A method and apparatus for alarming electrical equipment

Abstract

This invention provides a method and apparatus for power equipment alarm. The method includes: acquiring real-time measurement values ​​of the monitoring indicators according to set monitoring indicators; calculating state estimates of the monitoring indicators based on the real-time measurement values ​​and an optimal memory matrix dynamically updated based on historical measurement values ​​of the monitoring indicators, thereby generating a state estimation matrix; and generating a power equipment alarm result based on the state estimation matrix and a fault alarm threshold trust domain matrix dynamically updated based on historical measurement values ​​of the monitoring indicators. The method and apparatus effectively ensure the accuracy of the state estimates of the monitoring indicators calculated based on the optimal memory matrix, as well as the accuracy of determining the power equipment alarm result based on the state estimates. Furthermore, by establishing matrices for calculation of multiple monitoring indicators, the efficiency of calculating state estimates and power alarm results is greatly improved.

Description

Technical Field

[0001] This invention relates to the field of power equipment monitoring, and more specifically, to a power equipment alarm method and apparatus. Background Technology

[0002] Digital remote-reading meters for substations are intelligent monitoring instruments that enable local monitoring of primary equipment status parameters and remote data upload. They primarily encompass six types of meters: SF6 digital density meters, surge arrester digital leakage current meters, transformer digital oil thermometers, transformer digital oil level gauges, transformer digital gas relays, and instrument transformer digital oil pressure gauges. These meters utilize digital and communication technologies to monitor various non-electrical parameters (such as gas density, gas pressure, oil temperature, oil level, gas volume, and number of actuations) and electrical parameters (such as leakage current) in real time. Through digital data upload, monitoring data can be transmitted to cloud platforms or other management systems, enabling remote monitoring and management of substation equipment. Data processing and analysis can also facilitate alarm functions for power equipment.

[0003] These digital remote meters are widely used in critical equipment in substations, such as transformers, GIS (gas-insulated switchgear), surge arresters, and instrument transformers, enabling real-time monitoring of equipment status and data upload. Maintenance personnel can use this data to monitor equipment operation at any time, promptly identify faults and anomalies, and take measures to improve equipment stability and safety. However, current fault and anomaly alarms for power equipment typically rely on simple threshold alarms and rate-of-change alarms. These methods struggle to accurately and efficiently extract effective information from monitoring data, resulting in poor accuracy and efficiency in alarm reporting. Summary of the Invention

[0004] To address the problem that existing technologies for alarming substation power equipment faults and anomalies rely solely on threshold alarms and rate of change alarms, resulting in poor accuracy and efficiency, this invention provides a power equipment alarm method and apparatus.

[0005] According to one aspect of the present invention, a power equipment alarm method is provided, the method comprising:

[0006] According to the set monitoring indicators, the real-time measurement values ​​of the monitoring indicators are obtained, wherein the monitoring indicators include the operating indicators of the substation digital meters and the substation environmental parameters;

[0007] Based on the real-time measurement values ​​and the optimal memory matrix that is dynamically updated based on the historical measurement values ​​of the monitoring indicators, the state estimate values ​​of the monitoring indicators are calculated, and a state estimate matrix is ​​generated.

[0008] Power equipment alarm results are generated based on the state estimation matrix and the fault alarm threshold trust domain matrix, which is dynamically updated based on the historical measurements of the monitoring indicators.

[0009] According to another aspect of the present invention, the present invention provides an alarm device for power equipment, the device comprising:

[0010] The data acquisition module is used to acquire the real-time measurement values ​​of the monitoring indicators according to the set monitoring indicators, wherein the monitoring indicators include the operating indicators of the substation digital meters and the substation environmental parameters;

[0011] The data calculation module is used to calculate the state estimate of the monitoring indicator and generate the state estimate matrix based on the real-time measurement value and the optimal memory matrix that is dynamically updated based on the historical measurement value of the monitoring indicator.

[0012] The result output module is used to generate power equipment alarm results based on the state estimation matrix and the fault alarm threshold trust domain matrix that is dynamically updated based on the historical measurement values ​​of the monitoring indicators.

[0013] According to another aspect of the present invention, a computer-readable storage medium is provided, the storage medium storing a computer program for performing the methods described in any of the above aspects of the present invention.

[0014] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: a processor; a memory for storing executable instructions of the processor; the processor being configured to read the executable instructions from the memory and execute the instructions to implement the method described in any of the preceding aspects of the present invention.

[0015] The present invention discloses a power equipment alarm method and apparatus, wherein the method includes: acquiring real-time measurement values ​​of the monitoring indicators according to set monitoring indicators; calculating state estimates of the monitoring indicators based on the real-time measurement values ​​and an optimal memory matrix dynamically updated based on historical measurement values ​​of the monitoring indicators, thereby generating a state estimation matrix; and generating a power equipment alarm result based on the state estimation matrix and a fault alarm threshold trust domain matrix dynamically updated based on historical measurement values ​​of the monitoring indicators. The method and apparatus dynamically update the optimal memory matrix of the monitoring indicator measurement values ​​based on historical data of the monitoring indicators for power equipment alarms, and dynamically update the alarm threshold trust domain based on historical data of the monitoring indicators, thereby effectively ensuring the accuracy of the state estimates of the monitoring indicators calculated based on the optimal memory matrix, and the accuracy of determining the power equipment alarm result based on the state estimates. Furthermore, by establishing matrices for calculation of multiple monitoring indicators, the efficiency of calculating state estimates and power alarm results is greatly improved. Attached Figure Description

[0016] Exemplary embodiments of the present invention can be more fully understood by referring to the following figures:

[0017] Figure 1 A flowchart of a power equipment alarm method according to a preferred embodiment of the present invention;

[0018] Figure 2 A flowchart of a power equipment alarm method according to a preferred embodiment of the present invention;

[0019] Figure 3 This is a schematic diagram of the structure of an electronic device according to a preferred embodiment of the present invention. Detailed Implementation

[0020] Exemplary embodiments of the invention will now be described with reference to the accompanying drawings. However, the invention may be embodied in many different forms and is not limited to the embodiments described herein. These embodiments are provided to fully and completely disclose the invention and to fully convey its scope to those skilled in the art. The terminology used in the exemplary embodiments illustrated in the drawings is not intended to limit the invention. In the drawings, the same units / elements are referred to by the same reference numerals.

[0021] Unless otherwise stated, the terms used herein (including technical terms) have their common meaning as understood by one of ordinary skill in the art. Furthermore, it is understood that terms defined in commonly used dictionaries should be understood to have a meaning consistent with the context of their relevant field, and not to be interpreted as having an idealized or overly formal meaning.

[0022] Exemplary methods

[0023] Figure 1 This is a flowchart of a power equipment alarm method according to a preferred embodiment of the present invention. Figure 1 As shown, the power equipment alarm method of this preferred embodiment starts from step 101.

[0024] In step 101, the real-time measurement values ​​of the monitoring indicators are obtained according to the set monitoring indicators, wherein the monitoring indicators include the operating indicators of the substation digital meters and the substation environmental parameters.

[0025] In this preferred embodiment, the substation's digital meters collect meter operation data and environmental parameters in real time, and transmit various non-electrical parameters (such as gas density, gas pressure, oil temperature, oil level, gas volume, number of actuations, etc.) and electrical parameters (such as leakage current) to the substation monitoring host (or auxiliary control host and other monitoring data collection and aggregation system). The monitored power equipment includes: GIS, transformers, surge arresters, instrument transformers, etc.

[0026] Preferably, before generating the state estimation matrix, the method further includes: calculating the state estimate of the monitoring indicator based on the value of the indicator to be evaluated and the optimal memory matrix dynamically updated based on the historical measurement values ​​of the monitoring indicator.

[0027] When there are n monitoring indicators, the historical measurement values ​​of the n monitoring indicators at the sampling time are denoted as the observation vector X = [x1, x2, ..., xn]. n ] T ;

[0028] For multiple sensors that collect measurements of n monitoring indicators, select the n sensors with high trust scores to obtain historical measurement values ​​of the n monitoring indicators at m historical sampling times, generate m observation vectors, and generate an initial memory matrix D(T) based on the m observation vectors. The trust score is determined by the score in a pre-set scoring table based on expert experience, and T is the sampling period consisting of m sampling times.

[0029] The initial memory matrix D(T) is calculated using the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method to generate the corresponding optimized memory matrix. Specifically, the sliding window method is used to limit the time range of historical measurement values, the exponentially weighted moving average method is used to smooth the influence of historical measurement values ​​at different sampling times, the Kalman filter is used to predict and correct the historical measurement values ​​of the monitoring indicators, and the initial memory matrix is ​​adjusted in real time through incremental learning.

[0030] The optimal memory matrix is ​​generated based on the initial memory matrix and the optimized memory matrix.

[0031] In this preferred embodiment, since multiple sensors may collect data for a single monitoring indicator in a substation, to improve the accuracy of data acquisition, a confidence score is assigned to each sensor based on its characteristics, prioritizing sensors with higher confidence scores for data acquisition. Simultaneously, considering changes in operating conditions and the impact of data acquisition time on accuracy, historical data is continuously and dynamically updated to ensure the accuracy of power equipment status estimates calculated from real-time acquired data. This preferred embodiment employs a combination of four methods—sliding window update, exponentially weighted moving average, Kalman filtering, and incremental learning—to construct the optimal memory matrix for historical data. Specifically, a sliding window is used to limit the time range of the data, and an exponentially weighted moving average is used to smooth the influence of historical data (the exponentially weighted moving average takes a memory matrix generated based on high-confidence historical data as input, and then assigns a weight value to several historical memory matrices preceding this memory matrix and the memory matrix itself to obtain a new memory matrix. This method emphasizes the importance of the current sample and gradually de-emphasizes the importance of historical samples. The more recent the data is, the greater the weight, and vice versa. This can be set based on experience). Kalman filtering is used to dynamically predict and correct the measured values, and incremental learning is used to adjust the model in real time.

[0032] Preferably, the step of generating an optimal memory matrix based on the initial memory matrix and the optimized memory matrix is ​​wherein the formula for calculating the optimal memory matrix is:

[0033] D H (T+1)=γ1(T+1)*D w (T+1)+γ2(T+1)*D e (T+1)

[0034] +γ3(T+1)*D k (T+1)+γ4(T+1)*D l (T+1)

[0035]

[0036] In the formula, D H (T+1) is the optimal memory matrix for sampling period T+1, D w (T+1), D e (T+1), D k (T+1) and D l(T+1) represents the optimized memory matrix generated using the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method, respectively. γ1(T+1), γ2(T+1), γ3(T+1), and γ4(T+1) are the weight coefficients of the optimized memory matrices generated using the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method in calculating the optimal memory matrix, respectively. j (T) represents the value of the j-th column of the initial memory matrix D(T) at the sampling period T, where the values ​​of i represent the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method, respectively. This represents the value of the j-th column at sampling period T+1 when using the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method, respectively.

[0037] As shown in the formula for calculating the weight coefficients, the weight values ​​are dynamically adjusted based on the errors of the memory matrices obtained from the four methods. Methods with smaller errors will receive greater weights in their memory matrices. By dynamically adjusting these weight coefficients, the optimal memory matrix can adaptively meet the update requirements of the state estimates.

[0038] In step 102, the state estimate of the monitoring indicator is calculated based on the real-time measurement value and the optimal memory matrix that is dynamically updated based on the historical measurement value of the monitoring indicator, and a state estimate matrix is ​​generated.

[0039] In step 103, power equipment alarm results are generated based on the state estimation matrix and the fault alarm threshold trust domain matrix that is dynamically updated based on the historical measurement values ​​of the monitoring indicators.

[0040] Preferably, before generating the power equipment alarm result based on the state estimation matrix and the fault alarm threshold trust domain matrix dynamically updated based on historical measurements of the monitoring indicators, the method further includes:

[0041] Let X be an observation vector consisting of n monitoring indicators, and let f(X) be the distribution function of the fault alarm threshold of X. Using a Gaussian process model g(X) to approximate the objective function, the posterior distribution of the Gaussian process model g(X) is:

[0042] g(X)~N(μ(X),σ 2 (X))

[0043] In the formula, μ(X) is the mean function, σ 2 (X) is the covariance function, representing the predicted value and uncertainty of the objective function f(X), respectively. Its calculation formula is:

[0044] μ(X)=k(X,I)K -1 Y

[0045] σ 2 (X)=k(X,X)-k(X,I)K -1 k(I,X)

[0046] In the formula, k(X,I), k(X,X) and k(X,I) are kernel functions, K is the covariance matrix, I is the input matrix of the measured values ​​of the observation vector X, and Y is the observation matrix of the objective function f(X) obtained from the input matrix I.

[0047] The measured values ​​of the observation vector X obtained from the historical time period T and the observed values ​​of the corresponding objective function f(X) are used as training data to determine the expression of the Gaussian process model g(X);

[0048] The initial trust domain T1 of the fault alarm threshold trust domain matrix for the historical time period T is set as follows:

[0049]

[0050] In the formula, n is the dimension of the monitoring indicator, f(X0) is the initial search point of the objective function f(X), and Δ1 is the initial radius of the fault alarm threshold trust threshold.

[0051] Based on the expressions for the objective function f(X) and the Gaussian process model g(X), the observation vector X at sampling time t+(k+1)Δ is calculated. t Objective function observations and Gaussian process model prediction values Where, sampling time t is the last sampling time of the historical time period T, Δ t The sampling time interval is k, a natural number, with an initial value of 1.

[0052] Based on the observed values ​​of the objective function and Gaussian process model predictions and Calculate the ratio ρ between the actual improvement and the predicted improvement of the observed vector. t The calculation formula is as follows:

[0053]

[0054] According to the ratio The fault alarm threshold combination matrix is ​​dynamically updated using the initial trust field T1 of the matrix, where:

[0055] when At that time, expand the trust domain, Δ k+1 =γ e Δ k ;

[0056] when At that time, the trust domain remains unchanged, Δ k+1 =Δ k ;

[0057] when At that time, shrink the trust domain, Δ k+1 =γ s Δ k ;

[0058] In the formula, μ1 and μ2 are the thresholds for judging the trust domain update, and γ e γ is the magnification factor. e >1, γ s γ is the shrinkage coefficient, 0 < γ s <1.

[0059] In this preferred embodiment, by dynamically updating the trust domain of the alarm threshold based on historical data, it can better adapt to the actual operating conditions of the equipment, improve the accuracy of fault identification, and reduce false alarms.

[0060] Preferably, generating power equipment alarm results based on the state estimation matrix and the fault alarm threshold trust domain matrix dynamically updated based on historical measurements of the monitoring indicators includes:

[0061] When the state estimate value of the monitoring indicator in the state estimation matrix is ​​located in the corresponding fault alarm threshold trust domain matrix, the alarm result of the monitoring indicator is determined to be in an alarm state; otherwise, it is in a non-alarm state.

[0062] Output the alarm results of the power equipment based on the alarm results of n monitoring indicators.

[0063] The power equipment alarm method described in this preferred embodiment continuously updates the optimal memory matrix of the monitoring indicators and the state estimate corresponding to the monitoring indicators based on the historical measurement values ​​of the set power equipment monitoring indicators. This ensures the mean regression and dynamic adjustment of the state estimate obtained based on the real-time measurement values ​​of the monitoring indicators and the trust domain of the alarm threshold, greatly improving the accuracy and efficiency of power equipment alarms.

[0064] Exemplary device

[0065] Figure 2 This is a schematic diagram of a power equipment alarm device according to a preferred embodiment of the present invention. Figure 2 As shown, the power equipment alarm device 200 of this preferred embodiment includes:

[0066] The data acquisition module 201 is used to acquire the real-time measurement values ​​of the monitoring indicators according to the set monitoring indicators, wherein the monitoring indicators include the operating indicators of the substation digital meters and the substation environmental parameters.

[0067] The data calculation module 202 is used to calculate the state estimate of the monitoring index and generate a state estimate matrix based on the real-time measurement value and the optimal memory matrix that is dynamically updated based on the historical measurement value of the monitoring index.

[0068] The result output module 203 is used to generate power equipment alarm results based on the state estimation matrix and the fault alarm threshold trust domain matrix that is dynamically updated based on the historical measurement values ​​of the monitoring indicators.

[0069] Preferably, the device further includes a first update module, used for:

[0070] When there are n monitoring indicators, the historical measurement values ​​of the n monitoring indicators at the sampling time are denoted as the observation vector X = [x1, x2, ..., xn]. n ] T ;

[0071] For multiple sensors that collect measurements of n monitoring indicators, select the n sensors with high trust scores to obtain historical measurement values ​​of the n monitoring indicators at m historical sampling times, generate m observation vectors, and generate an initial memory matrix D(T) based on the m observation vectors. The trust score is determined by the score in a pre-set scoring table based on expert experience, and T is the sampling period consisting of m sampling times.

[0072] The initial memory matrix D(T) is calculated using the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method to generate the corresponding optimized memory matrix. Specifically, the sliding window method is used to limit the time range of historical measurement values, the exponentially weighted moving average method is used to smooth the influence of historical measurement values ​​at different sampling times, the Kalman filter is used to predict and correct the historical measurement values ​​of the monitoring indicators, and the initial memory matrix is ​​adjusted in real time through incremental learning.

[0073] The optimal memory matrix is ​​generated based on the initial memory matrix and the optimized memory matrix.

[0074] Preferably, the first update module generates an optimal memory matrix based on the initial memory matrix and the optimized memory matrix, wherein the formula for calculating the optimal memory matrix is:

[0075] D H (T+1)=γ1(T+1)*D w (T+1)+γ2(T+1)*D e(T+1)

[0076] +γ3(T+1)*D k (T+1)+γ4(T+1)*D l (T+1)

[0077]

[0078] In the formula, D H (T+1) is the optimal memory matrix for sampling period T+1, D w (T+1), D e (T+1), D k (T+1) and D l (T+1) represents the optimized memory matrix generated using the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method, respectively. γ1(T+1), γ2(T+1), γ3(T+1), and γ4(T+1) are the weight coefficients of the optimized memory matrices generated using the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method in calculating the optimal memory matrix, respectively. j (T) represents the value of the j-th column of the initial memory matrix D(T) at the sampling period T, where the values ​​of i represent the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method, respectively. This represents the value of the j-th column at sampling period T+1 when using the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method, respectively.

[0079] Preferably, the device further includes a second update module for:

[0080] Let X be an observation vector consisting of n monitoring indicators, and let f(X) be the distribution function of the fault alarm threshold of X. Using a Gaussian process model g(X) to approximate the objective function, the posterior distribution of the Gaussian process model g(X) is:

[0081] g(X)~N(μ(X),σ 2 (X))

[0082] In the formula, μ(X) is the mean function, σ 2 (X) is the covariance function, representing the predicted value and uncertainty of the objective function f(X), respectively. Its calculation formula is:

[0083] μ(X)=k(X,I)K -1 Y

[0084] σ 2 (X)=k(X,X)-k(X,I)K -1 k(I,X)

[0085] In the formula, k(X,I), k(X,X) and k(X,I) are kernel functions, K is the covariance matrix, I is the input matrix of the measured values ​​of the observation vector X, and Y is the observation matrix of the objective function f(X) obtained from the input matrix I.

[0086] The measured values ​​of the observation vector X obtained from the historical time period T and the observed values ​​of the corresponding objective function f(X) are used as training data to determine the expression of the Gaussian process model g(X);

[0087] The initial trust domain T1 of the fault alarm threshold trust domain matrix for the historical time period T is set as follows:

[0088]

[0089] In the formula, n is the dimension of the monitoring indicator, f(X0) is the initial search point of the objective function f(X), and Δ1 is the initial radius of the fault alarm threshold trust threshold.

[0090] Based on the expressions for the objective function f(X) and the Gaussian process model g(X), the observation vector X at sampling time t+(k+1)Δ is calculated. t Objective function observations and Gaussian process model prediction values Where, sampling time t is the last sampling time of the historical time period T, Δ t The sampling time interval is k, a natural number, with an initial value of 1.

[0091] Based on the observed values ​​of the objective function and Gaussian process model predictions and Calculate the ratio ρ between the actual improvement and the predicted improvement of the observed vector. t The calculation formula is as follows:

[0092]

[0093] According to the ratio The fault alarm threshold combination matrix is ​​dynamically updated using the initial trust field T1 of the matrix, where:

[0094] when At that time, expand the trust domain, Δ k+1 =γ e Δ k ;

[0095] when At that time, the trust domain remains unchanged, Δ k+1 =Δ k ;

[0096] when At that time, shrink the trust domain, Δ k+1 =γ s Δ k ;

[0097] In the formula, u1 and μ2 are the thresholds for judging the trust domain update, and γ e γ is the magnification factor. e >1, γ s γ is the shrinkage coefficient, 0 < γ s <1.

[0098] Preferably, the result output module 203 generates power equipment alarm results based on the state estimation matrix and a fault alarm threshold trust domain matrix dynamically updated based on historical measurements of the monitoring indicators, including:

[0099] When the state estimate value of the monitoring indicator in the state estimation matrix is ​​located in the corresponding fault alarm threshold trust domain matrix, the alarm result of the monitoring indicator is determined to be in an alarm state; otherwise, it is in a non-alarm state.

[0100] Output the alarm results of the power equipment based on the alarm results of n monitoring indicators.

[0101] The power equipment alarm device described in this preferred embodiment collects real-time measured values ​​of monitoring indicators. The steps for power equipment alarm based on the continuously updated optimal memory matrix and alarm threshold trust domain matrix are the same as those of the power equipment alarm method described in this invention, and the achieved technical effects are also the same. Therefore, they will not be repeated here.

[0102] Exemplary electronic devices

[0103] Figure 3 This is a schematic diagram of an electronic device according to a preferred embodiment of the present invention. The electronic device may be either or both of a first device and a second device, or a standalone device independent of them. The standalone device may communicate with the first device and the second device to receive the collected input signals from them. Figure 3 A block diagram of an electronic device according to an embodiment of the present disclosure is shown. Figure 3 As shown, the electronic device includes one or more processors 301 and memory 302.

[0104] The processor 301 may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions.

[0105] The memory 302 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 301 may execute the program instructions to implement the energy consumption anomaly diagnosis method based on enterprise energy consumption space of the various embodiments disclosed above, and / or other desired functions. In one example, the electronic device may also include an input device 303 and an output device 304, these components being interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0106] In addition, the input device 303 may also include, for example, a keyboard, a mouse, etc.

[0107] The output device 304 can output various information to the outside. The output device 304 may include, for example, a display, a speaker, a printer, and a communication network and its connected remote output devices, etc.

[0108] Of course, for the sake of simplicity, Figure 3 Only some of the components of the electronic device relevant to this disclosure are shown, omitting components such as buses, input / output interfaces, etc. In addition, the electronic device may include any other suitable components depending on the specific application.

[0109] Exemplary computer program products and computer-readable storage media

[0110] In addition to the methods and apparatus described above, embodiments of this disclosure may also be computer program products comprising computer program instructions that, when executed by a processor, cause the processor to perform the steps in the power equipment alarm methods according to various embodiments of this disclosure as described in the "Exemplary Methods" section of this specification.

[0111] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this disclosure. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on a user's computing device, partially on a user's computing device, as a standalone software package, partially on a user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0112] Furthermore, embodiments of this disclosure may also be computer-readable storage media storing computer program instructions thereon, which, when executed by a processor, cause the processor to perform the steps in the power equipment alarm methods according to various embodiments of this disclosure as described in the "Exemplary Methods" section above.

[0113] The computer-readable storage medium may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may, for example, include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0114] The basic principles of this disclosure have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this disclosure are merely examples and not limitations, and should not be considered as essential features of each embodiment of this disclosure. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the scope of this disclosure to the necessity of employing the aforementioned specific details for implementation.

[0115] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For system embodiments, since they largely correspond to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0116] The block diagrams of devices, apparatuses, devices, and systems disclosed herein are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0117] The methods and apparatus of this disclosure may be implemented in many ways. For example, they may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above-described order of steps for the methods is for illustrative purposes only, and the steps of the methods of this disclosure are not limited to the order specifically described above unless otherwise specifically stated. Furthermore, in some embodiments, this disclosure may also be implemented as a program recorded on a recording medium, the program including machine-readable instructions for implementing the methods according to this disclosure. Thus, this disclosure also covers recording media storing programs for performing the methods according to this disclosure.

[0118] It should also be noted that in the apparatus, devices, and methods of this disclosure, the components or steps are decomposable and / or recombinable. Such decomposition and / or recombination should be considered equivalent to the present disclosure. The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this disclosure. Therefore, this disclosure is not intended to be limited to the aspects shown herein, but rather to be carried out within the widest scope consistent with the principles and novel features disclosed herein.

[0119] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this disclosure to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. A method for alarming electrical equipment, characterized in that, The method includes: According to the set monitoring indicators, the real-time measurement values ​​of the monitoring indicators are obtained, wherein the monitoring indicators include the operating indicators of the substation digital meters and the substation environmental parameters; Based on the real-time measurement values ​​and the optimal memory matrix that is dynamically updated based on the historical measurement values ​​of the monitoring indicators, the state estimate values ​​of the monitoring indicators are calculated, and a state estimate matrix is ​​generated. Let X be an observation vector consisting of n monitoring indicators, and let f(X) be the distribution function of the fault alarm threshold of X. Using a Gaussian process model g(X) to approximate the objective function, the posterior distribution of the Gaussian process model g(X) is: In the formula, It is a mean function. Let be the covariance function, and let and represent the predicted value and uncertainty of the objective function f(X), respectively. The formula for calculating is: In the formula, , and K is the kernel function, I is the covariance matrix, Y is the input matrix of the measured values ​​of the observation vector X, and Y is the observation matrix of the objective function f(X) obtained from the input matrix I. The measured values ​​of the observation vector X obtained from the historical time period T and the observed values ​​of the corresponding objective function f(X) are used as training data to determine the expression of the Gaussian process model g(X); Set the fault alarm threshold for the historical time period T and the initial trust domain of the trust domain matrix. Its expression is: In the formula, n represents the dimension of the monitoring indicator. For the objective function The initial search point, The initial radius of the fault alarm threshold trust threshold; Based on the expressions for the objective function f(X) and the Gaussian process model g(X), the observation vector X at the sampling time is calculated. The objective function observation value f( ) and the Gaussian process model predicted value g( ), where sampling time t is the last sampling time of the historical time period T, The sampling time interval is k, a natural number, with an initial value of 1. Based on the observed values ​​of the objective function f( ) and f( ), Gaussian process model predicted value g( ) and g( Calculate the ratio of the actual improvement to the predicted improvement in the observed vector. The calculation formula is as follows: According to the ratio Initial trust region of the combination matrix with fault alarm threshold The fault alarm threshold combination matrix is ​​dynamically updated, where: when At the same time, expand the trust domain. ; when At the same time, keep the trust domain unchanged. ; when At that time, shrink the trust domain. ; In the formula, and To determine the threshold for trust domain updates, For the expansion factor, , The shrinkage coefficient, ; Power equipment alarm results are generated based on the state estimation matrix and the fault alarm threshold trust domain matrix that is dynamically updated based on the historical measurements of the monitoring indicators.

2. The method according to claim 1, characterized in that, The step of calculating the state estimate of the monitoring indicator based on the dynamically updated optimal memory matrix based on the real-time measurement value and the historical measurement value of the monitoring indicator, before generating the state estimate matrix, further includes: When there are n monitoring indicators, the historical measurement values ​​of the n monitoring indicators at the sampling time are recorded as the observation vector. ; For multiple sensors that collect measurements of n monitoring indicators, select the n sensors with high trust scores to obtain historical measurement values ​​of the n monitoring indicators at m historical sampling times, generate m observation vectors, and generate an initial memory matrix based on the m observation vectors. The trust score is determined by the score in a pre-set scoring table based on expert experience, and T is the sampling period consisting of m sampling times. The initial memory matrix is ​​updated using a sliding window method, an exponentially weighted moving average method, a Kalman filter method, and an incremental learning method. Calculations are performed to generate a corresponding optimized memory matrix. A sliding window is used to limit the time range of historical measurements, an exponentially weighted moving average is used to smooth the influence of historical measurements at different sampling times, Kalman filtering is used to predict and correct the historical measurements of monitoring indicators, and the initial memory matrix is ​​adjusted in real time through incremental learning. The optimal memory matrix is ​​generated based on the initial memory matrix and the optimized memory matrix.

3. The method according to claim 2, characterized in that, The optimal memory matrix is ​​generated based on the initial memory matrix and the optimized memory matrix, wherein the formula for calculating the optimal memory matrix is: In the formula, The optimal memory matrix for sampling period T+1. , , and The optimized memory matrices are generated using the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method, respectively. , , and The weight coefficients of the optimized memory matrices generated using the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method are respectively used in calculating the optimal memory matrix. Represents the initial memory matrix In the j-th column of the sampling period T, the values ​​of i represent the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method, respectively. This represents the value of the j-th column at sampling period T+1 when using the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method, respectively.

4. The method according to claim 1, characterized in that, The process of generating power equipment alarm results based on the state estimation matrix and the fault alarm threshold trust domain matrix dynamically updated based on historical measurements of the monitoring indicators includes: When the state estimate value of the monitoring indicator in the state estimation matrix is ​​located in the corresponding fault alarm threshold trust domain matrix, the alarm result of the monitoring indicator is determined to be in an alarm state; otherwise, it is in a non-alarm state. Output the alarm results of the power equipment based on the alarm results of n monitoring indicators.

5. An alarm device for electrical equipment, characterized in that, The device includes: The data acquisition module is used to acquire the real-time measurement values ​​of the monitoring indicators according to the set monitoring indicators, wherein the monitoring indicators include the operating indicators of the substation digital meters and the substation environmental parameters; The data calculation module is used to calculate the state estimate of the monitoring indicator and generate the state estimate matrix based on the real-time measurement value and the optimal memory matrix that is dynamically updated based on the historical measurement value of the monitoring indicator. The second update module is used for: Let X be an observation vector consisting of n monitoring indicators, and let f(X) be the distribution function of the fault alarm threshold of X. Using a Gaussian process model g(X) to approximate the objective function, the posterior distribution of the Gaussian process model g(X) is: In the formula, It is a mean function. Let be the covariance function, and let and represent the predicted value and uncertainty of the objective function f(X), respectively. The formula for calculating is: In the formula, , and K is the kernel function, I is the covariance matrix, Y is the input matrix of the measured values ​​of the observation vector X, and Y is the observation matrix of the objective function f(X) obtained from the input matrix I. The measured values ​​of the observation vector X obtained from the historical time period T and the observed values ​​of the corresponding objective function f(X) are used as training data to determine the expression of the Gaussian process model g(X); Set the fault alarm threshold for the historical time period T and the initial trust domain of the trust domain matrix. Its expression is: In the formula, n represents the dimension of the monitoring indicator. For the objective function The initial search point, The initial radius of the fault alarm threshold trust threshold; Based on the expressions for the objective function f(X) and the Gaussian process model g(X), the observation vector X at the sampling time is calculated. The objective function observation value f( ) and the Gaussian process model predicted value g( ), where sampling time t is the last sampling time of the historical time period T, The sampling time interval is k, a natural number, with an initial value of 1. Based on the observed values ​​of the objective function f( ) and f( ), Gaussian process model predicted value g( ) and g( Calculate the ratio of the actual improvement to the predicted improvement in the observed vector. The calculation formula is as follows: According to the ratio Initial trust region of the combination matrix with fault alarm threshold The fault alarm threshold combination matrix is ​​dynamically updated, where: when At the same time, expand the trust domain. ; when At the same time, keep the trust domain unchanged. ; when At that time, shrink the trust domain. ; In the formula, and To determine the threshold for trust domain updates, For the expansion factor, , The shrinkage coefficient, ; The result output module is used to generate power equipment alarm results based on the state estimation matrix and the fault alarm threshold trust domain matrix that is dynamically updated based on the historical measurement values ​​of the monitoring indicators.

6. The apparatus according to claim 5, characterized in that, The device further includes a first update module for: When there are n monitoring indicators, the historical measurement values ​​of the n monitoring indicators at the sampling time are recorded as the observation vector. ; For multiple sensors that collect measurements of n monitoring indicators, select the n sensors with high trust scores to obtain historical measurement values ​​of the n monitoring indicators at m historical sampling times, generate m observation vectors, and generate an initial memory matrix based on the m observation vectors. The trust score is determined by the score in a pre-set scoring table based on expert experience, and T is the sampling period consisting of m sampling times. The initial memory matrix is ​​updated using a sliding window method, an exponentially weighted moving average method, a Kalman filter method, and an incremental learning method. Calculations are performed to generate a corresponding optimized memory matrix. A sliding window is used to limit the time range of historical measurements, an exponentially weighted moving average is used to smooth the influence of historical measurements at different sampling times, Kalman filtering is used to predict and correct the historical measurements of monitoring indicators, and the initial memory matrix is ​​adjusted in real time through incremental learning. The optimal memory matrix is ​​generated based on the initial memory matrix and the optimized memory matrix.

7. The apparatus according to claim 6, characterized in that, The first update module generates an optimal memory matrix based on the initial memory matrix and the optimized memory matrix, wherein the formula for calculating the optimal memory matrix is: In the formula, The optimal memory matrix for sampling period T+1. , , and The optimized memory matrices are generated using the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method, respectively. , , and The weight coefficients of the optimized memory matrices generated using the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method are respectively used in calculating the optimal memory matrix. Represents the initial memory matrix In the j-th column of the sampling period T, the values ​​of i represent the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method, respectively. This represents the value of the j-th column at sampling period T+1 when using the sliding window update method, the exponentially weighted moving average method, the Kalman filter method, and the incremental learning method, respectively.

8. The apparatus according to claim 5, characterized in that, The result output module generates power equipment alarm results based on the state estimation matrix and a fault alarm threshold trust domain matrix dynamically updated based on historical measurements of the monitoring indicators, including: When the state estimate value of the monitoring indicator in the state estimation matrix is ​​located in the corresponding fault alarm threshold trust domain matrix, the alarm result of the monitoring indicator is determined to be in an alarm state; otherwise, it is in a non-alarm state. Output the alarm results of the power equipment based on the alarm results of n monitoring indicators.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program for executing the method described in any one of claims 1 to 4.

10. An electronic device, characterized in that, The electronic device includes: processor; Memory used to store the processor's executable instructions; The processor is configured to read the executable instructions from the memory and execute the instructions to implement the method described in any one of claims 1 to 4.

Images