Peak overcurrent protection system and method for three-phase rectification power supply

Abstract

The invention provides a peak overcurrent protection system and method for a three-phase rectification power supply, and relates to the technical field of circuit protection, and the method comprises the steps: carrying out the clustering analysis of multi-dimensional electric quantity data during the operation of the three-phase rectification power supply, constructing a state recognition model, and obtaining a risk early warning signal of the operation state of the three-phase rectification power supply; according to the risk early warning signal and the constraint condition of the three-phase rectification power supply, reactive current compensation amount is calculated, and a reactive current control instruction is obtained; executing the reactive current control instruction to inject reactive current into the three-phase rectification power supply, and monitoring to obtain change characteristics of electrical characteristics of a port of the three-phase rectification power supply; and performing protection logic judgment on the basis of the current amplitude change rate and the phase deviation characteristic in the change characteristic to obtain a fusion judgment result, and outputting a tripping signal and executing overcurrent isolation. According to the application, peak current advanced interception and multi-criterion collaborative decision based on intelligent early warning can be realized; therefore, over-current protection from passive cut-off to active defense is completed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of circuit protection, more particularly, the present application relates to a peak overcurrent protection system and method of three-phase rectifier power supply. BACKGROUND

[0002] Circuit protection technology is the cornerstone of ensuring the safe and stable operation of power electronic devices and systems. Among them, overcurrent protection, as one of the most core protection types, the core task is to quickly and reliably cut off the fault current when the system fails, such as short circuit, to prevent the expensive power devices and equipment from being damaged by overcurrent.

[0003] In the three-phase rectifier power supply system, peak overcurrent protection is particularly important. The protection method commonly used at present is mainly based on the fixed threshold comparison principle, that is, the real-time monitoring of the AC side or DC side current, when its amplitude exceeds the preset hardware or software action threshold, the protection device sends a trip instruction to drive the circuit breaker to break, which has inherent technical limitations, that is: first, the protection action lags behind the development of the fault, and can only be passively responded when the overcurrent has occurred and reached a dangerous level, which cannot suppress the first dangerous peak of the current, and the impact on the power electronic devices is still huge; second, the criterion is single and depends on the current amplitude, which is easy to cause unnecessary downtime due to misjudgment when the system is disturbed or subjected to a start-up impact, which will affect the continuity of power supply; in addition, it lacks the ability to predict and actively intervene in the fault, and belongs to the remedial passive protection, which cannot meet the requirements of protection speed, selectivity and reliability of high power electronic power systems. Therefore, how to realize the peak current early interception based on intelligent early warning and multi-criteria collaborative decision-making to complete the overcurrent protection from passive cutting to active defense is a difficult problem faced by the industry. SUMMARY

[0004] The present application provides a peak overcurrent protection system and method of three-phase rectifier power supply, which can realize peak current early interception based on intelligent early warning and multi-criteria collaborative decision-making to complete the overcurrent protection from passive cutting to active defense.

[0005] In the first aspect, the present application provides a peak overcurrent protection method of three-phase rectifier power supply, which comprises the following steps: Collecting multi-dimensional electrical quantity data during the operation of the three-phase rectifier power supply; Performing cluster analysis on the multi-dimensional electrical quantity data, and then constructing a state recognition model based on support vector domain description, classifying and recognizing the operating state of the three-phase rectifier power supply through the state recognition model, and obtaining a risk early warning signal; Calculating the reactive current compensation amount according to the risk early warning signal and the constraint condition of the three-phase rectifier power supply, and obtaining a reactive current control instruction; The reactive current control command is executed to inject reactive current into the three-phase rectifier power supply, thereby monitoring the changes in the electrical characteristics of the three-phase rectifier power supply port. Based on the current amplitude change rate and phase offset characteristics in the aforementioned change features, protection logic judgment is performed to obtain a fusion judgment result. When the fusion judgment result meets the action conditions, a trip signal is output and overcurrent isolation is performed.

[0006] In this embodiment, the action conditions specifically include: The DC-side current exceeds the first current threshold and the rate of change of the current exceeds the preset gradient; The current in any phase of the AC side exceeds the second current threshold and the duration exceeds the time window; The voltage and current phase angle deviation exceeds the safe range and continues for a preset period.

[0007] In this embodiment, the multidimensional electrical quantity data includes: AC three-phase voltage, AC three-phase current, DC output voltage, and DC output current.

[0008] In this embodiment, the clustering analysis of the multidimensional electrical quantity data to construct a state recognition model based on support vector domain description specifically includes: Cluster analysis was performed on the multidimensional electrical quantity data to obtain a cluster center set; The support vector field description algorithm is used to determine the equilibrium point subsets for different operating states based on the cluster center set; Spatial segmentation is performed on a subset of equilibrium points across all operating states to obtain a state recognition model.

[0009] In this embodiment, the risk warning signal obtained by classifying and identifying the operating status of the three-phase rectifier power supply through the state recognition model specifically includes: Calculate the membership degree of the equilibrium point subset of each operating state in the multidimensional electrical quantity data and the state recognition model; All membership degrees are compared with preset membership degree thresholds to filter out overcurrent risk patterns; The risk warning signal is determined based on the weighted values ​​of all membership degrees and the overcurrent risk pattern.

[0010] In this embodiment, the reactive current compensation amount is calculated based on the risk warning signal and the constraints of the three-phase rectified power supply, and the reactive current control command is obtained, specifically including: The risk level of the risk warning signal is analyzed, and then the risk level is mapped to the initial compensation value of reactive current. The direction of reactive current compensation is determined based on the overcurrent risk mode of the aforementioned risk warning signal. The initial compensation value is optimized by limiting the amplitude under the constraints of the three-phase rectified power supply to obtain the reactive current compensation amount. A reactive current control command is generated based on the compensation direction and the reactive current compensation amount.

[0011] In this embodiment, the process of executing the reactive current control command to inject reactive current into the three-phase rectified power supply, and then monitoring the changes in the electrical characteristics of the three-phase rectified power supply ports, specifically includes: The reactive current control command is executed, and the changes in AC three-phase voltage, AC three-phase current, DC output voltage, and DC output current after reactive current injection are collected. The dynamic response trajectory of the port's electrical characteristics is extracted based on the changed data; The current amplitude change rate and phase offset characteristics are generated by the dynamic response trajectory, thereby obtaining the change characteristics of the electrical characteristics of the three-phase rectifier power supply port.

[0012] In this embodiment, the protection logic judgment based on the current amplitude change rate and phase shift characteristics in the change features, and the resulting fusion judgment result specifically includes: The rate of change of current amplitude in the aforementioned change characteristics is compared with a preset current rate of change threshold to obtain a current amplitude judgment factor; The phase shift feature in the change characteristics is compared with a preset phase shift safety threshold to obtain the phase shift judgment factor; Logical operations are performed on the current amplitude judgment factor and the phase offset judgment factor to obtain a fusion judgment result.

[0013] In this embodiment, the constraints of the three-phase rectifier power supply refer to the operating parameter boundaries for stable operation. The constraints include: the maximum allowable overcurrent capability of the converter power devices, the overvoltage safety limit of the AC power grid, the stable operating range of the DC output voltage, and the upper limit of the fluctuation of the submodule capacitor voltage.

[0014] Secondly, this application provides a peak overcurrent protection system for a three-phase rectified power supply, used to implement a peak overcurrent protection method for a three-phase rectified power supply, the overcurrent protection system comprising: The data acquisition module is used to collect multi-dimensional electrical quantity data during the operation of the three-phase rectified power supply. The risk perception module is used to perform cluster analysis on the multidimensional electrical quantity data, and then construct a state recognition model based on support vector domain description. The state recognition model is used to classify and identify the operating state of the three-phase rectifier power supply and obtain risk warning signals. The collaborative calculation module is used to calculate the reactive current compensation amount based on the risk warning signal and the constraints of the three-phase rectifier power supply, and to obtain the reactive current control command. The suppression execution module is used to execute the reactive current control command to inject reactive current into the three-phase rectified power supply, and then monitor the changes in the electrical characteristics of the three-phase rectified power supply port. The fusion decision module is used to perform protection logic judgment based on the current amplitude change rate and phase offset characteristics in the change characteristics, obtain the fusion judgment result, and output a trip signal and perform overcurrent isolation when the fusion judgment result meets the action conditions.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: Multidimensional electrical quantity data of a three-phase rectifier power supply during operation are collected; cluster analysis is performed on the multidimensional electrical quantity data to construct a state recognition model based on support vector domain description; the operating state of the three-phase rectifier power supply is classified and identified through the state recognition model to obtain risk warning signals; reactive current compensation is calculated based on the risk warning signals and the constraints of the three-phase rectifier power supply to obtain reactive current control commands; reactive current is injected into the three-phase rectifier power supply by executing the reactive current control commands, and the changes in the electrical characteristics of the three-phase rectifier power supply ports are monitored; protection logic judgment is performed based on the current amplitude change rate and phase offset characteristics in the changes to obtain a fusion judgment result; when the fusion judgment result meets the action conditions, a trip signal is output and overcurrent isolation is performed.

[0016] Therefore, this application can realize peak current advance interception and multi-criteria collaborative decision-making based on intelligent early warning. First, by collecting multi-dimensional electrical quantity data of the three-phase rectifier power supply on the AC and DC sides, a high-dimensional operating feature space covering steady-state characteristics and transient disturbances is constructed, providing a comprehensive and high-resolution physical basis for early identification of fault trends. Then, clustering algorithms are used to extract cluster centers of operating modes, and a multi-state boundary model is constructed based on the support vector domain description method, so that the operating data forms a separable state domain in the high-dimensional space. Risk warning signals are generated by membership degree calculation, thereby breaking through the lag of traditional reliance on current amplitude exceeding the limit to take action and realizing early-stage intelligent prediction of overcurrent risks. Then, based on the risk level and risk mode of the risk warning signal, combined with the converter device capability, voltage stability and other operating constraints, the reactive current compensation is calculated. By injecting directional reactive current into the rectifier, the equivalent impedance and energy flow are actively changed to suppress the rapid rise of current. The trend effectively weakens the current peak before it forms, which helps reduce the initial impact on power devices. Subsequently, the dynamic response trajectory after reactive power compensation is collected, and two highly sensitive dynamic features, the rate of change of current amplitude and the phase shift of voltage and current, are extracted. By enhancing the perception of the disturbance response law, the shortcomings of traditional single amplitude criteria in monitoring latent faults and complex transient instability are compensated. Finally, the current amplitude change rate judgment factor and the phase shift judgment factor are logically fused, and the tripping action is triggered only when the abnormal change trend and the abnormal impedance characteristics occur simultaneously. This effectively avoids false actions caused by noise, grid fluctuations, etc., while maintaining high sensitivity, ensuring that overcurrent isolation is both timely and accurate. Through the synergistic effect of the above technology chain, this application realizes intelligent prediction of overcurrent risk, early suppression of peak current, dynamic identification of the nature of faults, and precise triggering of isolation action, which helps to improve the safety and reliability of three-phase rectifier power supplies in complex operating environments.

[0017] In summary, the technical solution adopted in this application can realize peak current advance interception based on intelligent early warning and multi-criteria collaborative decision-making, so as to complete overcurrent protection from passive interception to active defense. Attached Figure Description

[0018] To more clearly illustrate the technical solutions 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 embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of a peak overcurrent protection method for a three-phase rectified power supply provided in this application; Figure 2This is an exemplary flowchart for determining a state recognition model based on a support vector domain description, as provided in this application. Figure 3 This is an exemplary flowchart for determining reactive current control commands according to the present application; Figure 4 This is a modular structure diagram of a peak overcurrent protection system for a three-phase rectified power supply provided in this application. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] This application provides a peak overcurrent protection system and method for a three-phase rectifier power supply. The core of this system involves collecting multi-dimensional electrical quantity data during the operation of the three-phase rectifier power supply; performing cluster analysis on the multi-dimensional electrical quantity data to construct a state recognition model based on support vector domain description; classifying and identifying the operating state of the three-phase rectifier power supply using the state recognition model to obtain a risk warning signal; calculating the reactive current compensation amount based on the risk warning signal and the constraints of the three-phase rectifier power supply to obtain a reactive current control command; executing the reactive current control command to inject reactive current into the three-phase rectifier power supply, thereby monitoring the changes in the electrical characteristics of the three-phase rectifier power supply ports; performing protection logic judgment based on the current amplitude change rate and phase offset characteristics in the changes to obtain a fusion judgment result; and outputting a trip signal and performing overcurrent isolation when the fusion judgment result meets the action conditions.

[0022] Example 1: To better understand the above technical solution, the following will provide a detailed description of the technical solution in conjunction with the accompanying drawings and specific implementation methods. (Refer to...) Figure 1 As shown in the figure, this is a flowchart of a peak overcurrent protection method for a three-phase rectified power supply according to this embodiment of the present application. The overcurrent protection method includes the following steps: In step S1, multi-dimensional electrical quantity data of the three-phase rectified power supply during operation are collected.

[0023] It should be noted that, in this embodiment, the multidimensional electrical quantity data includes: AC three-phase voltage, AC three-phase current, DC output voltage, and DC output current; wherein, AC three-phase voltage refers to the potential difference between the three phase lines at the connection point between the three-phase rectifier power supply and the grid and the neutral point, which can be acquired using a three-phase voltage transformer; AC three-phase current refers to the instantaneous value of the current flowing through each phase line of the AC side of the three-phase rectifier power supply, which can be acquired using a current transformer; DC output voltage refers to the stable DC voltage established between the positive and negative terminals of the DC bus after rectification and conversion by the three-phase rectifier power supply, which can be acquired using a DC voltage sensor; DC output current refers to the current flowing from the DC side of the three-phase rectifier power supply to the load, which can be acquired using a DC current sensor.

[0024] In step S2, cluster analysis is performed on the multidimensional electrical quantity data to construct a state recognition model based on support vector domain description. The operating state of the three-phase rectifier power supply is classified and identified through the state recognition model to obtain a risk warning signal.

[0025] Preferably, in this embodiment, reference Figure 2 As shown, this figure is an exemplary flowchart for determining a state recognition model based on support vector domain description according to the present application. In this embodiment, the clustering analysis of the multidimensional electrical quantity data to construct the state recognition model based on support vector domain description can be implemented by the following steps: First, in step S21, cluster analysis is performed on the multidimensional electrical quantity data to obtain a cluster center set; Then, in step S22, the equilibrium point subsets for different operating states are determined based on the cluster center set using the support vector domain description algorithm. Finally, in step S23, the equilibrium point subset of all operating states is spatially segmented to obtain the state recognition model.

[0026] In specific implementation, firstly, the K-means clustering algorithm can be used to perform cluster analysis on the multidimensional electrical quantity data. That is, the three-phase AC voltage, three-phase AC current, DC output voltage, and DC output current at the same time stamp are arranged in order to obtain an 8-dimensional vector. Then, the 8-dimensional vectors at all times in the multidimensional electrical quantity data are used as the input of the K-means clustering algorithm, and the number of clusters K is set (for example, the initial value of K can be set to 1, and the value of K can be increased automatically through the iteration of the K-means clustering algorithm). Then, through the iteration of the K-means clustering algorithm, K cluster centers can be obtained. Then, the set of coordinates of all cluster centers is used as the cluster center set, where the cluster center represents an 8-dimensional vector. Then, a subset of data is constructed for each cluster center in the cluster center set. Specifically, for each cluster center, the Euclidean distance is calculated from the 8-dimensional vectors at all timestamps to that cluster center using the Euclidean distance mathematical expression. All 8-dimensional vectors are arranged in ascending order of Euclidean distance to obtain a high-dimensional feature space set. The high-dimensional feature space set is then divided equally according to the number K to obtain K subsets. The support vector domain description algorithm is then used to train the model on each of the K subsets. Specifically, the Gaussian radial basis function kernel is selected, and the support vector data description function in the sklearn library is used to train the model on the subsets. During the model training process, the smallest hypersphere containing the samples in the subset can be found in the high-dimensional feature space. The center coordinates of the smallest hypersphere can be used as an equilibrium point of the running state. The set of equilibrium points corresponding to the K subsets is then used as the equilibrium point subset. Finally, the Thiessen polygon graph segmentation algorithm is used to divide the high-dimensional feature space set into K non-overlapping convex polygon regions based on each equilibrium point. The rule of the Thiessen polygon graph segmentation algorithm is: for any point in the high-dimensional feature space set, the point is assigned to the smallest hypersphere region where the nearest equilibrium point is located; the mathematical decision model composed of the coordinates of each equilibrium point in the equilibrium point subset and the rule of the Thiessen polygon graph segmentation algorithm is used as the state recognition model.

[0027] It should be noted that in this application, different operating states refer to the categories of operating states of the three-phase rectifier power supply, including: normal operating state, known overcurrent risk state, and unknown abnormal state. The normal operating state refers to the state in which the three-phase rectifier power supply operates stably under rated parameters. Known overcurrent risk states include: shoot-through short circuit state, asymmetrical overload state, etc. Unknown abnormal states refer to states where the multi-dimensional electrical quantity data acquired in real time cannot be classified into known overcurrent risk states, indicating that the three-phase rectifier power supply has encountered a new fault mode requiring manual intervention. Furthermore, the balance point subset is defined by a support vector field description algorithm for each... The set of coordinates of the center of the smallest hypersphere established for each operating state, with each equilibrium point serving as the ideal value in the feature space for different operating states, is used to characterize the essential features of different states. The state recognition model is a classifier based on the Thiessen polygon graph theory, which can map electrical quantity data to a high-dimensional feature space and determine the operating state of the three-phase rectifier power supply based on the Thiessen polygon region corresponding to the electrical quantity data. In practical applications, an 8-dimensional vector composed of real-time acquired multi-dimensional electrical quantity data can be input into the state recognition model, and the state classification can be completed by comparing the distance of the 8-dimensional vector to each equilibrium point.

[0028] In this embodiment, the risk warning signal is obtained by classifying and identifying the operating status of the three-phase rectifier power supply through the state recognition model, which can be done in the following way: Calculate the membership degree of the equilibrium point subset of each operating state in the multidimensional electrical quantity data and the state recognition model; All membership degrees are compared with preset membership degree thresholds to filter out overcurrent risk patterns; The risk warning signal is determined based on the weighted values ​​of all membership degrees and the overcurrent risk pattern.

[0029] In specific implementation, firstly, the real-time collected multi-dimensional electrical quantity data is combined into an 8-dimensional vector, and the coordinates of each equilibrium point in the equilibrium point subset of the state recognition model are substituted into a Gaussian membership function. The membership degree of each operating state's equilibrium point subset is calculated using this Gaussian membership function, where the membership degree value is between 0 and 1. Then, a membership degree threshold (e.g., 0.7) is set, and all calculated membership degrees are compared with this threshold. If the membership degree of an operating state is greater than or equal to the threshold, the operating state belongs to that state mode. If the membership degree of all states is lower than this threshold, the current state is determined to be an unknown mode. The set of operating states consisting of all known state modes (e.g., "bridge arm short circuit state") with membership degrees exceeding the threshold and the "unknown mode" is considered as the overcurrent risk mode. Membership threshold is used to screen risk states. The value of membership threshold is between 0.6 and 1, and can be set according to the actual screening intensity. Finally, the membership of all operating states in the overcurrent risk mode is weighted and summed. The weight can be preset according to the severity of overcurrent caused by each operating state in historical data. The calculated weighted value is then compared with the risk action threshold (e.g., 0.8). That is, if the weighted value is greater than or equal to the risk action threshold, the highest risk level and specific risk mode are obtained. If the weighted value is less than the risk action threshold but the overcurrent risk mode set is not empty, a low risk level is obtained. The structured data containing the risk level and risk mode is then used as a risk warning signal. The risk action threshold is used to screen risk modes. The risk action threshold can be set according to the specifications of the three-phase rectifier power supply.

[0030] It should be noted that the membership degree in this application is used to quantify the similarity between electrical quantity data and preset operating state modes, which can provide accurate and quantifiable judgment basis for risk screening; the overcurrent risk mode refers to a set of one or more system state identifiers that may cause overcurrent faults, filtered by the membership degree threshold, representing a specific risk type; the risk warning signal includes the presence of risk in the three-phase rectifier power supply, the risk level (such as high, medium, low) and the risk mode, which is used to provide accurate and differentiated decision input for protection and control strategies (such as different degrees of reactive power compensation).

[0031] In step S3, the reactive current compensation amount is calculated based on the risk warning signal and the constraints of the three-phase rectified power supply to obtain the reactive current control command.

[0032] It should be noted that, in this embodiment, the constraints of the three-phase rectifier power supply refer to the operating parameter boundaries for stable operation. These constraints include: the maximum permissible overcurrent capacity of the converter power devices, the overvoltage safety limit of the AC power grid, the stable operating range of the DC output voltage, and the upper limit of the submodule capacitor voltage fluctuation. Specifically, the maximum permissible overcurrent capacity of the converter power devices refers to the maximum pulse current value that power switching devices such as insulated-gate bipolar transistors can withstand within their safe operating area, which can be set based on the rated collector current and maximum peak current parameters given in the three-phase rectifier power supply datasheet. The overvoltage safety limit of the AC power grid refers to the highest permissible operating voltage at the common connection point of the AC power grid to which the three-phase rectifier power supply is connected. The stable operating range of the DC output voltage refers to the minimum and maximum voltage range allowed for the DC bus voltage of the rectifier power supply to maintain normal operation. The upper limit of the submodule capacitor voltage fluctuation refers to the maximum permissible fluctuation amplitude of the supporting capacitor voltage of a single submodule in the modular multilevel converter during charging and discharging.

[0033] Preferably, in this embodiment, reference Figure 3 As shown, this figure is an exemplary flowchart of determining reactive current control commands according to the present application. In this embodiment, the reactive current compensation amount is calculated based on the risk warning signal and the constraints of the three-phase rectified power supply to obtain the reactive current control command. This can be achieved by the following steps: First, in step S31, the risk level of the risk warning signal is analyzed, and then the risk level is mapped to the initial compensation value of the reactive current. Secondly, in step S32, the direction of reactive current compensation is determined based on the overcurrent risk mode of the risk warning signal. Then, in step S33, the initial compensation value is limited and optimized by the constraints of the three-phase rectified power supply to obtain the reactive current compensation amount. Finally, in step S34, a reactive current control command is generated based on the compensation direction and the reactive current compensation amount.

[0034] In practice, firstly, the "risk level" field is read from the data structure of the risk warning signal, and a risk level-compensation value mapping table is established. This mapping table predefines the initial reactive current compensation value corresponding to different risk levels. For example, "high risk" level is mapped to -0.3 times the rated current, and "medium risk" level is mapped to -0.15 times the rated current. Then, the reactive current value is obtained by looking up the table according to the read risk level, and this reactive current value is used as the initial compensation value of reactive current. Secondly, the "overcurrent risk mode" field is read from the data structure of the risk warning signal, and a risk mode-compensation direction rule base is established. This rule base defines the compensation direction of reactive current under different risk modes. For example, for the "bridge arm short circuit" mode, the compensation direction is specified as negative (i.e., absorbing inductive reactive power from the grid); for the overcurrent risk caused by "voltage drop", the compensation direction is specified as positive (i.e., injecting inductive reactive power into the grid). The compensation direction of reactive current is obtained by matching the rules according to the read overcurrent risk mode in this rule base.

[0035] In addition, the initial compensation value is optimized by limiting the constraints of the three-phase rectifier power supply to obtain the reactive current compensation amount. Specifically, the constraints of the three-phase rectifier power supply are read from the parameter configuration manual. Based on the constraints, the upper and lower limits of the reactive current compensation amount allowed under the current operating conditions of the three-phase rectifier power supply are read. Then, the initial compensation value is limited to the upper and lower limits using a limiting function. The value processed by the limiting function is then used as the reactive current compensation amount. Finally, the compensation direction (positive or negative) and the reactive current compensation amount (a specific current value) are combined according to the instruction data format, such as a data packet including direction enumeration values ​​and current values. The combined data packet is used as a reactive current control instruction, which can be sent to the inner loop controller of the three-phase rectifier power supply.

[0036] It should be noted that the initial compensation value in this application is a reactive current reference value, which can quickly respond to the level of risk; the compensation direction is an indicator of whether the reactive current is injected or absorbed, which can ensure that the reactive current compensation action can correctly provide the expected support to the system voltage; the reactive current compensation amount is an executable reactive current value that incorporates safety constraints, which can ensure the safe and stable operation of the three-phase rectifier power supply while protecting the target; the reactive current control command is a data packet that includes reactive current control information (magnitude and direction), used to drive the rectifier to perform reactive current compensation.

[0037] In step S4, the reactive current control command is executed to inject reactive current into the three-phase rectified power supply, thereby monitoring the changes in the electrical characteristics of the three-phase rectified power supply port.

[0038] In this embodiment, the reactive current control command is executed to inject reactive current into the three-phase rectified power supply, and the changes in the electrical characteristics of the three-phase rectified power supply ports are then monitored. Specifically, this can be achieved in the following manner: The reactive current control command is executed, and the changes in AC three-phase voltage, AC three-phase current, DC output voltage, and DC output current after reactive current injection are collected. The dynamic response trajectory of the port's electrical characteristics is extracted based on the changed data; The current amplitude change rate and phase offset characteristics are generated by the dynamic response trajectory, thereby obtaining the change characteristics of the electrical characteristics of the three-phase rectifier power supply port.

[0039] In practice, the reactive current control command is first sent to the inner loop controller of the three-phase rectifier. The inner loop controller can adjust the control parameters of the three-phase rectifier according to the reactive current control command, thereby injecting reactive current corresponding to the magnitude of the reactive current control command. The three-phase AC voltage, three-phase AC current, DC output voltage, and DC output current after the reactive current injection are collected in real time through three-phase voltage transformers, current transformers, DC voltage sensors, and DC current sensors. The instantaneous values ​​of the collected three-phase AC voltage, three-phase AC current, DC output voltage, and DC output current are arranged in chronological order to form a sequence as change data. Then, a coordinate transformation algorithm is used to convert the AC three-phase voltage and three-phase current in the changing data into two mutually perpendicular DC components, resulting in the direct-axis component and the quadrature-axis component. With time as the horizontal axis, the DC output voltage, DC output current, direct-axis component, and quadrature-axis component are plotted as curves over time using a visualization library. These curves are then used as the dynamic response trajectory of the port electrical characteristics. The port electrical characteristics refer to the set of electrical parameters that characterize the operating state and energy interaction characteristics of the three-phase rectifier power supply at the AC connection port and DC output port, including: the amplitude, phase, and frequency characteristics of the AC voltage and current, and the stability of the DC voltage and current. Finally, differential calculations are used to calculate the amplitudes of the DC-side output voltage, DC-side output current, direct-axis component, and quadrature-axis component of the AC-side current in the dynamic response trajectory. The rate of change of these amplitudes over time is taken as the rate of change of current amplitude. The angle difference method is used to calculate the angle difference between the voltage vector and the current vector in the dynamic response trajectory, i.e., the angle difference between the direct-axis component and the quadrature-axis component. This angle difference is taken as the phase shift characteristic. The characteristic vector obtained by combining the rate of change of current amplitude and the phase shift characteristic is taken as the change characteristic of the electrical characteristics of the three-phase rectifier power supply port.

[0040] It should be noted that the dynamic response trajectory in this application refers to the dynamic response information of the port electrical quantity state transition process of the three-phase rectifier power supply after being subjected to reactive current disturbance; the current amplitude change rate refers to the rate of change of the amplitude of the AC side current composite vector with time, used to quantify the degree of rapid change of current in the three-phase rectifier power supply under disturbance, and is an indicator for judging overcurrent trend. The phase offset characteristic refers to the phase difference between the port voltage vector and the current vector, used to characterize the change of impedance characteristics of the three-phase rectifier power supply, and can serve as an important basis for judging the nature of the fault in the three-phase rectifier power supply; the change characteristics of the port electrical characteristics of the three-phase rectifier power supply are a low-dimensional data set containing amplitude change rate and phase offset information, which can be reduced in dimensionality and highlight the core dynamic information to provide effective input for the protection logic judgment of the three-phase rectifier power supply.

[0041] In step S5, protection logic judgment is performed based on the current amplitude change rate and phase offset characteristics in the change characteristics to obtain a fusion judgment result. When the fusion judgment result meets the action conditions, a trip signal is output and overcurrent isolation is performed.

[0042] In this embodiment, the protection logic judgment based on the current amplitude change rate and phase shift characteristics in the change features, and the fusion judgment result can be obtained in the following way: The rate of change of current amplitude in the aforementioned change characteristics is compared with a preset current rate of change threshold to obtain a current amplitude judgment factor; The phase shift feature in the change characteristics is compared with a preset phase shift safety threshold to obtain the phase shift judgment factor; Logical operations are performed on the current amplitude judgment factor and the phase offset judgment factor to obtain a fusion judgment result.

[0043] In practice, firstly, the pandas library is used to read the rate of change of current amplitude from the change features, and the rate of change of current amplitude is compared with a preset current rate of change threshold (e.g., 100A / ms). When the rate of change of current amplitude is greater than or equal to the current rate of change threshold, the current change is too fast, posing an overcurrent risk, and the current amplitude judgment factor can be set to 1. When the rate of change of current amplitude is less than the current rate of change threshold, the current change is considered to be within the safe range, and the current amplitude judgment factor can be set to 0. This Boolean variable with a value of 0 or 1 is then used as the current amplitude judgment factor. Next, the pandas library is used to read the phase offset features from the change features, and the phase offset features are compared with a preset phase offset safety threshold (e.g., 30 degrees). When the absolute value of the phase offset features is greater than or equal to the phase offset safety threshold, the system impedance characteristics are determined to be abnormal, posing a fault risk, and the phase offset judgment factor can be set to 1. When the absolute value of the phase offset features is less than the phase offset safety threshold, the phase relationship is considered normal, and the phase offset judgment factor can be set to 0. Then, the Boolean variable with a value of 0 or 1 is used as the phase offset judgment factor. Finally, logical operations are performed on the current amplitude judgment factor and the phase offset judgment factor to obtain the fusion judgment result. That is, the current amplitude judgment factor and the phase offset judgment factor are input into a logic AND gate circuit and processed by the logic AND operator of the logic AND gate circuit. For example, when both the current amplitude judgment factor and the phase offset judgment factor are 1, the logic operation output is 1, indicating that there is a risk of overcurrent fault. In other cases, the output is 0, and the system is in a safe state. Then, the logic output with a value of 0 or 1 is used as the fusion judgment result.

[0044] It should be noted that the current amplitude judgment factor in this application is a binary status identifier used to characterize whether an anomaly has occurred based solely on the current change rate dimension, quantifying continuous current change rate into a clear fault symptom signal; the phase offset judgment factor is a binary status identifier used to characterize whether an anomaly has occurred based solely on the voltage-current phase relationship dimension, quantifying continuous phase offset into a clear impedance anomaly signal; the fusion judgment result is the final decision signal that integrates both current change and phase offset information, and can ensure through logical AND operation that a fault is only determined when both current and phase are abnormal, thereby improving the accuracy and reliability of protection judgment and avoiding maloperation that may be caused by a single criterion; in addition, it should be noted that the current change rate threshold and phase offset safety threshold in this application can both be set according to the parameter configuration of the three-phase rectifier power supply.

[0045] In this embodiment, the action conditions specifically include: The DC-side current exceeds the first current threshold and the rate of change of the current exceeds the preset gradient; The current in any phase of the AC side exceeds the second current threshold and the duration exceeds the time window; The voltage and current phase angle deviation exceeds the safe range and continues for a preset period.

[0046] In specific implementation, when the fusion judgment result meets the action conditions, the output of the trip signal and the execution of overcurrent isolation can be carried out in the following way: First, the fusion judgment result is detected by a digital logic circuit judgment statement to see if it is 1. When the action conditions are detected, a high-level signal is output through the output interface of the digital logic circuit, and this high-level signal is used as the trip signal; then, the trip signal is transmitted to the circuit breaker's trip coil drive circuit through an optocoupler isolation circuit to drive the trip coil to be energized; finally, the three-phase rectified power supply is electrically isolated from the power grid through the mechanical tripping mechanism of the circuit breaker, completing the entire process of overcurrent protection; in addition, if the fusion judgment result is detected to be 0, an early warning signal is issued, which is used to remind manual intervention to check the overcurrent peak state of the three-phase rectified power supply.

[0047] It should be noted that the first current threshold refers to the maximum safe current value that the DC-side output circuit is allowed to operate for a long time. Preferably, the first current threshold can be set in the range of 1.2-1.5 times the rated current, which facilitates timely detection of severe overloads while ensuring that normal load fluctuations do not cause malfunctions. The preset gradient can be set to 100-200A / ms to facilitate rapid identification of abrupt changes in fault current such as short circuits. The second current threshold refers to the instantaneous overcurrent capability allowed by the AC-side power devices. Preferably, the second current threshold can be set in the range of 2-3 times the rated current to facilitate differentiation between normal start-up impacts and actual faults. The time window can be set according to actual needs and is not limited here. The safety range can be set in the range of 20°-30° to facilitate accurate identification of abnormal system impedance characteristics. The preset period can be set according to actual needs and is not limited here.

[0048] In summary, the technical solution adopted in this application can realize peak current advance interception based on intelligent early warning and multi-criteria collaborative decision-making, so as to complete overcurrent protection from passive interception to active defense.

[0049] Example 2: This application provides a peak overcurrent protection system for a three-phase rectified power supply, referring to... Figure 4 As shown, this figure is a modular structure diagram of a peak overcurrent protection system for a three-phase rectified power supply according to this application. The overcurrent protection system includes: Data acquisition module 100 is used to acquire multi-dimensional electrical quantity data during the operation of a three-phase rectified power supply; The risk perception module 200 is used to perform cluster analysis on the multidimensional electrical quantity data, and then construct a state recognition model based on support vector domain description. The state recognition model is used to classify and identify the operating state of the three-phase rectifier power supply to obtain a risk warning signal. The collaborative computing module 300 is used to calculate the reactive current compensation amount based on the risk warning signal and the constraints of the three-phase rectified power supply, and to obtain the reactive current control command. The suppression execution module 400 is used to execute the reactive current control command to inject reactive current into the three-phase rectified power supply, and then monitor the change characteristics of the electrical characteristics of the three-phase rectified power supply port. The fusion decision module 500 is used to perform protection logic judgment based on the current amplitude change rate and phase offset characteristics in the change characteristics, obtain the fusion judgment result, and output a trip signal and perform overcurrent isolation when the fusion judgment result meets the action conditions.

[0050] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0051] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compactdisc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0052] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

Claims

1. A peak overcurrent protection method for a three-phase rectifier power supply, characterized by, The overcurrent protection method comprises the following steps: Collecting multi-dimensional electrical quantity data during operation of the three-phase rectifier power supply; Performing cluster analysis on the multi-dimensional electrical quantity data, and then constructing a state recognition model based on support vector domain description, and classifying and recognizing the operating state of the three-phase rectifier power supply through the state recognition model to obtain a risk warning signal; Calculating a reactive current compensation amount according to the risk warning signal and a constraint condition of the three-phase rectifier power supply to obtain a reactive current control instruction; Injecting a reactive current into the three-phase rectifier power supply by executing the reactive current control instruction, and then monitoring to obtain a change characteristic of a port electrical characteristic of the three-phase rectifier power supply; Performing protection logic judgment based on a current amplitude change rate and a phase shift characteristic in the change characteristic to obtain a fusion judgment result, and outputting a trip signal and executing overcurrent isolation when the fusion judgment result meets an action condition.

2. A peak overcurrent protection method for a three-phase rectifier power supply as defined in claim 1, characterized by, The action condition specifically comprises: The direct current side current exceeds a first current threshold value and the current change rate exceeds a preset gradient; Any phase current on the alternating current side exceeds a second current threshold value and the duration exceeds a time window; The voltage and current phase angle offset exceeds a safe range and lasts for a preset period.

3. The peak over-current protection method of a three-phase rectifier power supply as claimed in claim 1, wherein, The multi-dimensional electrical quantity data comprises: three-phase voltage on the alternating current side, three-phase current on the alternating current side, output voltage on the direct current side, and output current on the direct current side.

4. The peak over-current protection method of a three-phase rectifier power supply as claimed in claim 1, wherein, The cluster analysis on the multi-dimensional electrical quantity data, and then constructing a state recognition model based on support vector domain description specifically comprises: Performing cluster analysis on the multi-dimensional electrical quantity data to obtain a cluster center set; Determining a balance point subset of different operating states based on the cluster center set through a support vector domain description algorithm; Performing spatial segmentation on the balance point subset of all operating states to obtain a state recognition model.

5. The peak overcurrent protection method of a three-phase rectifier power supply as claimed in claim 1, wherein, The classification and recognition of the operating state of the three-phase rectifier power supply through the state recognition model to obtain a risk warning signal specifically comprises: Calculating the membership degree of the multi-dimensional electrical quantity data and the balance point subset of each operating state in the state recognition model; Comparing all membership degrees with a preset membership degree threshold value, and then screening to obtain an overcurrent risk mode; Determining a risk warning signal based on the weighted value of all membership degrees and the overcurrent risk mode.

6. A peak over-current protection method for a three-phase rectifier power supply as defined in claim 1, wherein, The calculation of the reactive current compensation amount according to the risk warning signal and the constraint condition of the three-phase rectifier power supply to obtain the reactive current control instruction specifically comprises: Analyzing the risk level of the risk warning signal, and then mapping the risk level to an initial compensation value of the reactive current; Determining the compensation direction of the reactive current according to the overcurrent risk mode of the risk warning signal; Optimizing the initial compensation value through the constraint condition of the three-phase rectifier power supply to obtain a reactive current compensation amount; Generating a reactive current control instruction based on the compensation direction and the reactive current compensation amount.

7. A peak over-current protection method for a three-phase rectifier power supply as defined in claim 1, wherein, The injection of the reactive current into the three-phase rectifier power supply by executing the reactive current control instruction, and then monitoring to obtain a change characteristic of a port electrical characteristic of the three-phase rectifier power supply specifically comprises: Executing the reactive current control instruction, and collecting change data of the three-phase voltage on the alternating current side, the three-phase current on the alternating current side, the output voltage on the direct current side, and the output current on the direct current side after the injection of the reactive current; Extracting a dynamic response trajectory of the port electrical characteristic based on the change data; The current amplitude change rate and the phase offset feature are generated through the dynamic response trajectory, and then the change feature of the electrical characteristic of the three-phase rectifier power port is obtained.

8. The peak over-current protection method of a three-phase rectifier power supply as claimed in claim 1, wherein, The protection logic judgment is performed based on the current amplitude change rate and the phase offset feature in the change feature, and the fusion judgment result specifically includes: The current amplitude change rate in the change feature is compared with a preset current change rate threshold, and a current amplitude judgment factor is obtained. The phase offset feature in the change feature is compared with a preset phase offset safety threshold, and a phase offset judgment factor is obtained. The current amplitude judgment factor and the phase offset judgment factor are subjected to logical operation, and a fusion judgment result is obtained.

9. A peak over-current protection method for a three-phase rectifier power supply as defined in claim 1, wherein, The constraint condition of the three-phase rectifier power refers to the operating parameter boundary of stable operation, and the constraint condition includes: the maximum allowable overcurrent capacity of the converter power device, the overvoltage safety limit of the AC side power grid, the stable operating range of the DC side output voltage, and the fluctuation upper limit of the sub-module capacitor voltage.

10. A peak overcurrent protection system of a three-phase rectifier power supply for executing a peak overcurrent protection method of a three-phase rectifier power supply according to any one of claims 1 to 9, characterized by The overcurrent protection system includes: A data acquisition module is configured to acquire multi-dimensional electrical quantity data of the three-phase rectifier power during operation. A risk perception module is configured to perform clustering analysis on the multi-dimensional electrical quantity data, and then construct a state recognition model based on support vector domain description, classify and identify the operating state of the three-phase rectifier power through the state recognition model, and obtain a risk early warning signal. A collaborative calculation module is configured to calculate the reactive current compensation amount according to the risk early warning signal and the constraint condition of the three-phase rectifier power, and obtain a reactive current control instruction. An inhibition execution module is configured to execute the reactive current control instruction to inject reactive current into the three-phase rectifier power, and then monitor the change feature of the electrical characteristic of the three-phase rectifier power port. A fusion decision module is configured to perform protection logic judgment based on the current amplitude change rate and the phase offset feature in the change feature, and obtain a fusion judgment result. When the fusion judgment result meets the action condition, a trip signal is output and overcurrent isolation is performed.

Images