An inductor-based equipment monitoring and protection method and system

Abstract

The present invention discloses an inductor-based equipment monitoring and protection method and system, which relate to the technical field of inductor monitoring. By constructing operating state sample data, fault type sample data and operating life sample data, a fault judgment model, a fault type judgment model and an operating life prediction model are trained. For the inductor to be monitored, actual operating data is collected in real time, and based on the actual operating data and the fault judgment model, it is judged whether the inductor has a fault. Based on the actual operating data and the fault type judgment model, the fault type is judged and a fault warning is performed. Actual operating life data is formed based on all actual operating data and fault types of the inductor to be monitored during operation. Based on the actual operating life data and the operating life prediction model, the expected life of the inductor to be monitored is obtained, and a scrap warning is generated based on the expected life. The intelligence, safety and reliability level of the power system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of inductor monitoring, and in particular to an inductor-based equipment monitoring and protection method and system. Background Art

[0002] Power equipment is a key infrastructure for modern industrial production and people's livelihood security. Its safe and reliable operation is of great significance to ensuring the stability of the power system. As a core component in the power system, the status of the inductor is directly related to the operating efficiency and reliability of the entire system. Traditional inductor status monitoring methods mainly rely on manual inspections and regular testing, which have problems of lag and human subjective judgment. It is difficult to detect potential faults in time, and there is a lack of accurate assessment of the remaining life of the equipment.

[0003] The existing inductor condition monitoring technology has the following shortcomings: 1) It lacks intelligent fault judgment methods based on big data, making it difficult to accurately identify the operating status of the equipment; 2) It cannot effectively distinguish different types of faults, resulting in a blind spot in fault diagnosis; 3) For the remaining life assessment of equipment, it mainly relies on experience accumulation, lacks a systematic data-driven method, and has low prediction accuracy. These technical defects restrict the modernization level of power equipment monitoring, early warning, and life cycle management.

[0004] To this end, the present invention proposes an inductor-based equipment monitoring and protection method and system. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes an equipment monitoring and protection method and system based on an inductor, which improves the intelligence, safety and reliability level of the power system.

[0006] To achieve the above purpose, an inductor-based equipment monitoring and protection method is proposed, comprising the following steps:

[0007] Step 1: Collect historical operating data of several test inductors;

[0008] Step 2: Based on the historical operation data, construct the operation status sample data, fault type sample data and operation life sample data;

[0009] Step 3: training a fault judgment model based on the running status sample data, training a fault type judgment model based on the fault type sample data, and training an running life prediction model based on the running life sample data;

[0010] Step 4: For the inductor to be monitored, actual operation data is collected in real time, and based on the actual operation data and the fault judgment model, it is determined whether the inductor is faulty. If a fault occurs, go to step 5, otherwise continue to execute step 4;

[0011] Step 5: Based on the actual operation data and the fault type judgment model, determine the fault type and issue a fault warning;

[0012] Step 6: Based on all actual operation data and fault types of the monitored inductor during operation, actual operation life data is formed; based on the actual operation life data and the operation life prediction model, the expected life of the monitored inductor is obtained; and a scrap warning is generated based on the expected life;

[0013] The collecting of historical operating data of a plurality of test inductors comprises the following steps:

[0014] Step 11: for each test inductor, collect a set of physical parameters per unit time of the test inductor during its operating life cycle; the physical parameters include electrical parameters and auxiliary parameters; all sets of physical parameters per unit time constitute historical operating characteristic data;

[0015] Step 12: For each tested inductor, a running status label is added to each unit time in the running record of the inductor, and the running status labels of all unit times constitute historical running label data;

[0016] Step 13: For each tested inductor, select all unit times whose operation status labels are not 0 as fault nodes; label each fault node with a fault type label, and the fault type labels of all unit times constitute historical fault label data;

[0017] Step 14: Collect the operating life label of each test sensor;

[0018] Step 15: Historical operating characteristic data, historical fault label data and operating life labels of all tested sensors;

[0019] The running status sample data is constructed in the following manner:

[0020] The electrical parameters and auxiliary parameters of each unit time are combined into a set of state input vectors;

[0021] The state input vector and the operation state label of each unit time are combined into the operation state sample data;

[0022] The fault type sample data is constructed in the following manner:

[0023] Preset response time t;

[0024] For each fault node of each test inductor, the time series of the response time t before the fault node is taken as a fault time series, and the parameter values ​​of each electrical parameter and auxiliary parameter per unit time in the fault time series are collected. The parameter value sequence of each electrical parameter and auxiliary parameter in the fault time series is taken as the parameter value sequence of the parameter, and the parameter value sequence of each parameter of each fault node constitutes a type time series set;

[0025] The type time series set of the fault nodes of all the tested inductors and the fault type labels constitute the fault type sample data;

[0026] The operation life sample data is constructed in the following manner:

[0027] Collect the expected life P of the inductor in advance;

[0028] Preset time window size w;

[0029] For each test sensor, the fault coefficient of each time window is collected every time window from the start of operation; the fault coefficient is the product of the operation status label and the fault type label, and a total of Failure coefficients constitute the life prediction sequence;

[0030] The life prediction sequence and the operating life label of each test sensor are combined into operating life sample data;

[0031] The method of training the fault judgment model based on the running state sample data, training the fault type judgment model based on the fault type sample data, and training the running life prediction model based on the running life sample data is as follows:

[0032] Each state input vector is used as the input of the fault judgment model, and the fault judgment model takes the predicted value of the running state of the state input vector corresponding to the unit time as the output, and the predicted value of the running state is 0 or 1; the fault judgment model takes the running state label corresponding to the unit time as the prediction target, takes the difference between the predicted value of the running state and the running state label as the first prediction error, and takes minimizing the sum of squares of the first prediction errors of all unit times as the training target; the fault judgment model is trained until the sum of squares of the first prediction errors reaches convergence, and the training is stopped; the fault judgment model is one of the classification models;

[0033] Each type time series set is used as the input of the operating life prediction model, and the fault type judgment model uses the predicted value of the fault type of the fault node corresponding to the type time series set as the output; the fault type judgment model uses the fault type label of the corresponding fault node as the prediction target, the difference between the predicted value of the fault type and the fault type label as the second prediction error, and minimizes the sum of squares of the second prediction errors of all unit times as the training target; the fault type judgment model is trained until the sum of squares of the second prediction errors reaches convergence and the training is stopped;

[0034] Each life prediction sequence is used as an input of an operating life prediction model, and the operating life prediction model uses the life prediction value of the test inductor corresponding to the life prediction sequence as an output; the operating life prediction model uses the operating life label of the corresponding test inductor as a prediction target, uses the difference between the life prediction value and the operating life label as a third prediction error, and uses minimizing the sum of squares of the third prediction errors per unit time as a training target; the operating life prediction model is trained until the sum of squares of the third prediction errors reaches convergence, and the training is stopped;

[0035] The method of generating a scrapping warning based on expected life is:

[0036] The actual operating life data of the inductor to be monitored is obtained at each time window. If the difference between the actual operating life data and the operating time of the inductor to be monitored is less than a preset scrap warning threshold, a scrap warning is initiated.

[0037] An inductor-based equipment monitoring and protection system is proposed, which includes a historical data collection module, a sample construction module, a model training module and an equipment monitoring module; wherein each module is electrically connected;

[0038] A historical data collection module, which collects historical operation data of a number of test inductors and sends the historical operation data to a sample construction module;

[0039] A sample construction module constructs operation status sample data, fault type sample data and operation life sample data based on historical operation data, and sends the operation status sample data, fault type sample data and operation life sample data to the model training module;

[0040] A model training module trains a fault judgment model based on the running status sample data, trains a fault type judgment model based on the fault type sample data, trains an operating life prediction model based on the operating life sample data, and sends the fault judgment model, the fault type judgment model and the operating life prediction model to the equipment monitoring module;

[0041] The equipment monitoring module collects the actual operation data of the inductor to be monitored in real time, and judges whether the inductor is faulty based on the actual operation data and the fault judgment model. If a fault occurs, the fault type is judged based on the actual operation data and the fault type judgment model, and a fault warning is issued. The actual operation life data is constructed based on all the actual operation data and fault types of the inductor to be monitored during operation. Based on the actual operation life data and the operation life prediction model, the expected life of the inductor to be monitored is obtained, and a scrap warning is generated based on the expected life.

[0042] An electronic device comprises: a processor and a memory, wherein the memory stores a computer program that can be called by the processor; the processor executes the above-mentioned inductor-based equipment monitoring and protection method by calling the computer program stored in the memory.

[0043] A computer-readable storage medium stores instructions. When the instructions are executed on a computer, the computer is enabled to execute the above-mentioned inductor-based equipment monitoring and protection method.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] First, a large number of inductors running on site are collected as test samples to obtain their electrical parameters, auxiliary state parameters and fault maintenance record data of the whole life cycle; then, based on these historical operation data, operating state sample data, fault type sample data and operating life sample data set are constructed; then, the fault judgment model, fault type judgment model and life prediction model are trained; finally, the inductors actually running are monitored online, and fault warning, fault diagnosis and remaining life evaluation are carried out based on the model output to provide a decision-making basis for equipment maintenance. This method realizes the all-round monitoring and life cycle management of inductor status through data, modeling and intelligent means. 1) Based on big data and machine learning algorithms, accurate judgment of fault status, type and remaining life is achieved, avoiding the defects of subjective judgment; 2) Multi-dimensional time series data modeling is adopted to capture the dynamic changes in the operation process of the inductor, improve the timeliness and accuracy of monitoring; 3) Quantitative evaluation of equipment life is carried out to provide a scientific decision-making basis for maintenance and repair, and realize the full-cycle intelligent management of status and life; 4) No frequent manual inspection is required, saving labor costs and improving work efficiency. Thereby greatly improving the safety and reliability level of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is a flow chart of an inductor-based equipment monitoring and protection method in Embodiment 1 of the present invention;

[0047] Figure 2This is a module connection relationship diagram of an inductor-based equipment monitoring and protection system in Embodiment 2 of the present invention;

[0048] Figure 3 This is a schematic diagram of the structure of an electronic device in Embodiment 3 of the present invention;

[0049] Figure 4 This is a schematic diagram of the computer-readable storage medium structure in Example 4 of the present invention. DETAILED DESCRIPTION

[0050] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only 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 ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] Embodiment 1, as Figure 1 As shown, an inductor-based equipment monitoring and protection method includes the following steps:

[0052] Step 1: Collect historical operating data of several test inductors;

[0053] Step 2: Based on the historical operation data, construct the operation status sample data, fault type sample data and operation life sample data;

[0054] Step 3: training a fault judgment model based on the running status sample data, training a fault type judgment model based on the fault type sample data, and training an running life prediction model based on the running life sample data;

[0055] Step 4: For the inductor to be monitored, actual operation data is collected in real time, and based on the actual operation data and the fault judgment model, it is determined whether the inductor is faulty. If a fault occurs, go to step 5, otherwise continue to execute step 4;

[0056] Step 5: Based on the actual operation data and the fault type judgment model, determine the fault type and issue a fault warning;

[0057] Step 6: Based on all actual operation data and fault types of the monitored inductor during operation, actual operation life data is formed; based on the actual operation life data and the operation life prediction model, the expected life of the monitored inductor is obtained; and a scrap warning is generated based on the expected life;

[0058] Specifically, the test inductors are a number of specific inductors with measuring sensors pre-selected by the tester, which are used to collect their long-term operation data from the beginning of operation to the scrapping period, so as to provide data support for the establishment of the algorithm model;

[0059] Specifically, the collecting of historical operating data of a plurality of test inductors comprises the following steps:

[0060] Step 11: for each test inductor, collect a set of physical parameters per unit time of the test inductor during its operating life cycle; the physical parameters include electrical parameters and auxiliary parameters; all sets of physical parameters per unit time constitute historical operating characteristic data;

[0061] Specifically, the electrical parameters include but are not limited to excitation current, voltage, power, inductance, winding resistance, insulation resistance, etc.; the auxiliary parameters include but are not limited to temperature, vibration and rotation speed;

[0062] Step 12: For each tested inductor, a running status label is added to each unit time in the running record of the inductor, and the running status labels of all unit times constitute historical running label data;

[0063] Preferably, the operating state label is 0, 1 or 2, 0 means that the inductor is in a normal operating state in this unit time;

[0064] 1 means that the inductor has a minor fault or failure in this unit time;

[0065] 2 means that the inductor has serious fault or failure in this unit time;

[0066] It is understandable that the operating status label may be a label manually set by a tester in real time according to the operating status of the test inductor;

[0067] Step 13: For each tested inductor, select all unit times whose operation status labels are not 0 as fault nodes; label each fault node with a fault type label, and the fault type labels of all unit times constitute historical fault label data;

[0068] Specifically, the fault type label takes a value of 0, 1, 2...N, where N is all possible fault types of the tested inductor, including but not limited to winding open circuit, winding short circuit, core damage, insulation aging, dielectric leakage, corrosion, and mechanical damage;

[0069] Step 14: Collect the operating life label of each test sensor;

[0070] Step 15: Historical operating characteristic data, historical fault label data and operating life labels of all tested sensors;

[0071] Specifically, the operation life label is collected in the following manner:

[0072] The operating life label is obtained by calculating the time point when the test inductor is finally scrapped minus the time point when the inductor is put into use;

[0073] It is understandable that during the operation of the inductor, any failure or abnormal event that occurs will affect the final life of the inductor. Therefore, since different inductors have different types of failures and different lifespans, the scrapping node of each inductor is different, and targeted scrapping warnings are required;

[0074] Furthermore, the running status sample data is constructed in the following manner:

[0075] The electrical parameters and auxiliary parameters of each unit time are combined into a set of state input vectors;

[0076] The state input vector and the operation state label of each unit time are combined into the operation state sample data;

[0077] It is understandable that when different fault types occur in the inductor, its electrical performance and temperature performance over a period of time will be different. For example, when the winding is open-circuited: the inductance decreases or approaches 0, the winding resistance increases, and the excitation current decreases; when the winding is short-circuited: the inductance decreases, the excitation current increases, the heat generation increases, and the temperature rise is obvious; when the core is damaged: the inductance decreases, the excitation current fluctuates, and the winding resistance changes little; when the insulation ages: the insulation resistance decreases, the leakage current increases, and the heat generation increases; therefore, the timing judgment model can be used to classify the fault types;

[0078] The fault type sample data is constructed in the following manner:

[0079] Preset response time t;

[0080] For each fault node of each test inductor, the time series of the response time t before the fault node is taken as a fault time series, and the parameter values ​​of each electrical parameter and auxiliary parameter per unit time in the fault time series are collected. The parameter value sequence of each electrical parameter and auxiliary parameter in the fault time series is taken as the parameter value sequence of the parameter, and the parameter value sequence of each parameter of each fault node constitutes a type time series set;

[0081] The type time series set of the fault nodes of all the tested inductors and the fault type labels constitute the fault type sample data;

[0082] The operation life sample data is constructed in the following manner:

[0083] Collect the expected life P of the inductor in advance; it is understandable that the expected life may be a life parameter in a laboratory provided by the manufacturer, and when the inductor does not fail during operation, its theoretical life should be similar to the expected life;

[0084] A preset time window size w; the time window size may be a number of time units, such as 60 time units or 120 time units;

[0085] For each test sensor, the fault coefficient of each time window is collected every time window from the start of operation; the fault coefficient is the product of the operation status label and the fault type label, and a total of failure coefficients to form a life prediction sequence; it is understandable that when the actual life of the test sensor is less than the expected life, no failure coefficients can be generated during its operation cycle. If there are fault coefficients, the fault coefficients of the missing parts are set to 0;

[0086] For example, When it is 5, only four fault coefficients 0, 1, 0, 0 are generated in the operation cycle of a test sensor, and the life prediction sequence of the test sensor is [0, 1, 0, 0, 0];

[0087] The life prediction sequence and the operating life label of each test sensor are combined into operating life sample data;

[0088] It is understandable that when constructing the life prediction sequence, the operation data is segmented into time windows, which preserves the temporal continuity and order of the order and time interval of the faults. This time-based feature extraction method can better capture the aging and performance degradation trend of the inductor over time;

[0089] Furthermore, the method of training the fault judgment model based on the running state sample data, training the fault type judgment model based on the fault type sample data, and training the running life prediction model based on the running life sample data is as follows:

[0090] Each state input vector is used as the input of the fault judgment model, and the fault judgment model takes the predicted value of the running state of the state input vector corresponding to the unit time as the output, and the predicted value of the running state is 0 or 1; the fault judgment model takes the running state label corresponding to the unit time as the prediction target, takes the difference between the predicted value of the running state and the running state label as the first prediction error, and takes minimizing the sum of squares of the first prediction errors of all unit times as the training target; the fault judgment model is trained until the sum of squares of the first prediction errors reaches convergence, and the training is stopped; the fault judgment model is one of the classification models; the classification model is a DNN model, a SVM model, etc.;

[0091] Each type of time series set is used as the input of the operating life prediction model, and the fault type judgment model uses the predicted value of the fault type of the fault node corresponding to the type of time series set as the output; the fault type judgment model uses the fault type label of the corresponding fault node as the prediction target, the difference between the predicted value of the fault type and the fault type label as the second prediction error, and minimizes the sum of squares of the second prediction errors of all unit times as the training target; the fault type judgment model is trained until the sum of squares of the second prediction errors reaches convergence and the training is stopped; the fault type judgment model is one of the time series models; the time series model is an RNN model, an LSTM model, etc.;

[0092] Each life prediction sequence is used as the input of the operating life prediction model, and the operating life prediction model takes the life prediction value of the test inductor corresponding to the life prediction sequence as the output; the operating life prediction model takes the operating life label of the corresponding test inductor as the prediction target, takes the difference between the life prediction value and the operating life label as the third prediction error, and takes minimizing the sum of squares of the third prediction errors per unit time as the training target; the operating life prediction model is trained until the sum of squares of the third prediction error reaches convergence, and the training is stopped; the operating life prediction model is one of the time series models;

[0093] Furthermore, the actual operation data is a parameter value set of various electrical parameters and auxiliary parameters of the inductor to be monitored collected in real time;

[0094] The method of judging whether the inductor is faulty based on the actual operation data and the fault judgment model is as follows:

[0095] For each unit time, the parameter value set of the unit time is used as the input of the fault judgment model to obtain the predicted value of the operating state output by the fault judgment model. If the predicted value of the operating state is 0, it is judged that no fault occurs, otherwise, it is judged that a fault occurs;

[0096] The fault type is determined based on the actual operation data and the fault type judgment model, and a fault warning is performed:

[0097] For each unit time, based on the parameter value set of the inductor to be monitored in the previous t unit time, a corresponding type time series set is constructed, and the type time series set is input into the fault type judgment model to obtain the predicted value of the fault type output by the fault type judgment model;

[0098] The fault warning method may be a voice reminder, a network message reminder, or a background screen pop-up window.

[0099] The method of forming the actual operation life data based on all actual operation data and fault types of the inductor to be monitored during operation is as follows:

[0100] At each time window, the fault coefficient of the electrical device to be monitored is obtained and filled in by adding 0 Failure coefficients constitute the actual operating life data;

[0101] The method of generating a scrapping warning based on expected life is:

[0102] At each time window, the actual operating life data of the inductor to be monitored is obtained. If the difference between the actual operating life data and the operating time of the inductor to be monitored is less than a preset scrap warning threshold, a scrap warning is initiated;

[0103] The scrapping warning may be provided in the form of a voice reminder, a network message reminder, or a background screen pop-up window.

[0104] Embodiment 2, as Figure 2 As shown, an inductor-based equipment monitoring and protection system includes a historical data collection module, a sample construction module, a model training module, and an equipment monitoring module; wherein each module is electrically connected;

[0105] A historical data collection module, which collects historical operation data of a number of test inductors and sends the historical operation data to a sample construction module;

[0106] A sample construction module constructs operation status sample data, fault type sample data and operation life sample data based on historical operation data, and sends the operation status sample data, fault type sample data and operation life sample data to the model training module;

[0107] A model training module trains a fault judgment model based on the running status sample data, trains a fault type judgment model based on the fault type sample data, trains an operating life prediction model based on the operating life sample data, and sends the fault judgment model, the fault type judgment model and the operating life prediction model to the equipment monitoring module;

[0108] The equipment monitoring module collects the actual operation data of the inductor to be monitored in real time, and judges whether the inductor is faulty based on the actual operation data and the fault judgment model. If a fault occurs, the fault type is judged based on the actual operation data and the fault type judgment model, and a fault warning is issued. The actual operation life data is constructed based on all the actual operation data and fault types of the inductor to be monitored during operation. Based on the actual operation life data and the operation life prediction model, the expected life of the inductor to be monitored is obtained, and a scrap warning is generated based on the expected life.

[0109] Example 3, please refer to Figure 3 As shown, according to another aspect of the present application, an electronic device 500 is also provided. The electronic device 500 may include one or more processors and one or more memories. The memories store computer readable codes, and when the computer readable codes are run by one or more processors, the above-mentioned device monitoring and protection method based on inductors can be executed.

[0110] The method or system according to the embodiment of the present application can also be used by Figure 3 The electronic device architecture shown in FIG. Figure 3 As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to a network, an input / output component 506, a hard disk 507, etc. The storage device in the electronic device 500, such as the ROM 503 or the hard disk 507, may store an inductor-based device monitoring and protection method and system provided by the present application. Furthermore, the electronic device 500 may also include a user interface 508. Of course, Figure 3 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 3 One or more components of an electronic device are shown.

[0111] Example 4, please refer to Figure 4 As shown, a computer-readable storage medium 600 according to an embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are executed by the processor, an inductor-based equipment monitoring and protection method and system according to an embodiment of the present application described with reference to the above figures can be executed. The storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache). Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0112] In addition, according to the embodiments of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the present application provides a non-transitory machine-readable storage medium, which stores machine-readable instructions, and the machine-readable instructions can be executed by a processor to execute instructions corresponding to the method steps provided by the present application;

[0113] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center through a wired network or a wireless network. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0114] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0115] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0116] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only in one form, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0117] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0118] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0119] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An inductor-based equipment monitoring and protection method, characterized in that: The following steps are involved: Step 1: Collect historical operating data of several test inductors; Step 2: Based on the historical operation data, construct the operation status sample data, fault type sample data and operation life sample data; Step 3: training a fault judgment model based on the running status sample data, training a fault type judgment model based on the fault type sample data, and training an running life prediction model based on the running life sample data; Step 4: For the inductor to be monitored, actual operation data is collected in real time, and based on the actual operation data and the fault judgment model, it is determined whether the inductor is faulty. If a fault occurs, go to step 5, otherwise continue to execute step 4; Step 5: Based on the actual operation data and the fault type judgment model, determine the fault type and issue a fault warning; Step 6: Based on all actual operation data and fault types of the monitored inductor during operation, actual operation life data is formed; based on the actual operation life data and the operation life prediction model, the expected life of the monitored inductor is obtained; and a scrap warning is generated based on the expected life; The collecting of historical operating data of a plurality of test inductors comprises the following steps: Step 11: for each test inductor, collect a set of physical parameters per unit time of the test inductor during its operating life cycle; the physical parameters include electrical parameters and auxiliary parameters; all sets of physical parameters per unit time constitute historical operating characteristic data; Step 12: For each tested inductor, a running status label is added to each unit time in the running record of the inductor, and the running status labels of all unit times constitute historical running label data; Step 13: For each tested inductor, select all unit times whose operation status labels are not 0 as fault nodes; label each fault node with a fault type label, and the fault type labels of all unit times constitute historical fault label data; Step 14: Collect the operating life label of each test sensor; Step 15: Historical operating characteristic data, historical fault label data and operating life labels of all tested sensors; The running status sample data is constructed in the following manner: The electrical parameters and auxiliary parameters of each unit time are combined into a set of state input vectors; The state input vector and the operation state label of each unit time are combined into the operation state sample data; The fault type sample data is constructed in the following manner: Preset response time t; For each fault node of each test inductor, the time series of the response time t before the fault node is taken as a fault time series, and the parameter values ​​of each electrical parameter and auxiliary parameter per unit time in the fault time series are collected. The parameter value sequence of each electrical parameter and auxiliary parameter in the fault time series is taken as the parameter value sequence of the parameter, and the parameter value sequence of each parameter of each fault node constitutes a type time series set; The type time series set of the fault nodes of all the tested inductors and the fault type labels constitute the fault type sample data; The operation life sample data is constructed in the following manner: Collect the expected life P of the inductor in advance; Preset time window size w; For each test sensor, the fault coefficient of each time window is collected every time window from the start of operation; the fault coefficient is the product of the operation status label and the fault type label, and a total of Failure coefficients constitute the life prediction sequence; The life prediction sequence and operating life label of each test sensor are combined into operating life sample data.

2. The device monitoring and protection method based on an inductor according to claim 1 is characterized in that: The method of training the fault judgment model based on the running state sample data, training the fault type judgment model based on the fault type sample data, and training the running life prediction model based on the running life sample data is as follows: Each state input vector is used as an input of a fault judgment model, and the fault judgment model outputs a predicted value of the operating state per unit time corresponding to the state input vector, and the predicted value of the operating state is 0 or 1; The fault judgment model uses the running state label corresponding to the unit time as the prediction target, the difference between the predicted value of the running state and the running state label as the first prediction error, and minimizes the sum of squares of the first prediction errors of all unit times as the training target; Training the fault judgment model until the sum of squares of the first prediction error reaches convergence and then stops the training; the fault judgment model is one of the classification models; Each type of time series set is used as the input of the operating life prediction model, and the fault type judgment model uses the predicted value of the fault type of the fault node corresponding to the type of time series set as the output; The fault type judgment model uses the fault type label of the corresponding fault node as the prediction target, uses the difference between the predicted value of the fault type and the fault type label as the second prediction error, and uses minimizing the sum of squares of the second prediction errors of all unit times as the training target; The fault type judgment model is trained until the sum of squares of the second prediction error reaches convergence and the training is stopped; Each life prediction sequence is used as an input of an operating life prediction model, and the operating life prediction model uses the life prediction value of the test inductor corresponding to the life prediction sequence as an output; the operating life prediction model uses the operating life label of the corresponding test inductor as a prediction target, uses the difference between the life prediction value and the operating life label as a third prediction error, and uses minimizing the sum of squares of the third prediction errors of all unit times as a training target; The running life prediction model is trained until the sum of squares of the third prediction error reaches convergence and then the training is stopped.

3. The device monitoring and protection method based on an inductor according to claim 2 is characterized in that: The method of generating a scrapping warning based on expected life is: The actual operating life data of the inductor to be monitored is obtained at each time window. If the difference between the actual operating life data and the operating time of the inductor to be monitored is less than a preset scrap warning threshold, a scrap warning is initiated.

4. An inductor-based equipment monitoring and protection system, which is used to implement the inductor-based equipment monitoring and protection method according to any one of claims 1 to 3, characterized in that: It includes a historical data collection module, a sample construction module, a model training module and an equipment monitoring module; wherein each module is electrically connected to each other; A historical data collection module, which collects historical operation data of a number of test inductors and sends the historical operation data to a sample construction module; A sample construction module constructs operation status sample data, fault type sample data and operation life sample data based on historical operation data, and sends the operation status sample data, fault type sample data and operation life sample data to the model training module; A model training module trains a fault judgment model based on the running status sample data, trains a fault type judgment model based on the fault type sample data, trains an operating life prediction model based on the operating life sample data, and sends the fault judgment model, the fault type judgment model and the operating life prediction model to the equipment monitoring module; The equipment monitoring module collects the actual operation data of the inductor to be monitored in real time, and judges whether the inductor is faulty based on the actual operation data and the fault judgment model. If a fault occurs, the fault type is judged based on the actual operation data and the fault type judgment model, and a fault warning is issued. The actual operation life data is constructed based on all the actual operation data and fault types of the inductor to be monitored during operation. Based on the actual operation life data and the operation life prediction model, the expected life of the inductor to be monitored is obtained, and a scrap warning is generated based on the expected life.

5. An electronic device, characterized in that: include: A processor and a memory, wherein the memory stores a computer program that can be called by the processor; The processor executes the inductor-based equipment monitoring and protection method according to any one of claims 1 to 3 by calling the computer program stored in the memory.

6. A computer-readable storage medium, characterized in that: Instructions are stored, and when the instructions are run on a computer, the computer is caused to execute an inductor-based equipment monitoring and protection method and system as described in any one of claims 1 to 3.

Images