行物理参数监测，并将获取得到的传感器数据发送至嵌入式控制装置120。
传感器110的数量和类型均不唯一，可以是一个或多个，也可是一种或多种类型。根据传感器的类型不同，传感器数据的种类也会对应有所不同，传感器数据具体可包括电流数据、振动数据、温度数据和电压数据中的至少一种。可通过针对宿主电子系统电路板进行传感器110的布局设计，以监测所需的物理参数。本实施例中，传感器110包括连接嵌入式控制装置120的电流传感器、振动传感器、温度传感器和电压传感器，分别对宿主电子系统电路板进行监测，对应采集得到的传感器数据包括电流数据、振动数据、温度数据和电压数据。同时对宿主电子系统电路板的四种物理参数进行采集并发送至嵌入式控制装置120进行健康预测，确保预测结果更符合实际情况，提高了预测准确性。可以理解，传感器110的具体类型包括并不限于以上四种，还可包括其他可嵌入的传感器。
嵌入式控制装置120用于对传感器数据进行特征提取得到特征数据，并根据特征数据进行实时分析与预测，得到预测结果并显示。
通过对传感器数据进行特征提取得到特征数据，表征宿主电子系统电路板的物理状态，以便用作后续进行健康预测。嵌入式控制装置120对传感器数据进行特征提取的具体方式并不唯一，可以是对同一时刻对宿主电子系统电路板不同位置采集到的同类型传感器数据进行特征提取，得到特征数据；也可以是对宿主电子系统电路板同一位置采集到的多个同类型传感器数据进行特征提取，得到特征数据。特征数据的类型也不唯一，具体可包括均值或均方差。
以传感器数据包括温度数据、特征数据包括均值为例，可以是通过多个温度传感器对宿主电子系统电路板的不同位置同时进行温度监测，根据同一时刻采集得到的多个温度数据计算得到均值作为特征数据；也可以是通过温度传感器对宿主电子系统电路板的同一位置进行温度监测，获取每个采集周期内的多个温度数据计算得到均值作为特征数据。
嵌入式控制装置120的具体类型并不唯一，具体可以是SoPC嵌入式装置 或SoC嵌入式装置。本实施例中，嵌入式控制装置120为SoPC嵌入式装置，将处理器、存储器、I/O(输入/输出)口等系统设计及其他用户需要的功能模块集成到一个器件上，构建成一个可编程的片上系统。SoPC嵌入式装置具有灵活的设计方式，可裁减、可扩充、可升级，并具备软硬件的系统可编程能力。
在一个实施例中，如图2所示，嵌入式控制装置120包括FPGA(Field-Programmable Gate Array，现场可编程门阵列)逻辑器件121、嵌入式处理器122和显示器123，FPGA逻辑器件121连接传感器110，嵌入式处理器122连接FPGA逻辑器件121和显示器123。
嵌入式处理器122通过FPGA逻辑器件121获取传感器110输出的传感器数据，对传感器数据进行特征提取得到特征数据，并根据特征数据进行实时分析与预测得到预测结果，并将预测结果发送至显示器123进行显示。
此外，嵌入式控制装置120还包括连接嵌入式处理器122的存储器124和应用程序编程接口125。存储器124用于存储传感器数据以及预测结果等数据，应用程序编程接口125用于提供程序访问接口，以便进行应用程序的开发和访问，提高了操作便利性。
具体地，FPGA逻辑器件121可以是具有软核或硬核的FPGA逻辑器件。嵌入式处理器122内置有内核层、服务层和应用层。其中，内核层包括操作系统内核和传感器驱动程序，本实施例中，操作系统内核为面向SoPC的操作系统内核，传感器驱动程序用于结合操作系统内核，针对传感器进行驱动。服务层包括基于数据驱动的预测算法模型和应用编程接口，基于数据驱动的预测算法模型用于根据操作系统内核提供的接口获取传感器数据，针对传感器数据进行特征提取，并对特征数据进行实时分析与预测。应用层包括故障预测与健康管理APP(Application，应用程序)以及访问APP。故障预测与健康管理APP用于将预测结果发送至显示器123进行显示，为用户提供宿主电子系统电路板的实时健康监测及实时预测信息。工作人员可通过访问APP连接服务层所提供的应用编程接口，设计操作系统内核、传感器驱动程序、基于数据驱动的预测模型以及故障预测与健康管理APP的应用程序。
传感器110实时获取传感器数据，对所获得的传感器数据经FPGA逻辑器 件121、内核层中传感器驱动程序与操作系统内核传递后，在服务层中的基于数据驱动的预测模型中进行特征提取，并利用预测算法对特征数据进行实时预测，通过应用层中的故障预测与健康管理APP显示所获得的预测结果，为用户提供宿主电子系统电路板实时健康监测及实时预测信息。
嵌入式控制装置120根据特征数据进行实时分析与预测，得到预测结果的具体方式并不唯一，可以采用任何具有预测功能的算法进行分析和预测。具体地，可采用拓展卡尔曼滤波、无迹卡尔曼滤波和粒子滤波算法进行实时分析与预测，这些算法均是以卡尔曼滤波为基础。
给定N个量测输出数据y1,y2,...,yN，要预测p步之后系统状态xN+p。其中，滤波是指：
Figure PCTCN2016107704-appb-000001
一步预测和两步预测分别为：
Figure PCTCN2016107704-appb-000002
其中，
Figure PCTCN2016107704-appb-000003
为记号方便，简记
Figure PCTCN2016107704-appb-000004
具体而言，对于线性随机系统
Figure PCTCN2016107704-appb-000005
其中，A和C为预设矩阵，wk是均值为0，方差为Q不相关过程噪声，vk是均值为0，方差为R不相关量测噪声，且wk,vk不相关。
本实施例中，嵌入式控制装置120采用卡尔曼预测算法根据特征数据进行实时分析与预测，具体包括以下步骤：
对特征数据进行卡尔曼滤波滤波得到滤波数据。具体为：
Figure PCTCN2016107704-appb-000006
P′k＝APk-1AT+Q,
Kk＝P′kCT(CP′kCT+R)-1,
Figure PCTCN2016107704-appb-000007
Pk＝P′k-KkCP′k.
其中，k＝1,2,...,N，N为特征数据的个数，且
Figure PCTCN2016107704-appb-000008
已知；A和C为预设矩阵，AT,CT分别表示对矩阵A和C进行转置，Q和R分别为不相关过程噪声和不相关 量测噪声的方差，yk表示特征数据，
Figure PCTCN2016107704-appb-000009
表示滤波数据。
根据滤波数据进行卡尔曼预测得到预测结果。具体为：
Figure PCTCN2016107704-appb-000010
其中，
Figure PCTCN2016107704-appb-000011
为滤波数据，
Figure PCTCN2016107704-appb-000012
表示第k步的预测结果；N为特征数据的个数，p为预测步数。
上述电子设备健康监测预警系统，通过传感器110对电子设备的宿主电子系统电路板进行物理参数监测，获取传感器数据并发送至嵌入式控制装置120，传感器数据包括电流数据、振动数据、温度数据和电压数据中的至少一种。嵌入式控制装置120对传感器数据进行特征提取得到特征数据，并根据特征数据进行实时分析与预测，得到预测结果并显示，为用户提供宿主电子系统电路板实时健康监测及实时预测信息。可实时监测电子设备的性能退化过程，预测电子设备性能退化趋势，实现了对电子设备的故障预测与健康管理功能。
在一个实施例中，一种电子设备健康监测预警方法，基于上述电子设备健康监测预警系统实现。如图3所示，该方法包括以下步骤：
步骤S110：传感器对电子设备的宿主电子系统电路板进行物理参数监测，获取传感器数据并发送至嵌入式控制装置。
传感器设置于电子设备内，传感器的数量和类型均不唯一，可以是一个或多个，也可是一种或多种类型。根据传感器的类型不同，传感器数据的种类也会对应有所不同，传感器数据具体可包括电流数据、振动数据、温度数据和电压数据中的至少一种。可通过针对宿主电子系统电路板进行传感器的布局设计，以监测所需的物理参数。本实施例中，传感器包括电流传感器、振动传感器、温度传感器和电压传感器，分别对宿主电子系统电路板进行监测，对应采集得到的传感器数据包括电流数据、振动数据、温度数据和电压数据。同时对宿主电子系统电路板的四种物理参数进行采集并发送至嵌入式控制装置进行健康预测，确保预测结果更符合实际情况，提高了预测准确性。可以理解，传感器的具体类型包括并不限于以上四种，还可包括其他可嵌入的传感器。
步骤S120：嵌入式控制装置对传感器数据进行特征提取得到特征数据。
通过对传感器数据进行特征提取得到特征数据，表征宿主电子系统电路板的物理状态，以便用作后续进行健康预测。对传感器数据进行特征提取的具体方式并不唯一，可以是对同一时刻对宿主电子系统电路板不同位置采集到的同类型传感器数据进行特征提取，得到特征数据；也可以是对宿主电子系统电路板同一位置采集到的多个同类型传感器数据进行特征提取，得到特征数据。特征数据的类型也不唯一，具体可包括均值或均方差。嵌入式控制装置的具体类型并不唯一，具体可以是SoPC嵌入式装置或SoC嵌入式装置。
步骤S130：嵌入式控制装置根据特征数据进行实时分析与预测，得到预测结果并显示。
嵌入式控制装置根据特征数据进行实时分析与预测，得到预测结果的具体方式并不唯一，具体地，可采用拓展卡尔曼滤波、无迹卡尔曼滤波和粒子滤波算法进行实时分析与预测，这些算法均是以卡尔曼滤波为基础。本实施例中，如图4所示，步骤S130中采用卡尔曼预测算法根据特征数据进行实时分析与预测，具体包括步骤S132和步骤S134。
步骤S132：对特征数据进行卡尔曼滤波滤波得到滤波数据。具体为：
Figure PCTCN2016107704-appb-000013
P′k＝APk-1AT+Q,
Kk＝P′kCT(CP′kCT+R)-1,
Figure PCTCN2016107704-appb-000014
Pk＝P′k-KkCP′k.
其中，k＝1,2,...,N，N为特征数据的个数，且
Figure PCTCN2016107704-appb-000015
已知；A和C为预设矩阵，AT,CT分别表示对矩阵A和C进行转置，Q和R分别为不相关过程噪声和不相关量测噪声的方差，yk表示特征数据，
Figure PCTCN2016107704-appb-000016
表示滤波数据。
步骤S134：根据滤波数据进行卡尔曼预测得到预测结果。具体为：
Figure PCTCN2016107704-appb-000017
其中，
Figure PCTCN2016107704-appb-000018
为滤波数据，
Figure PCTCN2016107704-appb-000019
表示第k步的预测结果；N为特征数据的个数，p为预测步数。
上述电子设备健康监测预警方法，通过传感器对电子设备的宿主电子系统电路板进行物理参数监测，获取传感器数据并发送至嵌入式控制装置，传感器数据包括电流数据、振动数据、温度数据和电压数据中的至少一种。嵌入式控制装置对传感器数据进行特征提取得到特征数据，并根据特征数据进行实时分析与预测，得到预测结果并显示，为用户提供宿主电子系统电路板实时健康监测及实时预测信息。可实时监测电子设备的性能退化过程，预测电子设备性能退化趋势，实现了对电子设备的故障预测与健康管理功能。
以上所述实施例的各技术特征可以进行任意的组合，为使描述简洁，未对上述实施例中的各个技术特征所有可能的组合都进行描述，然而，只要这些技术特征的组合不存在矛盾，都应当认为是本说明书记载的范围。
以上所述实施例仅表达了本发明的几种实施方式，其描述较为具体和详细，但并不能因此而理解为对发明专利范围的限制。应当指出的是，对于本领域的普通技术人员来说，在不脱离本发明构思的前提下，还可以做出若干变形和改进，这些都属于本发明的保护范围。因此，本发明专利的保护范围应以所附权利要求为准。

Claims (9)

  1. 一种电子设备健康监测预警系统，其特征在于，包括传感器和嵌入式控制装置，所述传感器设置于电子设备内，所述嵌入式控制装置连接所述传感器，
    所述传感器用于对所述电子设备的宿主电子系统电路板进行物理参数监测，并将获取得到的传感器数据发送至所述嵌入式控制装置；所述传感器数据包括电流数据、振动数据、温度数据和电压数据中的至少一种；
    所述嵌入式控制装置用于对所述传感器数据进行特征提取得到特征数据，并根据所述特征数据进行实时分析与预测，得到预测结果并显示。
  2. 根据权利要求1所述的电子设备健康监测预警系统，其特征在于，所述传感器包括连接所述嵌入式控制装置的电流传感器、振动传感器、温度传感器和电压传感器。
  3. 根据权利要求1所述的电子设备健康监测预警系统，其特征在于，所述嵌入式控制装置包括FPGA逻辑器件、嵌入式处理器和显示器，所述FPGA逻辑器件连接所述传感器，所述嵌入式处理器连接所述FPGA逻辑器件和所述显示器，
    所述嵌入式处理器通过所述FPGA逻辑器件获取所述传感器输出的传感器数据，对所述传感器数据进行特征提取得到所述特征数据，并根据所述特征数据进行实时分析与预测得到所述预测结果，并将所述预测结果发送至所述显示器进行显示。
  4. 根据权利要求3所述的电子设备健康监测预警系统，其特征在于，所述嵌入式控制装置还包括连接所述嵌入式处理器的存储器和应用程序编程接口。
  5. 根据权利要求1-4任意一项所述的电子设备健康监测预警系统，其特征在于，所述嵌入式控制装置为SoPC嵌入式装置或SoC嵌入式装置。
  6. 一种电子设备健康监测预警方法，其特征在于，包括以下步骤：
    传感器对电子设备的宿主电子系统电路板进行物理参数监测，获取传感器数据并发送至嵌入式控制装置；所述传感器数据包括电流数据、振动数据、温 度数据和电压数据中的至少一种；
    所述嵌入式控制装置对所述传感器数据进行特征提取得到特征数据；
    所述嵌入式控制装置根据所述特征数据进行实时分析与预测，得到预测结果并显示。
  7. 根据权利要求6所述的电子设备健康监测预警方法，其特征在于，所述传感器包括电流传感器、振动传感器、温度传感器和电压传感器。
  8. 根据权利要求6所述的电子设备健康监测预警方法，其特征在于，所述特征数据包括均值或均方差值。
  9. 根据权利要求6所述的电子设备健康监测预警方法，其特征在于，所述嵌入式控制装置根据所述特征数据进行实时分析与预测，得到预测结果，包括：
    对所述特征数据进行卡尔曼滤波滤波得到滤波数据，具体为：
    Figure PCTCN2016107704-appb-100001
    Pk'＝APk-1AT+Q,
    Kk＝Pk'CT(CPk'CT+R)-1,
    Figure PCTCN2016107704-appb-100002
    Pk＝Pk'-KkCPk'.
    其中，k＝1,2,...,N，N为特征数据的个数，且P0,
    Figure PCTCN2016107704-appb-100003
    已知；A和C为预设矩阵，AT,CT分别表示对矩阵A和C进行转置，Q和R分别为不相关过程噪声和不相关量测噪声的方差，yk表示特征数据，
    Figure PCTCN2016107704-appb-100004
    表示滤波数据；
    根据所述滤波数据进行卡尔曼预测得到预测结果，具体为：
    Figure PCTCN2016107704-appb-100005
    其中，
    Figure PCTCN2016107704-appb-100006
    为滤波数据，
    Figure PCTCN2016107704-appb-100007
    表示第k步的预测结果；N为特征数据的个数，p为预测步数。
PCT/CN2016/107704 2016-08-12 2016-11-29 电子设备健康监测预警系统和方法 WO2018028086A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US15/556,702 US10458823B2 (en) 2016-08-12 2016-11-29 System and method for health monitoring and early warning for electronic device

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610665494.4A CN106323373A (zh) 2016-08-12 2016-08-12 电子设备健康监测预警系统和方法
CN201610665494.4 2016-08-12

Publications (1)

Publication Number Publication Date
WO2018028086A1 true WO2018028086A1 (zh) 2018-02-15

Family

ID=57739355

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/107704 WO2018028086A1 (zh) 2016-08-12 2016-11-29 电子设备健康监测预警系统和方法

Country Status (3)

Country Link
US (1) US10458823B2 (zh)
CN (1) CN106323373A (zh)
WO (1) WO2018028086A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115314401A (zh) * 2022-06-30 2022-11-08 中铁第四勘察设计院集团有限公司 接触网状态监测方法、装置、电子设备及存储介质
CN115389915A (zh) * 2022-10-27 2022-11-25 北京东远润兴科技有限公司 电路健康监测管理系统、监测方法及存储介质

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107329049B (zh) * 2017-08-21 2019-07-02 集美大学 一种基于卡尔曼滤波器的输电线路早期故障检测方法
CN107752171A (zh) * 2017-10-18 2018-03-06 河南科技大学第附属医院 一种重症监护儿童病号服
EP3502720B1 (en) * 2017-12-19 2022-11-09 Nxp B.V. An apparatus for prediction of failure of a functional circuit
CN109298265A (zh) * 2018-10-23 2019-02-01 东莞理工学院 电子设备故障检查装置
CN111000684B (zh) * 2019-09-27 2022-04-01 张兴利 一种基于多传感器融合的纸尿裤外置监测系统及监测方法
CN111854852A (zh) * 2020-08-25 2020-10-30 武汉迪航科技有限公司 一种用于机电设备安装的检测系统
CN112858872A (zh) * 2020-12-31 2021-05-28 华中光电技术研究所（中国船舶重工集团公司第七一七研究所） 电路板健康管理电路、装置、控制方法及电路板健康管理器
CN113934982B (zh) * 2021-10-18 2024-04-19 河北工业大学 基于振-电信号融合的断路器操作机构机械寿命预测方法
CN114238033B (zh) * 2021-12-28 2022-09-13 湖南云箭智能科技有限公司 板卡运行状态预警方法、装置、设备及可读存储介质
CN114613110B (zh) * 2022-01-29 2023-08-18 上海至冕伟业科技有限公司 基于消防物联网的消防水系统故障检测预警系统及方法
US20230418352A1 (en) * 2022-06-24 2023-12-28 Microsoft Technology Licensing, Llc Devices, systems, and methods for monitoring an electronic device in transit

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2008039492A (ja) * 2006-08-02 2008-02-21 Univ Nagoya 電気機器稼働状態推定システム
CN101825894A (zh) * 2010-04-30 2010-09-08 北京航空航天大学 一种六氟化硫高压断路器状态智能监测与健康管理系统
CN101963937A (zh) * 2010-09-29 2011-02-02 南京航空航天大学 飞控计算机系统剩余寿命预测方法
CN102156245A (zh) * 2011-03-11 2011-08-17 太原理工大学 一种矿用高压电缆在线故障诊断及预警方法
CN102316496A (zh) * 2011-09-07 2012-01-11 上海交通大学 无线传感器网络中基于卡尔曼滤波的数据融合方法
CN103019916A (zh) * 2012-12-14 2013-04-03 大连博控科技股份有限公司 便携电子设备温度监测系统
CN103176486A (zh) * 2011-12-26 2013-06-26 联想(北京)有限公司 一种应用于电子设备中的除湿方法及电子设备
CN204177480U (zh) * 2014-10-29 2015-02-25 天津滨海新区三格电子科技有限公司 一种基于Cortex-M4的多路振动信号采集系统

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10502594B2 (en) * 2013-04-29 2019-12-10 Emerson Electric (Us) Holding Corporation (Chile) Limitada Dynamic transducer with digital output and method for use
CN104750068B (zh) * 2015-02-13 2018-08-21 湖北锐世数字医学影像科技有限公司 一种多节点传感器网络的数据传输及控制装置

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2008039492A (ja) * 2006-08-02 2008-02-21 Univ Nagoya 電気機器稼働状態推定システム
CN101825894A (zh) * 2010-04-30 2010-09-08 北京航空航天大学 一种六氟化硫高压断路器状态智能监测与健康管理系统
CN101963937A (zh) * 2010-09-29 2011-02-02 南京航空航天大学 飞控计算机系统剩余寿命预测方法
CN102156245A (zh) * 2011-03-11 2011-08-17 太原理工大学 一种矿用高压电缆在线故障诊断及预警方法
CN102316496A (zh) * 2011-09-07 2012-01-11 上海交通大学 无线传感器网络中基于卡尔曼滤波的数据融合方法
CN103176486A (zh) * 2011-12-26 2013-06-26 联想(北京)有限公司 一种应用于电子设备中的除湿方法及电子设备
CN103019916A (zh) * 2012-12-14 2013-04-03 大连博控科技股份有限公司 便携电子设备温度监测系统
CN204177480U (zh) * 2014-10-29 2015-02-25 天津滨海新区三格电子科技有限公司 一种基于Cortex-M4的多路振动信号采集系统

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115314401A (zh) * 2022-06-30 2022-11-08 中铁第四勘察设计院集团有限公司 接触网状态监测方法、装置、电子设备及存储介质
CN115314401B (zh) * 2022-06-30 2024-03-15 中铁第四勘察设计院集团有限公司 接触网状态监测方法、装置、电子设备及存储介质
CN115389915A (zh) * 2022-10-27 2022-11-25 北京东远润兴科技有限公司 电路健康监测管理系统、监测方法及存储介质

Also Published As

Publication number Publication date
CN106323373A (zh) 2017-01-11
US20190154475A1 (en) 2019-05-23
US10458823B2 (en) 2019-10-29

Similar Documents

Publication Publication Date Title
WO2018028086A1 (zh) 电子设备健康监测预警系统和方法
CN108604360B (zh) 设施异常监测方法及其系统
JP6813522B2 (ja) 健全性監視及び障害特徴識別の方法及びシステム
US7653510B2 (en) Load calculating device and load calculating method
US10670648B2 (en) Systems and methods for determining whether a circuit is operating properly
JP6361175B2 (ja) 異常診断装置及びプログラム
WO2014103134A1 (en) Predicting a time of failure of a device
EP3933533B1 (en) Apparatus for diagnosing in vitro instruments
CN108460397B (zh) 设备故障类型的分析方法、装置、储存介质和电子设备
WO2018014502A1 (zh) 电梯的预诊断方法及装置
JP6880560B2 (ja) 故障予測装置、故障予測方法及び故障予測プログラム
JP2013196698A (ja) システム監視
Hu et al. Mahalanobis distance based approach for anomaly detection of analog filters using frequency features and Parzen window density estimation
KR102182226B1 (ko) 고장 검출 진단 시스템 및 진단방법
JP6274932B2 (ja) 予測システム、監視システム、運転支援システム、ガスタービン設備及び予測方法
KR20180126311A (ko) 대상 시스템의 상태 진단 및 원인 분석 시스템 및 방법
US11113166B2 (en) Monitoring system and method with baseboard management controller
WO2020044898A1 (ja) 機器状態監視装置及びプログラム
CN111290371A (zh) 物联网设备远程诊断的方法及装置、电子设备
US20220222402A1 (en) Information processing device, information processing method, and information processing program
KR20230125116A (ko) 머신러닝 클러스터링 알고리즘과 통계적 기법을 이용해 고장 데이터가 부재한 시스템의 결함 탐지 및 고장 진단 예측 방법
CN114297963A (zh) 基于K-Means小波神经网络的软电路故障诊断方法
CN114021744A (zh) 设备的剩余使用寿命的确定方法、装置和电子设备
EP3557421B1 (en) System and method for automatic generation of a diagnostic model
US11099525B2 (en) Control of safety input/output by non-safety system during safe times

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16912540

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 16912540

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 16912540

Country of ref document: EP

Kind code of ref document: A1