An automated operation and maintenance monitoring method and system for Internet of Things monitoring devices

Abstract

The present invention discloses an automated operation and maintenance monitoring method and system for Internet of Things monitoring devices, belonging to the technical field of operation and maintenance monitoring systems. Among them, the method part includes the following steps: S1, build an Internet of Things integrated operation and maintenance platform and connect multiple devices that need to be monitored for operation and maintenance; S2, perform anomaly identification on the reporting frequency and real-time performance of the devices to determine the operating conditions of the devices; S3, perform anomaly identification on the newly reported data of the devices. The system part includes an Internet of Things integrated operation and maintenance platform and multiple devices connected thereto. By building an integrated operation and maintenance platform for Internet of Things devices, monitoring the reporting frequency and real-time performance of device data, reconfirming the operating conditions of the devices, and at the same time building a numerical monitoring integrated model to identify outliers based on historical data, providing integrated device anomaly warning information, and replacing manual operation and maintenance with automated operation and maintenance, the problems of low efficiency and high cost caused by manual operation and maintenance are avoided.

Description

Technical Field

[0001] The present invention belongs to the technical field of operation and maintenance monitoring systems, and specifically relates to an automated operation and maintenance monitoring method and system for Internet of Things monitoring equipment. Background Art

[0002] With the construction of water conservancy informatization, the density and number of Internet of Things monitoring equipment are increasing continuously, and comprehensive, real-time, accurate and reliable data are further constructed, providing support for the construction of a smart water conservancy system with forecasting, early warning, rehearsal and emergency plan (referred to as "four predictions") functions, effectively improving the foresight of early warning and emergency response capabilities.

[0003] As the number of monitoring devices continues to increase, the difficulty of automated monitoring of equipment conditions has also increased. At the same time, the surge in monitoring data volume has also posed challenges to ensuring data quality, making it more difficult to identify outliers within large amounts of data. These issues not only affect the real-time and accuracy of monitoring data, but may also negatively impact early warning systems and emergency response mechanisms. Currently, operations and maintenance monitoring primarily relies on manual observation, a practice that has yet to establish an efficient operations and maintenance monitoring system. This is not only inefficient but also increases costs. Summary of the Invention

[0004] The present invention proposes an automated operation and maintenance monitoring method and system for Internet of Things monitoring equipment. By building an integrated operation and maintenance platform for Internet of Things equipment, the frequency and real-time reporting of equipment data are monitored, and the equipment working condition is reconfirmed. At the same time, an integrated numerical monitoring model is built to identify abnormal values based on historical data, and provide integrated equipment abnormality alarm information. Automated operation and maintenance replaces manual operation and maintenance to solve the problems of low efficiency and high cost caused by manual operation and maintenance.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is:

[0006] An automated operation and maintenance monitoring method for an Internet of Things monitoring device comprises the following steps:

[0007] S1. Build an integrated IoT operation and maintenance platform and connect multiple devices that require operation and maintenance monitoring;

[0008] S2. Use the IoT integrated operation and maintenance platform to identify abnormalities in the equipment's reporting frequency and real-time performance, and determine the equipment's operating condition.

[0009] S3. The IoT integrated operation and maintenance platform uses historical data as sample data to construct a univariate polynomial regression model and a multivariate linear regression model. It also uses the univariate polynomial regression model and the multivariate linear regression model to identify anomalies in the newly reported data of the equipment. If any data anomalies are found, the IoT integrated operation and maintenance platform will execute an alarm action.

[0010] Further, in step S2, the method for identifying abnormal reporting frequency includes the following steps:

[0011] S201. The operation and maintenance personnel configure the abnormal frequency identification rules, including the reporting frequency and time threshold;

[0012] S202. The device connects to the Internet of Things integrated operation and maintenance platform through a long connection;

[0013] S203. The Internet of Things integrated operation and maintenance platform detects the list of devices that have not reported data exceeding the time threshold through a scheduled task;

[0014] S204. Determine whether the long connection status of each device in the device list is alive;

[0015] S205. If the long connection status of the device is no longer alive, prompt for device offline abnormality. If the long connection status of the device is alive, continue;

[0016] S206. The Internet of Things integrated operation and maintenance platform actively sends a data query request to the devices with a surviving long connection status in the previous step and retries at most three times;

[0017] S207. If the Internet of Things integrated operation and maintenance platform does not receive a device data response three times, it is determined that the device is faulty and an abnormality prompt is given; if a device data response is received, it is not regarded as an abnormality and a network fluctuation prompt is given.

[0018] Further, in step S2, the method for identifying real-time abnormality includes the following steps:

[0019] S211. The operation and maintenance personnel configure the device reporting frequency rule;

[0020] S212. The device connects to the Internet of Things integrated operation and maintenance platform through a long connection;

[0021] S213. The Internet of Things integrated operation and maintenance platform detects the list of devices that have not reported data exceeding the time threshold through a scheduled task;

[0022] S214. Determine whether the long connection status of each device in the device list is alive;

[0023] S215. If the long connection status of the device is no longer alive, prompt for device offline abnormality; if the long connection status of the device is alive, continue;

[0024] S216. The Internet of Things integrated operation and maintenance platform actively sends a device local clock query request to the devices with a surviving long connection status in the previous step;

[0025] S217. After the Internet of Things integrated operation and maintenance platform receives the device response, determine whether the device clock matches the server clock;

[0026] S218. If the clocks match, it indicates that the device fails to report data on time, and a real-time fault of the device is prompted; if the clocks do not match, the Internet of Things integrated operation and maintenance platform actively issues an instruction to modify the local clock of the device and prompts a clock error of the device.

[0027] Further, in step S3, the method for identifying data anomalies includes the following steps:

[0028] S31. The Internet of Things integrated operation and maintenance platform stores the historical monitoring data of the device through the time series library;

[0029] S32. Train according to the historical monitoring data to construct a univariate polynomial regression model and a multiple linear regression model;

[0030] S33. After the Internet of Things integrated operation and maintenance platform receives the newly reported data of the device, substitute them into the two models for calculation respectively;

[0031] S34. If the newly reported data of the device is greater than the upper limit value of one of the regression models, it is determined that the data is abnormal and marked as abnormal data; otherwise, the data is normal and no processing is performed.

[0032] Further, in step S32, the construction method of the univariate polynomial regression model includes the following steps:

[0033] (1) Let the sample set T = [(x1, y1), (x2, y2), (x3, y3)...(x n , y n )], f(x i , w) = w0 + w1x1 + w2x2 + w3x3 + ··· + w im x im = ∑w j + x ji , where x is a univariate input, and w0, w1, w2,..., w M are M + 1 parameters;

[0034] (2) Use the least squares method to calculate the loss function, substitute the model and the training data, and obtain:

[0035]

[0036] (3) Take the partial derivative of w j and set it to 0 to obtain:

[0037]

[0038] (4) Solve the following system of linear equations and fit the polynomial coefficients w0, w1, w2,..., w M :

[0039]

[0040] (5) The calculated univariate polynomial regression model is:

[0041] and

[0042] Further, in step S32, the method for constructing the multiple linear regression model includes the following steps:

[0043] (1) Let the sample set: T = [(x 11 , x 21 ,..., x n1 , y1), (x 21 , x 22 ,..., x n2 , y2), (x 31 , x 32 ,..., x n3 , y3),...(x n1 , x n2 ,..., x nn , y n )]; where, y i = a1×x 1i + a2×x 2i +... + a n ×x ni + ε; denoted as the matrix expression:

[0044]

[0045] (2) Use the least squares method for fitting and solving to obtain the multiple linear regression model as:

[0046] y = xa + ε.

[0047] An Internet of Things monitoring device automated operation and maintenance monitoring system includes an Internet of Things integrated operation and maintenance platform and multiple devices. The Internet of Things integrated operation and maintenance platform is connected to the multiple devices and performs operation and maintenance monitoring on the multiple devices through the Internet of Things monitoring device automated operation and maintenance monitoring method.

[0048] Further, the Internet of Things integrated operation and maintenance platform is provided with a device condition determination module and an abnormal data identification module;

[0049] The device condition determination module is used to identify abnormal reporting frequencies and real-time anomalies of the devices;

[0050] The abnormal data identification module is used to identify data anomalies of the devices.

[0051] Further, the abnormal data recognition module performs abnormal recognition on the newly reported data of the device by constructing a univariate polynomial regression model and a multiple linear regression model.

[0052] Due to the adoption of the above technical solutions, the present invention has the following beneficial effects:

[0053] 1. The present invention provides an integrated automatic operation and maintenance monitoring solution, covering device operating conditions and data anomaly monitoring, avoiding abnormal underreporting and manual monitoring, improving operation and maintenance efficiency, and saving labor costs.

[0054] 2. The present invention can customize device operating condition determination rules for different devices, improving flexibility and scalability.

[0055] 3. The present invention trains a regression model based on historical monitoring data, performs mathematical statistical analysis through data change rules, and then determines abnormal values, with higher accuracy.

[0056] 4. The present invention provides an integrated operation and maintenance monitoring platform, facilitating the centralized presentation of various abnormal problems, and helping administrators and operation and maintenance personnel quickly discover and solve problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 is a flowchart of abnormal reporting frequency recognition in the automatic operation and maintenance monitoring method for Internet of Things monitoring devices proposed by the present invention;

[0058] Figure 2 is a flowchart of real-time anomaly recognition in the automatic operation and maintenance monitoring method for Internet of Things monitoring devices proposed by the present invention;

[0059] Figure 3 is a flowchart of data anomaly recognition in the automatic operation and maintenance monitoring method for Internet of Things monitoring devices proposed by the present invention; DETAILED DESCRIPTION OF THE EMBODIMENTS

[0060] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0061] Embodiment 1

[0062] The present invention proposes an automatic operation and maintenance monitoring method for Internet of Things monitoring devices, which includes steps S1, S2, and S3.

[0063] S1. Build an integrated operation and maintenance platform for the Internet of Things and connect multiple devices that need to be monitored for operation and maintenance.

[0064] S2. Identify anomalies in the reporting frequency and real-time performance of devices through the Internet of Things integrated operation and maintenance platform, and determine the operating conditions of the devices.

[0065] Among them, the method for identifying anomalies in the reporting frequency includes steps S201 - S207.

[0066] S201. The operation and maintenance personnel configure the rules for identifying anomalies in the reporting frequency, including the reporting frequency and time threshold. The reporting frequency is, for example, once per hour or once every five minutes.

[0067] S202. The device connects to the Internet of Things integrated operation and maintenance platform through a long connection.

[0068] S203. The Internet of Things integrated operation and maintenance platform detects the list of devices that have not reported data beyond the time threshold through a scheduled task.

[0069] S204. Determine whether the long connection status of each device in the device list is alive respectively.

[0070] S205. If the long connection status of the device is no longer alive, prompt for device offline anomaly; if the long connection status of the device is alive, continue.

[0071] S206. The Internet of Things integrated operation and maintenance platform actively sends a data query request to the devices with a live long connection status in the previous step and retries up to three times.

[0072] S207. If the Internet of Things integrated operation and maintenance platform does not receive a device data response three times, it is determined that the device has a fault and an anomaly prompt is given; if a device data response is received, it is not regarded as an anomaly and a network fluctuation prompt is given.

[0073] That is to say, the Internet of Things integrated operation and maintenance platform is responsible for storing the data reporting records of each device. After exceeding the required reporting frequency and time threshold, the platform actively sends a data query request to the device and retries three times. If a device data response can be received, it indicates that the data reporting omission may be caused by network fluctuations and is not determined as an anomaly. If no response is received three times, it is determined that the device is abnormal.

[0074] The method for identifying anomalies in real-time performance includes steps S211 - S218.

[0075] S211. The operation and maintenance personnel configure the device reporting frequency rules.

[0076] S212. The device connects to the Internet of Things integrated operation and maintenance platform through a long connection.

[0077] S213. The Internet of Things integrated operation and maintenance platform detects the list of devices that have not reported data beyond the time threshold through a scheduled task.

[0078] S214. Determine whether the long connection status of each device in the device list is alive respectively.

[0079] S215. If the long connection status of the device is no longer alive, prompt an abnormal device offline; if the long connection status of the device is alive, continue to execute.

[0080] S216. The Internet of Things integrated operation and maintenance platform actively sends a device local clock query request to the device with a long connection status alive in the previous step.

[0081] S217. After the Internet of Things integrated operation and maintenance platform receives the device response, determine whether the device clock matches the server clock.

[0082] S218. If the clocks match, it means that the device has not reported data on time, and prompt a device real-time fault; if the clocks do not match, the Internet of Things integrated operation and maintenance platform actively sends an instruction to modify the device local clock and prompts a device clock error.

[0083] That is to say, the Internet of Things integrated operation and maintenance platform determines whether the actual received device data reporting time is consistent with the required time according to the reporting frequency configured by the device and the platform server time. If not, the platform actively sends a local clock query request to the device to determine whether the real-time problem is caused by inconsistent clocks. If the clocks are inconsistent, the platform actively sends a time calibration request to the device. In addition, the platform records real-time exceptions caused by non-clock problems.

[0084] The present invention can customize device condition determination rules for different devices, improving flexibility and scalability.

[0085] S3. The Internet of Things integrated operation and maintenance platform uses historical data as sample data to construct a univariate polynomial regression model and a multiple linear regression model, and performs anomaly identification on newly reported data of the device through the univariate polynomial regression model and the multiple linear regression model. If data anomalies are found, the Internet of Things integrated operation and maintenance platform performs an alarm action.

[0086] Among them, the method for data anomaly identification includes steps S31 - S34.

[0087] S31. The Internet of Things integrated operation and maintenance platform stores device historical monitoring data through a time series library.

[0088] S32. Train according to the historical monitoring data to construct a univariate polynomial regression model and a multiple linear regression model.

[0089] Specifically, the construction method of the univariate polynomial regression model includes the following steps.

[0090] (1) Let the sample set T = [(x1, y1), (x2, y2), (x3, y3)...(xn ,y n )],f(x i ,w)=w0+w1x1+w2x2+w3x3+···+w im x im =∑w j +x ji , where x is a single variable input, w0,w1,w2,...,w M There are M+1 parameters.

[0091] (2) Use the least squares method to calculate the loss function, substitute the model and training data, and get:

[0092]

[0093] (3) For w j Taking the partial derivative and setting it to 0, we get:

[0094]

[0095] (4) Solve the following linear equations and fit the polynomial coefficients w0, w1, w2, ..., w M .

[0096]

[0097] (5) The univariate polynomial regression model is calculated as:

[0098]

[0099] The method for constructing a multiple linear regression model includes the following steps.

[0100] (1) Let the sample set: T = [(x 11 ,x 21 ,...,x n1 ,y1),(x 21 ,x 22 ,...,x n2 ,y2),(x 31 ,x 32 ,...,x n3 ,y3),...(x n1 ,x n2 ,...,x nn ,y n )]; where y i =a1×x 1i +a2×x 2i +...+a n ×x ni +ε; expressed as a matrix expression:

[0101]

[0102] (2) The least squares method is used for fitting and solving to obtain the multiple linear regression model as follows:

[0103] y = xa + ε.

[0104] The present invention trains a regression model based on historical monitoring data, conducts mathematical statistical analysis through the data change law, and then determines outliers, with higher accuracy.

[0105] S33. After the Internet of Things integrated operation and maintenance platform receives the newly reported data of the device, it substitutes them into the two models for calculation respectively.

[0106] S34. If the newly reported data of the device is greater than the upper limit value of one of the regression models, it is determined that the data is abnormal and marked as abnormal data; otherwise, the data is normal and no processing is performed.

[0107] The present invention builds an Internet of Things device integrated operation and maintenance platform to monitor the data reporting frequency and real-time performance of the device, reconfirm the device working conditions, and at the same time builds a numerical monitoring integrated model to identify outliers based on historical data and provide integrated device anomaly warning information. The present invention covers the monitoring of device working conditions and data anomalies, avoids abnormal missed reports and manual monitoring, improves the operation and maintenance efficiency, and saves labor costs. The present invention provides an integrated operation and maintenance monitoring platform, which is convenient for various abnormal problems to be presented centrally, and helps administrators and operation and maintenance personnel quickly discover and solve problems.

[0108] Embodiment 2

[0109] The present invention also proposes an Internet of Things monitoring device automated operation and maintenance monitoring system, which includes an Internet of Things integrated operation and maintenance platform and multiple devices. The Internet of Things integrated operation and maintenance platform is connected to the multiple devices and performs operation and maintenance monitoring on the multiple devices through the Internet of Things monitoring device automated operation and maintenance monitoring method.

[0110] The Internet of Things integrated operation and maintenance platform is provided with a device working condition determination module and an abnormal data identification module. The device working condition determination module is used for identifying abnormal reporting frequency and real-time performance anomalies of the device. The abnormal data identification module is used for identifying data anomalies of the device. The abnormal data identification module identifies the newly reported data of the device as abnormal by constructing a univariate polynomial regression model and a multiple linear regression model.

[0111] The present invention utilizes the historical monitoring data of the device to train a univariate polynomial regression model and a multiple linear regression model, judges outliers for the device-reported data, and improves accuracy. At the same time, it monitors the reporting frequency and real-time performance of the device, realizes flexible monitoring customization through configuration, and adds a secondary verification actively triggered by the platform to improve the robustness of anomaly monitoring. Overall, it realizes the integrated automatic operation and maintenance monitoring of water conservancy IoT monitoring devices, improves operation and maintenance efficiency, and reduces labor costs.

[0112] The above description is a detailed description of the preferred and feasible embodiments of the present invention, but the embodiments are not intended to limit the patent application scope of the present invention. Any equivalent changes or modifications made under the technical spirit prompted by the present invention shall fall within the patent scope covered by the present invention.

Claims

1. An automated operation and maintenance monitoring method for Internet of Things monitoring devices, characterized in that, It includes the following steps: S1. Build an Internet of Things integrated operation and maintenance platform and connect multiple devices that need to be monitored for operation and maintenance; S2. Through the Internet of Things integrated operation and maintenance platform, perform anomaly identification on the reporting frequency and real-time performance of the devices to determine the operating conditions of the devices; S3. The Internet of Things integrated operation and maintenance platform uses historical data as sample data to construct a univariate polynomial regression model and a multiple linear regression model, and through the univariate polynomial regression model and the multiple linear regression model, perform anomaly identification on the newly reported data of the devices. If data anomalies are found, the Internet of Things integrated operation and maintenance platform executes an alarm action; Among them, the method for data anomaly identification includes the following steps: S31. The Internet of Things integrated operation and maintenance platform stores the historical monitoring data of the devices through a time series library; S32. Train according to the historical monitoring data to construct a univariate polynomial regression model and a multiple linear regression model; S33. After the Internet of Things integrated operation and maintenance platform receives the newly reported data of the devices, substitute them into the two models for calculation respectively; S34. If the newly reported data of the device is greater than the upper limit value of one of the regression models, it is determined that the data is abnormal and marked as abnormal data; otherwise, the data is normal and no processing is performed; Specifically, in step S32, the construction method of the univariate polynomial regression model includes the following steps: (1) Let the sample set T = [(x1, y1), (x2, y2), (x3, y3)...(x n , y r )], where x is a single-variable input, and w0, w1, w2, ..., w M are M + 1 parameters; (2) Use the least squares method to calculate the loss function, substitute the model and the training data, and obtain: (3) Take the partial derivative with respect to w j Take the partial derivative and set it to 0, we get: (4) Solve the following system of linear equations and fit the polynomial coefficients w0, w1, w2,..., w M : (5) Calculate to obtain the univariate polynomial regression model as: and The construction method of the multiple linear regression model includes the following steps: (1) Let the sample set be: T = [(x 11 , x 21 ,..., x n1 , y1), (x 21 , x 22 ,..., x n2 , y2), (x 31 , x 32 ,..., x n3 , y3),... (x n1 , x n2 ,..., x nn , y n )]; where y i = a1 × x 1i + a2 × x 2i +... + a n × x ni + ε; Denote it as the matrix expression: (2) Adopt the least squares method for fitting and solution to obtain the multiple linear regression model as: y = xa + ε.

2. The automated operation and maintenance monitoring method for an Internet of Things monitoring device according to claim 1, characterized in that: In step S2, the method for anomaly identification of the reporting frequency includes the following steps: S201. The operation and maintenance personnel configure the reporting frequency anomaly identification rules, including the reporting frequency and the time threshold; S202. The device connects to the Internet of Things integrated operation and maintenance platform through a long connection; S203. The Internet of Things integrated operation and maintenance platform detects the list of devices that have not reported data exceeding the time threshold through a scheduled task; S204. Respectively judge whether the long connection status of each device in the device list is alive; S205. If the long connection status of the device is no longer alive, prompt for device offline anomaly. If the long connection status of the device is alive, continue to execute; S206. The Internet of Things integrated operation and maintenance platform actively sends a data query request to the device with the long connection status alive in the previous step and retries at most three times; S207. If the Internet of Things integrated operation and maintenance platform does not receive a device data response three times, it is determined that the device has a fault and an anomaly prompt is given; if a device data response is received, it is not regarded as an anomaly and a network fluctuation prompt is given.

3. The automated operation and maintenance monitoring method for an Internet of Things monitoring device according to claim 2, characterized in that: In step S2, the method for anomaly identification of real-time performance includes the following steps: S211. The operation and maintenance personnel configure the device reporting frequency rules; S212. The device connects to the Internet of Things integrated operation and maintenance platform through a long connection; S213. The Internet of Things integrated operation and maintenance platform detects the list of devices that have not reported data exceeding the time threshold through a scheduled task; S214. Respectively judge whether the long connection status of each device in the device list is alive; S215. If the long connection status of the device is no longer alive, prompt for an abnormal device offline; if the long connection status of the device is alive, continue to execute; S216. The Internet of Things integrated operation and maintenance platform actively sends a device local clock query request to the device with a long connection status alive in the previous step; S217. After the Internet of Things integrated operation and maintenance platform receives the device response, determine whether the device clock matches the server clock; S218. If the clocks match, it indicates that the device has not reported data on time, and prompt for a device real-time fault; if the clocks do not match, the Internet of Things integrated operation and maintenance platform actively sends an instruction to modify the device local clock and prompt for a device clock error.

4. An automated operation and maintenance monitoring system for Internet of Things monitoring devices, characterized in that: It includes an Internet of Things integrated operation and maintenance platform and multiple devices. The Internet of Things integrated operation and maintenance platform is connected to multiple devices, and performs operation and maintenance monitoring on multiple devices through the Internet of Things monitoring device automated operation and maintenance monitoring method described in any one of claims 1-3.

5. The automated operation and maintenance monitoring system for an Internet of Things monitoring device according to claim 4, wherein: The Internet of Things integrated operation and maintenance platform is provided with a device working condition determination module and an abnormal data identification module; The device working condition determination module is used to identify abnormal reporting frequencies and real-time abnormalities of the device; The abnormal data identification module is used to identify data abnormalities of the device.

6. The automated operation and maintenance monitoring system for an Internet of Things monitoring device according to claim 5, characterized in that: The abnormal data identification module identifies abnormalities in the newly reported data of the device by constructing a single-variable polynomial regression model and a multiple linear regression model.

Images