Fault self-healing method and system applicable to primary and secondary integrated power distribution equipment

Abstract

This application relates to the field of fault self-healing technology for power distribution equipment, specifically to a fault self-healing method and system applicable to integrated primary and secondary power distribution equipment. The method includes: determining the energy density value of the power distribution equipment at each moment based on the energy distribution of the differential current signal in the frequency domain at each moment, and the time interval from any moment within a preset time period to each moment; determining a fault discrimination factor based on the energy density value and switching state of the power distribution equipment at all moments before the current moment, as well as all modal components and their center frequencies; and determining a fault judgment threshold based on the fault discrimination factor to determine whether to trigger fault self-healing of the power distribution equipment. This application solves the problem that traditional methods are difficult to accurately identify fault characteristics under complex operating conditions. By fusing dynamic thresholds with multimodal features, it improves the accuracy and reliability of fault identification during the fault self-healing process.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fault self-healing of power distribution equipment, in particular to a fault self-healing method and system suitable for primary and secondary fusion power distribution equipment. BACKGROUND

[0002] Due to the construction of smart grid and large-scale access of distributed energy, the problem of time-consuming and laborious fault disposal of distribution network is becoming more and more serious, and the primary and secondary fusion power distribution equipment fault self-healing technology emerges as the times require and develops rapidly. The primary and secondary fusion power distribution equipment fault self-healing technology can realize rapid and accurate positioning of faults, greatly reduce the fault outage time and influence area, change the original fault disposal from hours to minutes or even seconds, greatly improve the power supply reliability, improve the user satisfaction, and reduce the operation workload and operation danger of the operation and maintenance personnel.

[0003] With the large-scale access of new energy, the distribution network has changed from the traditional single power supply radial structure to a multi-power active network, which has caused significant changes in the amplitude and direction of fault current, thereby destroying the premise condition of traditional overcurrent protection based on "one-way flow of current", resulting in misjudgment of fault direction. At the same time, high-resistance grounding faults are easily submerged by load fluctuations and noise due to weak fault current, further increasing the difficulty of fault identification. The existing technology mainly relies on threshold comparison of power frequency electrical quantities to extract faults, which is difficult to accurately extract effective fault features under complex working conditions, thereby reducing the accuracy and reliability of fault identification in the fault self-healing process. SUMMARY

[0004] In order to solve the above technical problems, the purpose of the present application is to provide a fault self-healing method and system suitable for primary and secondary fusion power distribution equipment, and the technical solution adopted is as follows:

[0005] In the first aspect, the present application provides a fault self-healing method suitable for primary and secondary fusion power distribution equipment, which comprises the following steps:

[0006] Obtain the forward difference current data of the power distribution equipment at all times before each time to form the difference current signal at each time;

[0007] Determine the energy feature value of the power distribution equipment at each time based on the energy distribution of the difference current signal in the frequency domain at each time; determine the time feature value of any time within a preset time length before each time to the time interval of each time, and combine the energy feature value to determine the energy density value of the power distribution equipment at each time;

[0008] A mode decomposition algorithm is applied to the energy density values ​​at all times prior to the current time to obtain all mode components. Based on the energy density values ​​and switching states of the power distribution equipment at all times prior to the current time, as well as all mode components and their center frequencies, the fault confidence probability and feature importance vector of the power distribution equipment at the current time are determined. The fault discrimination factor of the power distribution equipment at the current time is determined by combining the fault confidence probability and the feature importance vector.

[0009] Based on the fault discrimination factor, the fault threshold correction factor at the current moment is determined to obtain the fault judgment threshold of the power distribution equipment at the current moment, and to determine whether the fault self-healing of the power distribution equipment is triggered at the current moment.

[0010] Preferably, the method for determining the energy characteristic values ​​of the power distribution equipment at each time point is as follows:

[0011] The time-frequency energy matrix of the differential current signal at each time point is obtained by using a time-frequency conversion algorithm. The gradient matrix of the time-frequency energy matrix is ​​calculated, and the maximum value in the gradient matrix is ​​taken as the energy characteristic value of the power distribution equipment at each time point.

[0012] Preferably, the time characteristic value at any given moment is the exponential result of the time interval from any moment to each moment within a preset time period prior to each moment.

[0013] Preferably, the expression for the energy density value of the power distribution equipment at each time point is: In the formula, Indicates time Energy density value of the power distribution equipment; This represents the energy characteristic value of the power distribution equipment at time t; This represents the time characteristic value at time t; T represents the preset duration.

[0014] Preferably, the method for determining the fault confidence probability and feature importance vector of the power distribution equipment at the current moment is as follows:

[0015] The energy density, switching status, all modal components and their center frequencies of the power distribution equipment at all times before the current time are used as input to the random forest algorithm. The classification labels are set as fault and non-fault. The proportion of the number of all decision trees corresponding to the fault label in the classification results is used as the fault confidence probability of the power distribution equipment at the current time.

[0016] The energy density value, switching state, all modal components and their center frequencies are denoted as various features. The normalized value of the information gain of each feature in each decision tree is calculated. The mean of the normalized values ​​of the information gain of each feature in all decision trees is used as the feature importance value of each feature. The feature importance values ​​of all features are used to form the feature importance vector of the power distribution equipment at the current moment.

[0017] Preferably, the fault discrimination factor of the power distribution equipment at the current moment is the result of positive fusion of the modulus of the feature importance vector of the power distribution equipment at the current moment and the fault confidence probability.

[0018] Preferably, the expression for the fault threshold correction factor at the current moment is: In the formula, This represents the fault threshold correction factor at the current moment; This represents the fault discrimination factor of the power distribution equipment at the current moment; This represents the preset adjustment coefficient; norm() represents the normalization function.

[0019] Preferably, the fault determination threshold of the power distribution equipment at the current moment is the result of multiplying the preset basic fault determination threshold by the fault threshold correction factor at the current moment.

[0020] Preferably, determining whether the fault self-healing of the power distribution equipment is triggered at the current moment includes:

[0021] If the normalized value of the fault discrimination factor of the power distribution equipment at the current moment is greater than or equal to the fault discrimination threshold, then the power distribution equipment fault self-healing is triggered; otherwise, the power distribution equipment fault self-healing is not triggered. Secondly, embodiments of this application also provide a fault self-healing system suitable for integrated primary and secondary power distribution equipment, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described fault self-healing methods suitable for integrated primary and secondary power distribution equipment.

[0022] This application has at least the following beneficial effects:

[0023] This application constructs an energy density value based on the energy distribution of the differential current signal in the frequency domain at each time point, and the time interval from any time point within a preset duration to each time point. This effectively distinguishes fault discharge from non-fault factors such as load fluctuations and harmonic interference. Especially in weak fault scenarios such as high-resistance grounding faults, it can still accurately capture fault energy mutations and avoid missed detections. At the same time, the energy density value can highlight energy mutations near the current time point through a time decay mechanism, suppressing the influence of long-term interference signals, thereby reducing the false detection rate. Furthermore, this application extracts multi-dimensional features of fault signals and quantifies their importance by combining modal decomposition and random forest, constructing a fault discrimination factor. This effectively improves the accuracy and robustness of fault identification under complex operating conditions, significantly reduces the false detection rate, and helps enhance the self-healing capability of power distribution equipment. In summary, this application effectively copes with complex operating condition interference by using a dynamic threshold mechanism and multi-modal feature fusion, combined with the fault discrimination factor to adaptively adjust the fault judgment threshold, significantly improving the accuracy and reliability of fault identification during the fault self-healing process. Attached Figure Description

[0024] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 A flowchart illustrating the steps of a fault self-healing method for integrated primary and secondary power distribution equipment provided in one embodiment of this application;

[0026] Figure 2 This is a schematic diagram of the fault determination threshold extraction process provided in one embodiment of this application. Detailed Implementation

[0027] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the fault self-healing method and system applicable to primary and secondary integrated power distribution equipment proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0028] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0029] The following description, in conjunction with the accompanying drawings, details the specific scheme of the fault self-healing method and system applicable to primary and secondary integrated power distribution equipment provided in this application.

[0030] Please see Figure 1 The diagram illustrates a flowchart of a fault self-healing method for integrated primary and secondary power distribution equipment according to an embodiment of this application. The method includes the following steps:

[0031] Step S1: Obtain the forward differential current data of the power distribution equipment at all times prior to each time point, and form the differential current signal at each time point.

[0032] The forward differential current data of the power distribution equipment at all previous times is obtained using high-precision current transformers deployed in the primary and secondary integrated distribution network equipment, forming the differential current signal at each time point. The specific acquisition process is as follows:

[0033] The three-phase current data of the power distribution equipment is collected in real time. The vector synthesis formula is used to integrate the three-phase current data of the power distribution equipment at all times before each time into a current signal. The forward differential current data at each time in the current signal is calculated. That is, the current signal value between each time and the previous time is used as the forward differential current data at each time. The forward differential current data of the power distribution equipment at all times before each time are used to form the differential current signal at each time. At this time, the differential current signal is a zero-sequence current signal.

[0034] Furthermore, the switch status sensor is used to acquire the on / off position signal of the power distribution equipment, i.e., the switch status signal. The switch status signal is processed by a digital sampler to ensure accurate capture and timing consistency of its status changes. At each moment, the on and off states of the switch are represented by 0 and 1 respectively. Therefore, the obtained switch status signal of the power distribution equipment is a series of digital signals with amplitudes of 0 and 1.

[0035] It should be noted that the acquisition frequency of the three-phase current data and the acquisition frequency of the switch status signal are both set to 10kHz. In actual applications, as other implementation methods, implementers may also set their own data acquisition frequency according to specific circumstances. This embodiment does not impose any special restrictions.

[0036] The vector synthesis of three-phase currents is a well-known technique and will not be elaborated further.

[0037] Step S2: Based on the energy distribution of the differential current signal in the frequency domain at each time point, determine the energy characteristic value of the power distribution equipment at each time point; based on the time interval from any time point within a preset time period before each time point to each time point, determine the time characteristic value of any time point, and combine the energy characteristic value to determine the energy density value of the power distribution equipment at each time point.

[0038] Because the transient process of a fault in a primary and secondary integrated power distribution equipment is affected by non-fault factors such as load harmonics and backfeed current from distributed power sources, traditional current amplitude criteria are difficult to accurately distinguish between faults and interferences, resulting in location errors. Real faults are usually accompanied by significant energy release and sustained characteristics. For example, current surges caused by arc reignition or insulation breakdown manifest as a sharp increase in energy, rich high-frequency components, and high energy concentration at the actual moment in the time-frequency domain. In contrast, signals from non-fault interference factors have smooth amplitude fluctuations, narrow spectral distributions, and no significant energy concentration points. Therefore, this embodiment determines the energy characteristic value of the power distribution equipment at each moment based on the energy distribution of the differential current signal in the frequency domain. Based on the time interval from any moment within a preset duration prior to each moment, the time characteristic value of any moment is determined. Combined with the energy characteristic value, the energy density value of the power distribution equipment at each moment is determined to identify whether a fault has occurred in the power distribution equipment. The specific process is as follows:

[0039] First, this embodiment determines the energy characteristic value of the power distribution equipment at each time point based on the energy distribution of the differential current signal in the frequency domain at each time point, specifically:

[0040] In this embodiment, the differential current signal at each time point is used as the input to the S-transform time-frequency analysis algorithm. The frequency resolution is set to 0.1~1kHz to cover the main frequency of the traveling wave, and the window function width is set to 5ms to approximate the duration of the initial transient of the fault. The output time-frequency energy matrix is ​​used to characterize the distribution of fault discharge energy in the time-frequency domain.

[0041] It should be noted that the columns of the time-frequency energy matrix represent each frequency, the rows represent each time, and the element values ​​represent energy.

[0042] The S-transform time-frequency analysis algorithm is a well-known technique, and the specific process of obtaining the time-frequency energy matrix using it will not be elaborated here.

[0043] Furthermore, the gradient matrix of the time-frequency energy matrix is ​​calculated. Specifically, for each element in each row of the time-frequency energy matrix and a predetermined number of subsequent elements, a neighborhood of each element is formed. The range of all elements in the neighborhood, i.e., the energy range, is calculated as the energy gradient of each element. Following the energy gradient calculation method described above, all elements in the time-frequency energy matrix are traversed to obtain the energy gradient of all elements, forming the time-frequency gradient matrix.

[0044] It should be noted that the preset quantity is set manually. In this embodiment, the preset quantity is 5. In actual applications, as other implementation methods, implementers can also set it according to specific circumstances. This embodiment does not impose any special restrictions.

[0045] Furthermore, in this embodiment, the time feature value of any given moment is determined based on the time interval between any moment within a preset duration prior to each moment. Specifically:

[0046] In this embodiment, the exponentialized result of the time interval from any time point within a preset time period prior to each time point is used as the time feature value of any time point.

[0047] It should be noted that in this embodiment, the exponential function value with the natural constant as the base and the time interval from any time to each time within a preset time period before each time as the independent variable is used as the exponential result of the time interval from any time to each time within a preset time period before each time. In actual application, as other implementation methods, implementers may also use other exponential functions according to specific circumstances. This embodiment does not impose any special restrictions.

[0048] Furthermore, in this embodiment, based on the time characteristic value at any given time and in combination with the energy characteristic value, the energy density value of the power distribution equipment at each time is determined, specifically as follows:

[0049] As one implementation method, in this embodiment, time Energy density value of power distribution equipment The expression is: In the formula, This represents the energy characteristic value of the power distribution equipment at time t; This represents the time characteristic value at time t; T represents the preset duration.

[0050] It should be noted that the preset duration is set manually. In this embodiment, the preset duration is 20ms. In actual applications, as other implementation methods, implementers can also set it according to specific circumstances. This embodiment does not impose any special restrictions.

[0051] Based on the energy density values ​​of the power distribution equipment at each time point, it can be understood that the energy density value characterizes the distribution intensity of fault discharge energy in the time domain. The energy characteristic value reflects the intensity of the discharge at the fault point. If the energy characteristic value at a time point before the current time is larger, it indicates that the fault energy change at that time point before the current time is more violent. Furthermore, if a certain time point is closer to the current time, it indicates that the energy at the current time is higher, and the corresponding energy density value is larger, indicating that it is more likely to be caused by the intense discharge at the fault point, and that the power distribution equipment is more likely to be faulty at the current time.

[0052] Conversely, if the energy characteristic value at a certain time before the current time is smaller, it indicates that the energy change at that time before the current time is more gradual. Furthermore, if a certain time is farther away from the current time, it indicates that the energy at the current time is lower, and the corresponding energy density value is smaller, indicating that it is more likely to be caused by normal fluctuations or interference, and that the possibility of a fault in the power distribution equipment at the current time is smaller.

[0053] Thus, this embodiment constructs an energy density value by combining time-frequency energy distribution and time decay characteristics, which can effectively distinguish between fault discharge and interference signals, and significantly improve the accuracy and robustness of fault identification.

[0054] Step S3: Apply a mode decomposition algorithm to the energy density values ​​at all times before the current time to obtain all mode components; based on the energy density values ​​and switching states of the power distribution equipment at all times before the current time, as well as all mode components and their center frequencies, determine the fault confidence probability and feature importance vector of the power distribution equipment at the current time; combine the fault confidence probability and the feature importance vector to determine the fault discrimination factor of the power distribution equipment at the current time.

[0055] In complex power systems, once a fault occurs in the integrated primary and secondary power distribution equipment, the fault signal it generates is often mixed with various factors such as load fluctuations from daily electricity consumption, harmonic interference, and current backfeeding from distributed power sources. This makes it very difficult to identify the fault by relying solely on the magnitude of the current or simple judgment criteria. Not only is it easy to make misjudgments, but it is also impossible to accurately determine whether the fault has actually occurred, resulting in a significant reduction in the reliability of fault identification.

[0056] Therefore, this embodiment further employs a mode decomposition algorithm on the energy density values ​​at all times prior to the current time to obtain all mode components; based on the energy density values ​​and switching states of the power distribution equipment at all times prior to the current time, as well as all mode components and their center frequencies, the fault confidence probability and feature importance vector of the power distribution equipment at the current time are determined; combining the fault confidence probability and the feature importance vector, the fault discrimination factor of the power distribution equipment at the current time is determined, and the true fault features are accurately extracted from them. The specific process is as follows:

[0057] In this embodiment, the energy density values ​​at all times prior to the current time are first used as the input of the mode decomposition algorithm, and all mode components are output. In this embodiment, the mode decomposition algorithm is the variational mode decomposition algorithm. In practical applications, as other implementation methods, implementers may also use other mode decomposition methods such as empirical mode decomposition algorithms according to specific circumstances. This embodiment does not impose any special restrictions on the selection of mode decomposition algorithms.

[0058] Variational mode decomposition algorithm is a well-known technique, and the specific process of obtaining modal components using it will not be elaborated here.

[0059] Furthermore, this embodiment determines the fault confidence probability and feature importance vector of the power distribution equipment at the current moment based on the energy density values ​​and switching states of the power distribution equipment at all previous moments, as well as all modal components and their center frequencies. Specifically:

[0060] In this embodiment, the energy density value, switching state, all modal components and their center frequencies of the power distribution equipment at all times before the current time are used as input to the random forest algorithm. The classification labels are set as fault and non-fault, the number of decision trees is set to D, and the maximum depth is set to V layers to prevent overfitting and maintain the logic of topological correlation. The proportion of the number of all decision trees corresponding to the fault label in the classification results is used as the fault confidence probability of the power distribution equipment at the current time.

[0061] Furthermore, the energy density value, switching state, all modal components and their center frequencies are denoted as various features. The normalized value of the information gain of each feature in each decision tree is calculated. The mean of the normalized values ​​of the information gain of each feature in all decision trees is used as the feature importance value of each feature. The feature importance values ​​of all features are used to form the feature importance vector of the power distribution equipment at the current moment.

[0062] It should be noted that the values ​​of the number of decision trees D and the maximum depth V are set manually. In this embodiment, the number of decision trees D is 200 and the maximum depth V is 8. In actual applications, as other implementation methods, implementers can also set them according to specific circumstances. This embodiment does not impose any special restrictions.

[0063] The random forest algorithm and the method for calculating information gain are well-known techniques, and the principles of the random forest algorithm and the calculation process of information gain will not be elaborated here.

[0064] Furthermore, this embodiment combines the fault confidence probability and the feature importance vector to determine the fault discrimination factor of the power distribution equipment at the current moment, specifically:

[0065] In this embodiment, the result of positively fusing the modulus of the feature importance vector of the power distribution equipment at the current moment with the fault confidence probability is used as the fault discrimination factor of the power distribution equipment at the current moment.

[0066] It should be understood that positive fusion refers to combining two or more indicators through addition or multiplication to obtain a comprehensive indicator, thereby more comprehensively and accurately assessing a phenomenon or problem. This fusion method is not limited to simple arithmetic operations, but can also include more complex statistical models and analytical methods. Implementers can choose according to specific circumstances, and this embodiment does not impose any special restrictions.

[0067] Preferably, in this embodiment, the product of the modulus of the feature importance vector of the power distribution equipment at the current moment and the fault confidence probability is used as the fault discrimination factor of the power distribution equipment at the current moment. In practical applications, as other implementation methods, implementers may also adopt other positive fusion methods such as sum values ​​according to specific circumstances. This embodiment does not impose any special restrictions.

[0068] Based on the fault discrimination factor of the power distribution equipment at the current moment, it can be understood that the fault discrimination factor reflects the confidence of the fault occurrence. If the magnitude of the feature importance vector at the current moment is larger, it means that the power distribution equipment is more likely to fail at the current moment. Therefore, the corresponding fault discrimination factor is larger. At the same time, if the confidence probability of the power distribution equipment at the current moment is larger, it means that the confidence of the power distribution failure at the current moment is higher. Therefore, the corresponding fault discrimination factor is larger.

[0069] Conversely, if the magnitude of the feature importance vector is smaller at the current moment, it means that the probability of a fault in the power distribution equipment at the current moment is smaller, and therefore the corresponding fault discrimination factor is smaller. At the same time, if the confidence probability of a fault in the power distribution equipment at the current moment is smaller, it means that the confidence of a fault occurring in the power distribution at the current moment is lower, and therefore the corresponding fault discrimination factor is smaller.

[0070] Thus, by combining mode decomposition and random forest, multidimensional features of fault signals are extracted and their importance is quantified to construct fault discrimination factors, thereby effectively improving the accuracy and robustness of fault identification under complex working conditions, significantly reducing the false judgment rate, and enhancing the self-healing capability of power distribution equipment.

[0071] S4: Based on the fault discrimination factor, determine the fault threshold correction factor at the current moment to obtain the fault judgment threshold of the power distribution equipment at the current moment, and determine whether the fault self-healing of the power distribution equipment is triggered at the current moment.

[0072] Due to various interference factors affecting the power grid, the operating conditions of the integrated fault-tolerant device vary greatly. A single fixed threshold value is no longer sufficient to accurately identify faults under different operating conditions, which can affect the fault identification accuracy of the fault self-healing system to some extent. Therefore, to improve the reliability and adaptability of the fault self-healing system, based on the fault discrimination factor obtained in step 3 above, this embodiment determines the fault threshold correction factor at the current moment to obtain the fault judgment threshold of the power distribution equipment at the current moment, and determines whether the fault self-healing of the power distribution equipment is triggered at the current moment. The specific process is as follows:

[0073] As one implementation method, in this embodiment, the fault threshold correction factor at the current moment... The expression is: In the formula, This represents the fault discrimination factor of the power distribution equipment at the current moment; This represents the preset adjustment coefficient; norm() represents the normalization function.

[0074] It should be noted that the preset adjustment coefficient is set manually. In this embodiment, the preset adjustment coefficient is 0.6. In actual application, as other implementation methods, implementers can also set it according to specific circumstances. This embodiment does not impose any special restrictions.

[0075] Based on the fault threshold correction factor at the current moment, it can be understood that if the fault discrimination factor at the current moment is larger, it indicates a higher confidence level in the existence of the fault. Therefore, reducing the fault threshold correction factor can lower the fault determination threshold and improve the fault detection rate. Conversely, if the fault discrimination factor at the current moment is smaller, it indicates a lower confidence level in the existence of the fault. Therefore, increasing the fault threshold correction factor can raise the fault determination threshold, reduce the false alarm rate, and enhance the reliability of the system's discrimination.

[0076] Furthermore, in this embodiment, the result of multiplying the preset basic fault judgment threshold by the fault threshold correction factor at the current moment is used as the fault judgment threshold of the power distribution equipment at the current moment.

[0077] Preferably, the schematic diagram of the fault determination threshold extraction process provided in this embodiment is as follows: Figure 2 As shown.

[0078] It should be noted that the preset basic fault judgment threshold is set manually, and the value of the judgment threshold is 0.1 to 1. In this embodiment, the preset basic fault judgment threshold is 0.3. In actual application, the implementer can also set it according to the specific situation. This embodiment does not impose any special restrictions.

[0079] If the normalized value of the fault discrimination factor of the power distribution equipment is greater than or equal to the fault judgment threshold at the current moment, the power distribution equipment fault self-healing is triggered. Conversely, if the normalized value of the fault discrimination factor of the power distribution equipment is less than the fault judgment threshold at the current moment, the power distribution equipment fault self-healing is not triggered.

[0080] Thus, this embodiment effectively addresses complex operating conditions and significantly improves the accuracy and reliability of fault identification during the fault self-healing process by using a dynamic threshold mechanism and multimodal feature fusion, combined with fault discrimination factors to adaptively adjust the fault determination threshold.

[0081] Based on the same inventive concept as the above methods, this application also provides a fault self-healing system suitable for primary and secondary integrated power distribution equipment, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described fault self-healing methods suitable for primary and secondary integrated power distribution equipment.

[0082] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, specific embodiments of this specification have been described above. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0083] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0084] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the principles of this application should be included within the protection scope of this application.

Claims

1. A fault self-healing method for a secondary fused power distribution device, comprising: The method comprises the following steps: Obtaining the forward difference current data of the power distribution equipment at all time points before each time point to form the difference current signal at each time point; Based on the energy distribution of the difference current signal at each time point in the frequency domain, the energy characteristic value of the power distribution equipment at each time point is determined; based on the time interval from any time point within a preset time length before each time point to the time point, the time characteristic value of the any time point is determined, and the energy density value of the power distribution equipment at each time point is determined in combination with the energy characteristic value; Using a modal decomposition algorithm on the energy density values at all time points before the current time point to obtain all modal components; based on the energy density values and the switching state of the power distribution equipment at all time points before the current time point, and all modal components and their center frequencies, the fault confidence probability and the feature importance vector of the power distribution equipment at the current time point are determined; the fault discrimination factor of the power distribution equipment at the current time point is determined by comprehensively considering the fault confidence probability and the feature importance vector; Based on the fault discrimination factor, the fault threshold correction factor at the current time point is determined to obtain the fault determination threshold of the power distribution equipment at the current time point, and it is determined whether to trigger the fault self-healing of the power distribution equipment at the current time point; The determination method of the fault confidence probability and the feature importance vector of the power distribution equipment at the current time point is: The energy density values, the switching state, all modal components and their center frequencies of the power distribution equipment at all time points before the current time point are taken as the input of the random forest algorithm, the classification label is set as fault and non-fault, and the proportion of the number of fault labels corresponding to all decision trees in the classification results in all decision trees is taken as the fault confidence probability of the power distribution equipment at the current time point; The energy density values, the switching state, all modal components and their center frequencies are recorded as various features, the normalized values of the information gain of each feature in each decision tree are calculated, the mean value of the normalized values of the information gain of all features in all decision trees is taken as the feature importance value of each feature, and the feature importance values of all features are combined to form the feature importance vector of the power distribution equipment at the current time point.

2. The fault self-healing method for a secondary fused power distribution device of claim 1, wherein, The determination method of the energy characteristic value of the power distribution equipment at each time point is: The time-frequency energy matrix of the difference current signal at each time point is obtained by using a time-frequency conversion algorithm, the gradient matrix of the time-frequency energy matrix is calculated, and the maximum value in the gradient matrix is taken as the energy characteristic value of the power distribution equipment at each time point.

3. The fault self-healing method for a secondary power distribution system of claim 1, wherein, The time characteristic value of any time point is the exponential result of the time interval from any time point within a preset time length before each time point to the time point.

4. The fault self-healing method for a secondary power distribution system of claim 1, wherein, The expression of the energy density value of the power distribution equipment at each time point is: ; in the formula, represents the energy density value of the power distribution equipment at time point ; represents the energy characteristic value of the power distribution equipment at time point t; represents the time characteristic value at time point t; and T represents a preset time length.

5. The fault self-healing method for a secondary power distribution system of claim 1, wherein, The fault discrimination factor of the power distribution equipment at the current time point is the modulus value of the feature importance vector of the power distribution equipment at the current time point and the positive fusion result of the fault confidence probability.

6. The fault self-healing method for a secondary power distribution system of claim 1, wherein, The expression of the failure threshold correction factor at the current time is: ; wherein, represents the failure threshold correction factor at the current time; represents the failure discrimination factor of the power distribution device at the current time; represents a preset adjustment coefficient; and norm() represents a normalization function.

7. The fault self-healing method for a secondary power distribution system of claim 1, wherein, The fault determination threshold of the power distribution equipment at the current time point is the product of a preset basic fault determination threshold and the fault threshold correction factor at the current time point.

8. The fault self-healing method for a secondary power distribution system of claim 1, wherein, The determination of whether to trigger the fault self-healing of the power distribution equipment at the current time point comprises: If the normalized value of the fault discrimination factor of the power distribution equipment at the current time point is greater than or equal to the fault determination threshold, the fault self-healing of the power distribution equipment is triggered, otherwise, the fault self-healing of the power distribution equipment is not triggered.

9. A fault self-healing system for a secondary fused power distribution device, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein, The processor implements the steps of the fault self-recovery method for a secondary fusion power distribution device according to any one of claims 1-8 when executing the computer program.

Images