A device data analysis and management system and method applying a temperature sensor

Abstract

The present invention discloses a device data analysis management system and method using a temperature sensor, and relates to the technical field of device data analysis. The present invention generates device fault feature evaluation data after analyzing device operation data; calculates device fault feature values ​​according to feature data when the device fails in history; analyzes the abnormal correlation between the fault feature values ​​and the corresponding first environmental data; constructs a fault feature fluctuation model according to the abnormal correlation combination; establishes an environmental correlation fitting model according to the environmental data association set, and constructs an environmental confidence fitting model through the environmental correlation fitting models of two or more other environmental data; analyzes the operation of the device using the feature fluctuation model; generates a fault risk warning according to the operation of the device; and judges the true situation of the fault risk warning according to the environmental confidence fitting model. It can reduce the monitoring cost of users and improve the comprehensiveness and reliability of the monitoring system.

Description

Technical Field

[0001] The present invention relates to the technical field of device data analysis, and specifically to a device data analysis management system and method applying a temperature sensor. Background Art

[0002] A temperature sensor can monitor the temperature change of a device in real time. By monitoring the temperature change of the device in real time, the temperature sensor can help detect potential problems such as overheating of the device in a timely manner and avoid damage to the device due to overheating. Correspondingly, when a device fails, a large amount of heat energy is often generated, and there is a hidden correlation between the data monitored by the temperature sensor and the operating state of the device.

[0003] In the prior art, using a temperature sensor alone to analyze the data of a device is easily affected by the environment, resulting in inaccurate judgment; if the sensor fails, the device state cannot be accurately monitored; a single data source cannot comprehensively reflect the complex operating state of the device, lacking comprehensiveness and reliability.

[0004] Therefore, the present invention discloses a device data analysis management system and method applying a temperature sensor to solve the above problems. Summary of the Invention

[0005] The purpose of the present invention is to provide a device data analysis management system and method applying a temperature sensor to solve the problems raised in the prior art.

[0006] To achieve the above purpose, the present invention provides the following technical solution: A device data analysis management method applying a temperature sensor, the method comprising the following steps:

[0007] S1: Install sensors on the device, use the sensors to collect environmental data of the device; extract the device operation data when a failure occurs during the historical operation of the device, and generate device failure characteristic evaluation data after analyzing the device operation data;

[0008] S2: Calculate the device failure characteristic value according to the characteristic data when the device fails in history; analyze the abnormal correlation degree between the failure characteristic value and the corresponding first environmental data; generate an abnormal correlation combination according to the abnormal correlation degree, and construct a failure characteristic fluctuation model according to the abnormal correlation combination;

[0009] S3: Extract other environmental data except the first environmental data when the device fails in history, form an environmental data association set according to the first environmental data and other environmental data, establish an environmental association fitting model according to the environmental data association set, and construct an environmental confidence fitting model through the environmental association fitting models of two or more other environmental data;

[0010] S4: Collect real-time first environmental data, analyze the operation of the device using the feature fluctuation model; generate a fault risk warning based on the operation of the device; judge the authenticity of the fault risk warning according to the environmental confidence fitting model.

[0011] According to the above solution, in S1, the following content is included:

[0012] S101: Mark and extract the device operation data when each device fails, and generate a device fault data set; the device operation data includes current value, voltage value, operating power, and load.

[0013] S102: Set weights for each device operation data, and calculate the fault operation value of the device according to the device fault data set of each device and the weights corresponding to the device operation data; form a set of the fault operation values of device DE i and denote it as SDE i ={SDE （i，1） , SDE （i，2） , …, SDE （i，J）}; i ∈ [1, I], where I represents the total number of faulty devices; where SDE （i，j） represents the j-th fault operation value in the fault operation value set of device DE i ; j ∈ [1, J]; J represents the number of fault operation values in the fault operation value set.

[0014] S103: Calculate the fault difference of the device according to the fault operation value set of each device:

[0015] CDF i = max{SSDE i max - RSSDE i , SSDE i max - VSSDE i};

[0016] Among them, CDF i represents the fault difference of device DE i , max{} represents the maximum value operation, SSDE i max represents the maximum value of the fault operation values in the fault operation value set SDE i , RSSDE i represents the root mean square value of the fault operation values in the fault operation value set SDE i , and VSSDE i represents the mean value of the fault operation values in the fault operation value set SDE i .

[0017] S104: Generate a set of device fault differences according to the calculation formula in S103, denoted as CDF = {CDF 1 , CDF 2 , …, CDF I}; Calculate the mean of the set of device fault differences, denoted as the fault feature evaluation data VCDF of the device.

[0018] According to the above solution, the first environmental data is temperature sensor data.

[0019] According to the above solution, in S2, analyze the abnormal correlation degree between the fault feature value and the corresponding first environmental data, including the following content:

[0020] S201: Extract the fault feature evaluation data VCDF when the device fails, and form a set denoted as FCD = {VCDF 1 , VCDF 2 , …, VCDF N}, where N represents the total number of fault feature evaluation data in the fault feature evaluation data set; Analyze and calculate the fault feature value according to the fault feature evaluation data set:

[0021] ;

[0022] where FCV represents the fault feature value, VCDF n represents the nth fault feature evaluation data in the fault feature evaluation data set, n ∈ [1, N]; μ represents the mean of the fault feature evaluation data set, σ represents the standard deviation of the fault feature evaluation data set; exp[] represents the exponential function with the real number e as the base;

[0023] S202: Extract the fault occurrence times when the fault feature evaluation data is greater than the fault feature value from the fault feature evaluation data set, and form a set denoted as TG = {TG 1 , TG 2 , …, TG M}; where TG m represents the mth fault time in the fault time set; M represents the total number of fault times in the fault time set, M < N; Set an environmental first threshold for the first environmental value. If the first environmental value is greater than or equal to the environmental first threshold, mark the first environmental value. If the first environmental value is less than the environmental first threshold, do not process the first environmental value; Generate a first environmental abnormal time set for all the marked first environmental value corresponding times; Denote it as TFE = {TFE 1 , TFE 2 ,..., TFE Q}, where TFE 1 , TFE 2,..., TFE Q respectively represent the moments when the 1st, 2nd,..., Qth marked first environmental values are greater than or equal to the threshold;

[0024] S203: Calculate the abnormal correlation degree between the set of fault moments and the set of first environmental abnormal moments within the same monitoring period:

[0025] A(TG→TFE)=num 2 [TG∩TFE]÷(M×Q);

[0026] where A(TG→TFE) represents the abnormal correlation degree between the set of fault moments TG and the set of first environmental abnormal moments TFE, and num 2 [TG∩TFE] represents the square of the total number of moments in the intersection set of the set of fault moments TG and the set of first environmental abnormal moments TFE.

[0027] Through the calculation of the abnormal correlation degree, the set of fault moments and the set of first environmental abnormal moments can be effectively closely related, providing data support for the establishment of subsequent various models.

[0028] According to the above scheme, a fault feature fluctuation model is constructed in S2, including the following:

[0029] S211: Set a threshold for the abnormal correlation degree, and obtain the corresponding set of fault moments and the set of first environmental abnormal moments when the abnormal correlation degree is less than the threshold; extract the data pairs composed of the corresponding fault feature values and first environmental values when the fault moments and the first environmental abnormal moments are the same, denoted as (TEM a , FCV a ); where TEM a represents the first environmental data of the a-th data pair, and FCV a represents the fault feature value of the a-th data pair;

[0030] S212: Take the coordinate point (0,0) as the origin, the x-axis corresponds to the first environmental data and the y-axis corresponds to the fault feature value, construct a plane rectangular coordinate system, mark the coordinate points corresponding to the data pairs in the plane rectangular coordinate system, and connect the adjacent marked coordinate points in the plane rectangular coordinate system in ascending order of the x-axis coordinate values, and use the function corresponding to the obtained broken line as the constructed fault feature fluctuation model.

[0031] By establishing the fault feature fluctuation model, the operation state and fault state of the device can be effectively judged through the first environmental data. The administrator can set regular monitoring for the device operation data, without the need to detect the device operation data for a long time, which can effectively reduce energy consumption and reduce the monitoring cost of users.

[0032] According to the above solution, in S3, the following contents are included:

[0033] S301: Set the data extraction duration for other environmental data, and extract the other environmental data except the first environmental data before the fault moment involved in the fault feature fluctuation model according to the data extraction duration; generate an environmental data association set based on the extracted first environmental data and other environmental data, denoted as EDA = {(TEM b , TTEM （c，b） ) | b ∈ [1, B], c ∈ [1, C]}; where B represents the number of data groups in the environmental data association set, C represents the number of categories of other environmental data, TEM b represents the first environmental value of the b-th data group, and TTEM （c，b） represents the other environmental value of the b-th data group of the c-th other environmental data;

[0034] S302: Based on the environmental data association set, establish an environmental association fitting model:

[0035] TTEM = β c 1 × TEM + β c 2 ;

[0036] where β c 1 and β c 2 represent fitting coefficients, TEM represents the independent variable of the first environmental value, TTEM represents the dependent variable of the other environmental value, and use the least squares method to calculate and solve β c 1 and β c 2 in the environmental association fitting model;

[0037] S303: Set a threshold for the first fitting coefficient β c 1 , extract the first fitting coefficients greater than the threshold, and calculate the mean value, denoted as the first fitting coefficient eigenvalue β 1 ; extract the second fitting coefficient β c 2 corresponding to the first fitting coefficients greater than the threshold, and calculate the mean value, denoted as the second fitting coefficient eigenvalue β 2 ; according to the first fitting coefficient eigenvalue and the second fitting coefficient eigenvalue, construct an environmental confidence fitting model: TTEM = β 1 × TEM + β 2 .

[0038] By means of a multi-environment correlation fitting model, an environmental confidence fitting model is further constructed, which can effectively avoid the situation that a single data source cannot comprehensively reflect the complex operating state of the device and further improve the accuracy of the system.

[0039] According to the above solution, in S4, the following contents are included:

[0040] S401: Collect real-time first environmental data, input the real-time first environmental data into the fault feature fluctuation model, calculate the real-time fault feature value, set a threshold for the real-time fault feature value. If the real-time fault feature value is greater than the threshold, it is determined that there is a fault risk; if the real-time fault feature value is less than or equal to the threshold, it is determined that there is no fault risk.

[0041] S402: Input the real-time first environmental data corresponding to the fault risk into the environmental confidence fitting model to calculate the real-time other environmental values; extract the real-time environmental data collected by the sensor when there is a fault risk, and calculate the average value of the real-time environmental data. If the average value of the real-time environmental data is greater than the real-time other environmental values, it is determined that the fault risk is credible, and the fault of the device is maintained; if the average value of the real-time environmental data is less than or equal to the real-time other environmental values, it is determined that the fault risk is not credible, and the first environmental data acquisition sensor is calibrated.

[0042] Calculating the real-time fault feature value of the device through the first environmental data can effectively judge the operating state of the device according to the model constructed based on historical data; the present invention further determines the fault risk through other environmental data, which can effectively avoid the inaccurate judgment caused by the environmental interference of the first environmental data, judge the abnormality of the first environmental sensor, and remind the administrator to calibrate, which can effectively guarantee the accuracy of subsequent monitoring and further improve the comprehensiveness and reliability of the system.

[0043] Another aspect of the present application provides a device data analysis and management system applying a temperature sensor. The system is implemented by applying the above-mentioned device data analysis and management method applying a temperature sensor. The system includes an environment and operation data acquisition module, a fault feature analysis module, an environmental fitting model construction module, and a warning judgment and calibration module.

[0044] The environment and operation data acquisition module is used to collect environmental data of the device by installing sensors on the device; extract the device operation data when a fault occurs during the historical operation of the device, and generate fault feature evaluation data of the device after analyzing the device operation data.

[0045] The fault feature analysis module is used to calculate the device fault feature value according to the feature data when the device fails in history; analyze the abnormal correlation degree between the fault feature value and the corresponding first environmental data; generate an abnormal correlation combination according to the abnormal correlation degree, and construct a fault feature fluctuation model according to the abnormal correlation combination;

[0046] The environmental fitting model construction module is used to extract other environmental data except the first environmental data when the device fails in history, form an environmental data association set according to the first environmental data and other environmental data, establish an environmental association fitting model according to the environmental data association set, and construct an environmental confidence fitting model through the environmental association fitting models of two or more other environmental data;

[0047] The warning judgment calibration module is used to collect real-time first environmental data, analyze the operation condition of the device by using the feature fluctuation model; generate a fault risk warning according to the operation condition of the device; judge the true situation of the fault risk warning according to the environmental confidence fitting model.

[0048] The environmental and operation data acquisition module includes an environmental data acquisition unit and a fault feature evaluation unit;

[0049] The environmental data acquisition unit installs sensors on the device and uses the sensors to collect environmental data of the device;

[0050] The fault feature evaluation unit marks and extracts the device operation data when each device fails, and generates a device fault data set; sets weights for each device operation data, calculates the fault operation value of the device; forms a set of the fault operation values of the device, generates a device fault difference set according to the set of the fault operation values of each device, and calculates the mean value of the device fault difference set, which is recorded as the fault feature evaluation data of the device;

[0051] The fault feature analysis module includes an abnormal correlation degree calculation unit and a fault feature model construction unit;

[0052] The abnormal correlation degree calculation unit analyzes and calculates the fault feature value according to the fault feature evaluation data set; generates a fault time set and a first environmental abnormal time set, and analyzes the abnormal correlation degree between the fault time set and the first environmental abnormal time set within the same monitoring period;

[0053] The fault feature model construction unit is used to set a threshold for the abnormal correlation degree, and when the abnormal correlation degree is less than the threshold, obtain the corresponding set of fault moments and the first set of environmental abnormal moments; extract the data pairs composed of the corresponding fault feature values and the first environmental values when the fault moments and the first environmental abnormal moments are the same; take the coordinate point (0, 0) as the origin, the x-axis corresponds to the first environmental data and the y-axis corresponds to the fault feature values, construct a plane rectangular coordinate system, mark the coordinate points corresponding to the data pairs in the plane rectangular coordinate system, and connect the adjacent marked coordinate points in the plane rectangular coordinate system in ascending order of the x-axis coordinate values, and use the function corresponding to the obtained broken line as the constructed fault feature fluctuation model.

[0054] The environmental fitting model construction module includes an environmental correlation fitting model construction unit and an environmental confidence fitting model construction unit;

[0055] The environmental correlation fitting model construction unit is used to set a data acquisition duration for other environmental data, and extract other environmental data except the first environmental data before the fault moment involved in the fault feature fluctuation model according to the data acquisition duration; generate an environmental data correlation set based on the extracted first environmental data and other environmental data, and establish an environmental correlation fitting model based on the environmental data correlation set;

[0056] The environmental confidence fitting model construction unit is used to set a threshold for the first fitting coefficient, extract the first fitting coefficients greater than the threshold, and calculate the mean value, denoted as the first fitting coefficient eigenvalue; extract the second fitting coefficients corresponding to the first fitting coefficients greater than the threshold, and calculate the mean value, denoted as the second fitting coefficient eigenvalue; construct an environmental confidence fitting model according to the first fitting coefficient eigenvalue and the second fitting coefficient eigenvalue.

[0057] The warning judgment and calibration module includes a fault risk judgment unit and a fault risk calibration unit;

[0058] The fault risk judgment unit is used to collect real-time first environmental data, substitute the real-time first environmental data into the fault feature fluctuation model, calculate the real-time fault feature value, set a threshold for the real-time fault feature value, if the real-time fault feature value is greater than the threshold, it is determined that there is a fault risk; if the real-time fault feature value is less than or equal to the threshold, it is determined that there is no fault risk;

[0059] The fault risk calibration unit inputs the real-time first environmental data corresponding to the fault risk into the environmental confidence fitting model to calculate the real-time other environmental values; extracts the real-time environmental data collected by the sensor when there is a fault risk, and calculates the mean value of the real-time environmental data; if the mean value of the real-time environmental data is greater than the real-time other environmental values, it is determined that the fault risk is credible, and the fault of the device is maintained; if the mean value of the real-time environmental data is less than or equal to the real-time other environmental values, it is determined that the fault risk is not credible, and the first environmental data acquisition sensor is calibrated.

[0060] Compared with the prior art, the beneficial effects of the present invention are as follows: By calculating the abnormal correlation degree, the fault time set and the first environmental abnormal time set can be effectively and closely related, providing data support for the establishment of subsequent various models; By establishing a fault feature fluctuation model, the operation state and fault state of the device can be effectively judged through the first environmental data. The administrator can set regular monitoring of the device operation data, without the need to detect the device operation data for a long time, which can effectively reduce energy consumption and reduce the monitoring cost of users; Through multiple environmental correlation fitting models, an environmental confidence fitting model is further constructed, which can effectively avoid the inability of a single data source to comprehensively reflect the complex operation state of the device, and further improve the accuracy of the system; By calculating the real-time fault feature value of the device through the first environmental data, the operation state of the device can be effectively judged according to the model constructed based on historical data; The present invention further determines the fault risk through other environmental data, which can effectively avoid the first environmental data being affected by environmental interference, resulting in inaccurate judgment, judge the abnormality of the first environmental sensor, and remind the administrator to calibrate, which can effectively ensure the accuracy of subsequent monitoring, and further improve the comprehensiveness and reliability of the system. Description of the Drawings

[0061] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention, and do not constitute a limitation to the present invention. In the drawings:

[0062] Figure 1 It is a schematic flowchart of a device data analysis and management method applying a temperature sensor according to the present invention;

[0063] Figure 2 It is a schematic structural diagram of a device data analysis and management system applying a temperature sensor according to the present invention. Detailed Embodiments

[0064] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with 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 the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0065] Please refer to Figure 1 , the present invention provides a technical solution: A method for analyzing and managing device data applied with a temperature sensor, the method comprising the following steps:

[0066] S1: By installing sensors on the device, use the sensors to collect environmental data of the device; extract the device operation data when a failure occurs during the historical operation of the device, and generate device failure characteristic evaluation data after analyzing the device operation data;

[0067] According to the above solution, in S1, the following contents are included:

[0068] S101: Mark and extract the device operation data when each device fails, and generate a device failure data set; the device operation data includes current value, voltage value, operating power, and load;

[0069] S102: Set weights for each device operation data, and calculate the failure operation value of the device according to the device failure data set of each device and the weights corresponding to the device operation data; Denote the set composed of the failure operation values of device DE i as SDE i = {SDE （i，1） , SDE （i，2） , …, SDE （i，J）}; i ∈ [1, I], where I represents the total number of failed devices; where SDE （i，j） represents the j-th failure operation value in the failure operation value set of device DE i ; j ∈ [1, J]; J represents the number of failure operation values in the failure operation value set;

[0070] S103: Calculate the failure difference of the device according to the failure operation value sets of each device:

[0071] CDF i = max{SSDE i max - RSSDE i , SSDE i max - VSSDE i};

[0072] Wherein, CDFi Denote the device as DE i The fault difference value, max{} represents the maximum value operation, SSDE i max Denote the set of fault operation values as SDE i The maximum value of the fault operation values in the set of fault operation values SDE, RSSDE i Denote the set of fault operation values as SDE i The root mean square value of the fault operation values in the set of fault operation values SDE, VSSDE i Denote the set of fault operation values as SDE i The mean value of the fault operation values in the set of fault operation values SDE;

[0073] S104: According to the calculation formula in S103, generate a set of device fault difference values, denoted as CDF = {CDF 1 , CDF 2 , …, CDF I}; Calculate the mean value of the set of device fault difference values, denoted as the fault feature evaluation data VCDF of the device.

[0074] S2: Calculate the device fault feature value according to the feature data when the device fails in history; Analyze the abnormal correlation degree between the fault feature value and the corresponding first environmental data; Generate an abnormal correlation combination according to the abnormal correlation degree, and construct a fault feature fluctuation model according to the abnormal correlation combination;

[0075] According to the above solution, the first environmental data is the temperature sensor data.

[0076] According to the above solution, in S2, analyzing the abnormal correlation degree between the fault feature value and the corresponding first environmental data includes the following content:

[0077] S201: Extract the fault feature evaluation data VCDF when the device fails, and form a set denoted as FCD = {VCDF 1 , VCDF 2 , …, VCDF N}, where N represents the total number of fault feature evaluation data in the set of fault feature evaluation data; Analyze and calculate the fault feature value according to the set of fault feature evaluation data:

[0078] ;

[0079] where FCV represents the fault feature value, VCDF n represents the nth fault feature evaluation data in the set of fault feature evaluation data, n ∈ [1, N]; μ represents the mean value of the set of fault feature evaluation data, σ represents the standard deviation of the set of fault feature evaluation data; exp[] represents the exponential function with the real number e as the base;

[0080] S202: Extract the fault occurrence times when the fault feature evaluation data is greater than the fault feature value from the fault feature evaluation data set, and form a set denoted as TG = {TG 1 , TG 2 , …, TG M}; where TG m represents the m-th fault time in the fault time set; M represents the total number of fault times in the fault time set, M < N; Set an environmental first threshold for the first environmental value. If the first environmental value is greater than or equal to the environmental first threshold, mark the first environmental value. If the first environmental value is less than the environmental first threshold, do not process the first environmental value; Generate a first environmental anomaly time set for all the marked first environmental value corresponding times; Denote it as TFE = {TFE 1 , TFE 2 ,..., TFE Q}, where TFE 1 , TFE 2 ,..., TFE Q respectively represent the times when the 1st, 2nd,..., Q-th first environmental values greater than or equal to the threshold are marked;

[0081] S203: Calculate the anomaly correlation degree between the fault time set and the first environmental anomaly time set within the same monitoring period:

[0082] A(TG → TFE) = num 2 [TG ∩ TFE] ÷ (M × Q);

[0083] where A(TG → TFE) represents the anomaly correlation degree between the fault time set TG and the first environmental anomaly time set TFE, and num 2 [TG ∩ TFE] represents the square of the total number of times in the intersection set of the fault time set TG and the first environmental anomaly time set TFE.

[0084] Example 1, in this example, the fault time set is {1S, 2S, 3S, 4S, 5S, 6S, 7S};

[0085] The first environmental anomaly time set is {2S, 3S, 4S, 5S, 6S, 7S, 8S};

[0086] So TG ∩ TFE is {1S, 2S, 3S, 4S, 5S, 6S, 7S, 8S};

[0087] A(TG → TFE) = 8 2 ÷(7 × 7) ≈ 1.31;

[0088] According to the above solution, in S2, construct a fault feature fluctuation model, including the following content:

[0089] S211: Set a threshold for the abnormal correlation degree. When the abnormal correlation degree is less than the threshold, obtain the corresponding set of fault moments and the first set of environmental abnormal moments; extract the data pairs composed of the corresponding fault feature values and the first environmental values when the fault moments and the first environmental abnormal moments are the same, denoted as (TEM a , FCV a ); where TEM a represents the first environmental data of the a-th data pair, and where FCV a represents the fault feature value of the a-th data pair;

[0090] S212: Taking the coordinate point (0, 0) as the origin, with the x-axis corresponding to the first environmental data and the y-axis corresponding to the fault feature value, construct a plane rectangular coordinate system. Mark the coordinate points corresponding to the data pairs in the plane rectangular coordinate system, and connect the adjacent marked coordinate points in the plane rectangular coordinate system in ascending order of the x-axis coordinate values. Take the function corresponding to the obtained broken line as the established fault feature fluctuation model.

[0091] S3: Extract other environmental data other than the first environmental data when the equipment fails in history. According to the first environmental data and other environmental data, form an environmental data association set. Based on the environmental data association set, establish an environmental association fitting model, and construct an environmental confidence fitting model through the environmental association fitting models of two or more other environmental data;

[0092] According to the above solution, in S3, it includes the following content:

[0093] S301: Set the data acquisition duration for other environmental data, and extract other environmental data other than the first environmental data before the fault moment involved in the fault feature fluctuation model according to the data acquisition duration; generate an environmental data association set according to the extracted first environmental data and other environmental data, denoted as EDA = {(TEM b , TTEM （c，b） ) | b ∈ [1, B], c ∈ [1, C]}; where B represents the number of data groups in the environmental data association set, C represents the number of categories of other environmental data, TEM b represents the first environmental value of the b-th data group, and TTEM （c，b） represents the other environmental value of the c-th type of other environmental data in the b-th data group;

[0094] S302: Based on the environmental data association set, establish an environmental association fitting model:

[0095] TTEM = β c 1 ×TEM + β c 2 ;

[0096] where β c 1 and β c 2 represent fitting coefficients, TEM represents the independent variable of the first environmental value, TTEM represents the dependent variable of other environmental values, and the least squares method is used to calculate and solve β c 1 and β c 2 in the environmental correlation fitting model;

[0097] S303: Set a threshold for the first fitting coefficient β c 1 Extract the first fitting coefficients greater than the threshold and calculate the mean value, denoted as the first fitting coefficient eigenvalue β 1 ; Extract the second fitting coefficient β c 2 corresponding to the first fitting coefficients greater than the threshold, and calculate the mean value, denoted as the second fitting coefficient eigenvalue β 2 ; According to the first fitting coefficient eigenvalue and the second fitting coefficient eigenvalue, construct an environmental confidence fitting model: TTEM = β 1 ×TEM + β 2 .

[0098] S4: Collect real-time first environmental data, analyze the operation status of the device using the characteristic fluctuation model; generate a fault risk warning according to the operation status of the device; judge the authenticity of the fault risk warning according to the environmental confidence fitting model.

[0099] According to the above solution, in S4, it includes the following content:

[0100] S401: Collect real-time first environmental data, substitute the real-time first environmental data into the fault characteristic fluctuation model, calculate the real-time fault characteristic value, set a threshold for the real-time fault characteristic value, if the real-time fault characteristic value is greater than the threshold, it is determined that there is a fault risk; if the real-time fault characteristic value is less than or equal to the threshold, it is determined that there is no fault risk;

[0101] S402: Substitute the real-time first environmental data corresponding to the fault risk into the environmental confidence fitting model, calculate the real-time other environmental values; extract the real-time environmental data collected by the sensor when there is a fault risk, and calculate the mean value of the real-time environmental data; if the mean value of the real-time environmental data is greater than the real-time other environmental values, judge that the fault risk is credible and maintain the device's fault; if the mean value of the real-time environmental data is less than or equal to the real-time other environmental values, judge that the fault risk is not credible and calibrate the first environmental data acquisition sensor.

[0102] Please refer to Figure 2, the present invention provides a technical solution: On another aspect of this application, a device data analysis and management system applying a temperature sensor is provided. The system is implemented by applying the above-mentioned device data analysis and management method applying a temperature sensor. The system includes an environment and operation data acquisition module, a fault feature analysis module, an environment fitting model construction module, and a warning judgment and calibration module;

[0103] The environment and operation data acquisition module is used to collect environment data of the device by installing sensors on the device; extract the device operation data when a fault occurs during the historical operation of the device, and generate device fault feature evaluation data after analyzing the device operation data;

[0104] The fault feature analysis module is used to calculate the device fault feature value according to the feature data when a fault occurs in the device in history; analyze the abnormal correlation degree between the fault feature value and the corresponding first environment data; generate an abnormal correlation combination according to the abnormal correlation degree, and construct a fault feature fluctuation model according to the abnormal correlation combination;

[0105] The environment fitting model construction module is used to extract other environment data except the first environment data when a fault occurs in the device in history, form an environment data association set according to the first environment data and other environment data, establish an environment association fitting model according to the environment data association set, and construct an environment confidence fitting model through the environment association fitting models of two or more other environment data;

[0106] The warning judgment and calibration module is used to collect real-time first environment data, analyze the operation condition of the device by using the feature fluctuation model; generate a fault risk warning according to the operation condition of the device; judge the true situation of the fault risk warning according to the environment confidence fitting model.

[0107] The environment and operation data acquisition module includes an environment data acquisition unit and a fault feature evaluation unit;

[0108] The environment data acquisition unit collects environment data of the device by installing sensors on the device;

[0109] The fault feature evaluation unit marks and extracts the device operation data when a fault occurs in each device, and generates a device fault data set; sets weights for each device operation data, calculates the fault operation value of the device; forms a set of the fault operation values of the device, generates a device fault difference set according to the set of the fault operation values of each device, and calculates the mean value of the device fault difference set, which is recorded as the device fault feature evaluation data;

[0110] The fault feature analysis module includes an abnormal correlation degree calculation unit and a fault feature model construction unit;

[0111] The abnormal correlation degree calculation unit analyzes and calculates the fault feature values based on the fault feature evaluation data set; generates the fault time set and the first environmental abnormal time set, and analyzes the abnormal correlation degree between the fault time set and the first environmental abnormal time set within the same monitoring period;

[0112] The fault feature model construction unit is used to set a threshold for the abnormal correlation degree, and obtain the corresponding fault time set and the first environmental abnormal time set when the abnormal correlation degree is less than the threshold; extract the data pairs composed of the corresponding fault feature values and the first environmental values when the fault time and the first environmental abnormal time are the same; take the coordinate point (0,0) as the origin, the x-axis corresponds to the first environmental data and the y-axis corresponds to the fault feature value, construct a plane rectangular coordinate system, mark the coordinate points corresponding to the data pairs in the plane rectangular coordinate system, and connect the adjacent marked coordinate points in the plane rectangular coordinate system in ascending order of the x-axis coordinate values, and use the function corresponding to the obtained broken line as the constructed fault feature fluctuation model.

[0113] The environmental fitting model construction module includes an environmental correlation fitting model construction unit and an environmental confidence fitting model construction unit;

[0114] The environmental correlation fitting model construction unit is used to set the data acquisition duration for other environmental data, and extract other environmental data except the first environmental data before the fault time involved in the fault feature fluctuation model according to the data acquisition duration; generate an environmental data correlation set based on the extracted first environmental data and other environmental data, and establish an environmental correlation fitting model based on the environmental data correlation set;

[0115] The environmental confidence fitting model construction unit is used to set a threshold for the first fitting coefficient, extract the first fitting coefficients greater than the threshold, and calculate the mean value, denoted as the first fitting coefficient eigenvalue; extract the second fitting coefficients corresponding to the first fitting coefficients greater than the threshold, and calculate the mean value, denoted as the second fitting coefficient eigenvalue; construct an environmental confidence fitting model according to the first fitting coefficient eigenvalue and the second fitting coefficient eigenvalue.

[0116] The early warning judgment and calibration module includes a fault risk judgment unit and a fault risk calibration unit;

[0117] The fault risk judgment unit is used to collect real-time first environmental data, substitute the real-time first environmental data into the fault feature fluctuation model, calculate the real-time fault feature value, set a threshold for the real-time fault feature value, if the real-time fault feature value is greater than the threshold, it is determined that there is a fault risk; if the real-time fault feature value is less than or equal to the threshold, it is determined that there is no fault risk;

[0118] The fault risk calibration unit inputs the real-time first environmental data corresponding to the existing fault risk into the environmental confidence fitting model to calculate the real-time other environmental values; extracts the real-time environmental data collected by the sensor when there is a fault risk, and calculates the mean value of the real-time environmental data; if the mean value of the real-time environmental data is greater than the real-time other environmental values, it is determined that the fault risk is credible, and the fault of the device is maintained; if the mean value of the real-time environmental data is less than or equal to the real-time other environmental values, it is determined that the fault risk is not credible, and the first environmental data acquisition sensor is calibrated.

[0119] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0120] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present invention. Any reference signs in the claims should not be regarded as limiting the claims involved.

Claims

1. A device data analysis and management method using a temperature sensor, characterized in that: The method comprises the following steps: S1: Install sensors on the equipment and use them to collect environmental data. Extract equipment operation data when failures occurred during the equipment's historical operation, analyze the equipment operation data, and generate equipment failure feature evaluation data. S2: Calculate the equipment fault characteristic value according to the fault characteristic evaluation data when the equipment fails in history; analyze the abnormal correlation between the fault characteristic value and the corresponding first environment data; generate an abnormal correlation combination according to the abnormal correlation, and construct a fault characteristic fluctuation model according to the abnormal correlation combination; S3: extracting other environmental data except the first environmental data when the device fails in the history, forming an environmental data association set according to the first environmental data and other environmental data, establishing an environmental association fitting model according to the environmental data association set, and constructing an environmental confidence fitting model through the environmental association fitting models of two or more other environmental data; S4: collect real-time first environment data, and use the fault characteristic fluctuation model to analyze the operation status of the equipment; generate a fault risk warning based on the operation status of the equipment; Determine the true situation of fault risk warning based on the environmental confidence fitting model; In S2, analyzing the abnormal correlation between the fault characteristic value and the corresponding first environment data includes the following contents: S201: Extract the fault feature evaluation data VCDF when the equipment fails, and form a set denoted as FCD={VCDF1, VCDF2, ..., VCDF N }, where N represents the total number of fault feature evaluation data in the fault feature evaluation data set; the fault feature value is calculated based on the fault feature evaluation data set analysis: ； Where FCV represents the fault characteristic value, VCDF n represents the nth fault feature evaluation data in the fault feature evaluation data set, n∈[1,N]; μ represents the mean of the fault feature evaluation data set, and σ represents the standard deviation of the fault feature evaluation data set; S202: Extract the fault occurrence time when the fault feature evaluation data is greater than the fault feature value from the fault feature evaluation data set, and form a set denoted as TG={TG1, TG2, ..., TG M }; where TG m represents the mth fault moment in the fault moment set; M represents the total number of fault moments in the fault moment set, M<N; set the first environmental threshold for the first environmental value, if the first environmental value is greater than or equal to the first environmental threshold, mark the first environmental value, if the first environmental value is less than the first environmental threshold, do not process the first environmental value; generate the first environmental abnormal moment set for all the marked first environmental values ​​corresponding moments; denoted as TFE={TFE1, TFE2, ..., TFE Q }, where TFE1, TFE2, ..., TFE Q Respectively represent the moments when the 1st, 2nd, ..., Qth marked first environment values ​​are greater than or equal to the threshold; S203: Calculate the abnormal correlation between the fault time set and the first environment abnormal time set in the same monitoring period: A (TG → TFE) = number 2 [TG∩TFE]÷(M×Q); A(TG→TFE) represents the abnormal correlation between the fault time set TG and the first environmental abnormal time set TFE, num 2 [TG∩TFE] represents the square of the total number of moments in the intersection of the fault time set TG and the first environment abnormal time set TFE.

2. The device data analysis and management method using a temperature sensor according to claim 1, characterized in that: In S1, the following contents are included: S101: marking and extracting the equipment operation data of each equipment when a fault occurs, and generating an equipment fault data set; the equipment operation data includes current value, voltage value, operation power and load; S102: Set weights for each device operation data, and calculate the device failure operation value according to the device failure data set of each device and the weight corresponding to the device operation data; i The fault operation value set is recorded as SDE i ={SDE （i，1） , SDE （i，2） , …, SDE （i，J） }; i∈[1,I], where I represents the total number of faulty devices; where SDE （i，j） Indicates the device DE i The jth fault operation value in the fault operation value set; j∈[1,J]; J represents the number of faulty operation values ​​in the faulty operation value set; S103: Calculate the fault difference of each device according to the fault operation value set of each device: CDF i =max{SSDE i max -RSSDE i ,SSDE i max -VSSDE i }; Among them, CDF i Indicates the device DE i Fault difference, max{} indicates maximum value operation, SSDE i max Represents the fault operation value set SDE i The maximum value of the fault operation value, RSSDE i Represents the fault operation value set SDE i The root mean square value of the fault operation value, VSSDE i Represents the fault operation value set SDE i The mean of the fault operation values; S104: Generate a set of equipment fault difference values ​​according to the calculation formula in S103, denoted as CDF={CDF1, CDF2, …, CDF I }; Calculate the mean of the equipment fault difference set and record it as the equipment fault feature evaluation data VCDF.

3. The device data analysis and management method using a temperature sensor according to claim 1, characterized in that: The first environmental data is temperature sensor data.

4. The device data analysis and management method using a temperature sensor according to claim 2, characterized in that: The fault characteristic fluctuation model is constructed in S2, including the following: S211: Set a threshold for the abnormal correlation, and obtain the corresponding fault time set and the first environment abnormal time set when the abnormal correlation is less than the threshold; extract the corresponding fault feature value and the first environment value to form a data pair when the fault time and the first environment abnormal time are the same, recorded as (TEM a , FCV a ); where TEM a Represents the first environment data of the a-th data pair, where FCV a represents the fault characteristic value of the ath data pair; S212: With the coordinate point (0,0) as the origin, the x-axis corresponding to the first environmental data and the y-axis corresponding to the fault characteristic value, a plane rectangular coordinate system is constructed, the coordinate points corresponding to the data are marked in the plane rectangular coordinate system, and the adjacent marked coordinate points in the plane rectangular coordinate system are connected in order from small to large in the x-axis coordinate value, and the function corresponding to the obtained broken line is used as the constructed fault characteristic fluctuation model.

5. The device data analysis and management method using a temperature sensor according to claim 4, characterized in that: In S3, include the following: S301: Set a data acquisition time for other environmental data, and extract other environmental data except the first environmental data before the fault time involved in the fault feature fluctuation model according to the data acquisition time; generate an environmental data association set based on the extracted first environmental data and other environmental data, recorded as EDA={(TEM b ,TTEM （c，b） )|b∈[1,B],c∈[1,C]}; where B represents the number of data groups in the environmental data association set, C represents the number of categories of other environmental data, TEM b Indicates the first environmental value of the bth data group, TTEM （c，b） Indicates other environmental values ​​of the bth data group of the cth type of other environmental data; S302: Establishing an environment association fitting model based on the environment data association set: TTEM=β c 1×TEM+β c 2; where β c 1 and β c 2 represents the fitting coefficient, TEM represents the independent variable of the first environmental value, TTEM represents the dependent variable of other environmental values, and the β in the environmental association fitting model is fitted using the least squares method. c 1 and β c 2. Perform calculations and solve; S303: For the first fitting coefficient β c 1 Set the threshold, extract the first fitting coefficient greater than the threshold, and calculate the mean, which is recorded as the first fitting coefficient eigenvalue β1; extract the second fitting coefficient β corresponding to the first fitting coefficient greater than the threshold c 2, and calculate the mean, which is recorded as the second fitting coefficient eigenvalue β2; according to the first fitting coefficient eigenvalue and the second fitting coefficient eigenvalue, the environmental confidence fitting model is constructed: TTEM=β1×TEM+β2.

6. The device data analysis and management method using a temperature sensor according to claim 5, characterized in that: In S4, the following contents are included: S401: collecting real-time first environment data, bringing the real-time first environment data into a fault characteristic fluctuation model, calculating a real-time fault characteristic value, setting a threshold for the real-time fault characteristic value, and determining that a fault risk exists if the real-time fault characteristic value is greater than the threshold; If the real-time fault characteristic value is less than or equal to the threshold, it is determined that there is no fault risk; S402: Bring the real-time first environment data corresponding to the fault risk into the environment confidence fitting model to calculate the real-time other environment values; extract the real-time environment data collected by the sensor when the fault risk exists, and calculate the mean of the real-time environment data; if the mean of the real-time environment data is greater than the real-time other environment values, it is determined that the fault risk is credible, and the equipment fault is maintained; If the mean value of the real-time environmental data is less than or equal to other real-time environmental values, it is determined that the fault risk is not credible, and the first environmental data acquisition sensor is calibrated.

7. A device data analysis and management system using a temperature sensor, the system being applied to a device data analysis and management method using a temperature sensor as claimed in any one of claims 1 to 6, characterized in that: The system includes an environment and operation data acquisition module, a fault feature analysis module, an environment fitting model construction module and an early warning judgment calibration module; The environment and operation data acquisition module is used to collect environmental data of the equipment by installing sensors on the equipment; extract the equipment operation data when the equipment fails during historical operation, and generate equipment failure feature evaluation data after analyzing the equipment operation data; The fault feature analysis module is used to calculate the equipment fault feature value according to the fault feature evaluation data when the equipment fails in history; analyze the abnormal correlation between the fault feature value and the corresponding first environment data; generate an abnormal correlation combination according to the abnormal correlation, and construct a fault feature fluctuation model according to the abnormal correlation combination; The environment fitting model construction module is used to extract other environment data except the first environment data when the equipment fails in the history, form an environment data association set according to the first environment data and other environment data, establish an environment association fitting model according to the environment data association set, and construct an environment confidence fitting model through the environment association fitting models of two or more other environment data; The early warning judgment calibration module is used to collect real-time first environment data, analyze the operation status of the equipment using the fault characteristic fluctuation model; and generate a fault risk early warning according to the operation status of the equipment; The actual situation of fault risk warning is judged according to the environmental confidence fitting model.

Images