Power quality self-adaptive diagnosis method and device for power grid, electric energy meter and medium

Abstract

The application provides a power quality adaptive diagnosis method and device of a power grid, an electric energy meter and a medium, wherein the method is applied to a B-level three-phase rail type electric energy meter, the method comprises the following steps: generating a dynamic behavior baseline of a plurality of power quality parameters, and generating an adaptive floating threshold corresponding to each power quality parameter based on the dynamic behavior baseline; monitoring the power quality parameters in real time; if any power quality parameter is monitored to exceed the corresponding adaptive floating threshold, a diagnosis event is triggered; based on the diagnosis event, a multi-dimensional data set corresponding to a power quality parameter triggering time point is captured; a preset root inference engine is used to analyze and process the multi-dimensional data set, and an electric energy quality diagnosis result of the current power grid is obtained. Thus, the B-level three-phase rail type electric energy meter is used for fault diagnosis of power quality, the operation and maintenance efficiency is improved, and the demand of the "double high" power grid for rapid intelligent operation and maintenance is met.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computers, in particular to an adaptive power quality diagnosis method and device for a power grid, an electric energy meter and a medium. BACKGROUND

[0002] With the deepening of the "double carbon" strategy, the high proportion of new energy represented by wind power and photovoltaic power is connected to the grid, and a large number of power electronic devices such as frequency converters and energy storage systems are widely used, which together constitute the significant characteristics of modern "double high" (high proportion of new energy, high proportion of power electronic devices) power grid. Among them, "double high" power grid optimizes energy structure, but also brings frequent and rapid switching action of power electronic devices and high sensitivity to disturbances.

[0003] In related technologies, a B-level three-phase rail type electric energy meter is usually used for monitoring and recording power quality events. The B-level three-phase rail type electric energy meter is a three-phase electric energy metering device with a precision level conforming to international B-level standards. With its accurate metering, flexible installation and rich functions, it is suitable for complex industrial and commercial energy management.

[0004] However, the B-level three-phase rail type electric energy meter can only realize sensing and cannot realize fault diagnosis of power quality, making it difficult to meet the urgent needs of "double high" power grid for rapid intelligent operation and maintenance. SUMMARY

[0005] Therefore, the embodiments of the present application provide an adaptive power quality diagnosis method and device for a power grid, an electric energy meter and a medium to solve the problem that the B-level three-phase rail type electric energy meter can only realize sensing and cannot realize fault diagnosis of power quality, making it difficult to meet the urgent needs of "double high" power grid for rapid intelligent operation and maintenance.

[0006] In a first aspect, the embodiments of the present application provide an adaptive power quality diagnosis method for a power grid, comprising:

[0007] Generating a dynamic behavior baseline of a plurality of power quality parameters, and generating an adaptive floating threshold corresponding to each power quality parameter based on the dynamic behavior baseline;

[0008] Real-time monitoring of the power quality parameters, if any power quality parameter exceeds the corresponding adaptive floating threshold, triggering a diagnosis event;

[0009] Based on the diagnosis event, capturing a multi-dimensional data set corresponding to the triggering time point of the any power quality parameter, the multi-dimensional data set comprising: three-phase current waveform data, three-phase voltage waveform data and switch state data synchronized in time with the waveform data;

[0010] The preset root cause inference engine is used to analyze and process the multi-dimensional data set, and obtain a power quality diagnosis result of the current power grid.

[0011] In an optional embodiment, the method further comprises:

[0012] For each of the power quality parameters, a sliding time window is used to count corresponding historical data;

[0013] According to the historical data corresponding to each of the power quality parameters, an exponential weighted moving average algorithm is used to calculate a dynamic average value and a dynamic standard deviation of each of the power quality parameters;

[0014] According to the dynamic average value and the dynamic standard deviation of each of the power quality parameters, a dynamic behavior baseline of each of the power quality parameters is generated;

[0015] Based on the dynamic behavior baseline, an adaptive floating threshold corresponding to each of the power quality parameters is generated.

[0016] In an optional embodiment, the method further comprises:

[0017] The dynamic average value of each of the power quality parameters is subtracted by a preset multiple of the dynamic standard deviation to obtain the adaptive floating threshold corresponding to the power quality parameter.

[0018] In an optional embodiment, the method further comprises:

[0019] A preset ring buffer is used to temporarily store high-resolution three-phase current waveform data, three-phase voltage waveform data, and waveform data time-synchronized switching value state data within a preset time period in real time;

[0020] The method further comprises:

[0021] Based on the diagnosis event, the preset ring buffer is locked, and multi-dimensional data sets within each of a preset time period before and after the triggering time point of any of the power quality parameters are captured.

[0022] In an optional embodiment, the multi-dimensional data set further comprises: a positive sequence, a negative sequence, or a zero sequence current time sequence curve calculated and obtained according to the three-phase current waveform data;

[0023] Alternatively, a positive sequence, a negative sequence, or a zero sequence voltage time sequence curve calculated and obtained according to the three-phase voltage waveform data.

[0024] In an optional embodiment, the preset root cause inference engine is used to analyze and process the multi-dimensional data set to obtain an electric energy quality diagnosis result of the current power grid, including:

[0025] The preset root cause inference engine is used to check the current time curve and the voltage time curve. If there is a current or voltage increase amplitude greater than a preset amplitude and a phase difference within a preset interval near 180 degrees, the current power grid is diagnosed as an upstream power grid fault. If there is the current or voltage increase amplitude greater than the preset amplitude and the phase difference within a preset interval near 0 degrees, the current power grid is diagnosed as a downstream user side fault.

[0026] In an optional embodiment, the preset root cause inference engine is used to analyze and process the multi-dimensional data set to obtain an electric energy quality diagnosis result of the current power grid, including:

[0027] The preset diagnosis model in the preset root cause inference engine is used to analyze and process the multi-dimensional data set, and an electric energy quality diagnosis result of the current power grid is output. The preset diagnosis model is obtained by training using a power grid sample data set. The power grid sample data includes a mapping relationship between multi-dimensional data features of a known power grid event sample and an event root cause.

[0028] In a second aspect, the embodiments of the present application further provide an electric energy quality adaptive diagnosis device of a power grid, including:

[0029] A generation module is configured to generate dynamic behavior baselines of a plurality of electric energy quality parameters, and generate adaptive floating threshold values corresponding to each of the electric energy quality parameters based on the dynamic behavior baselines.

[0030] A monitoring module is configured to monitor the electric energy quality parameters in real time. If any electric energy quality parameter exceeds the corresponding adaptive floating threshold value, a diagnosis event is triggered.

[0031] An acquisition module is configured to capture a multi-dimensional data set corresponding to a trigger time point of the any electric energy quality parameter based on the diagnosis event. The multi-dimensional data set includes three-phase current waveform data, three-phase voltage waveform data, and switch state data synchronized with the waveform data in time.

[0032] The acquisition module is further configured to use a preset root cause inference engine to analyze and process the multi-dimensional data set to obtain an electric energy quality diagnosis result of the current power grid.

[0033] Thirdly, this application also provides a Class B three-phase rail-mounted energy meter, including: a data acquisition module, a processor, a memory, a clock module, and a communication module. When the Class B three-phase rail-mounted energy meter is running, the data acquisition module, the processor, the clock module, and the communication module execute machine-readable instructions stored in the memory to perform the method described in any of the first aspects.

[0034] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the method described in any of the first aspects.

[0035] This application provides a method, device, energy meter, and medium for adaptive power quality diagnosis of power grids. The method is applied to a Class B three-phase DIN rail energy meter and includes: generating dynamic behavioral baselines for multiple power quality parameters; generating adaptive floating thresholds for each power quality parameter based on the dynamic behavioral baselines; monitoring power quality parameters in real time; triggering a diagnostic event if any power quality parameter exceeds its corresponding adaptive floating threshold; capturing a multi-dimensional dataset corresponding to the trigger time point of the power quality parameter based on the diagnostic event; analyzing and processing the multi-dimensional dataset using a preset root cause inference engine; and obtaining the current power quality diagnostic results of the power grid. This method improves operation and maintenance efficiency by using a Class B three-phase DIN rail energy meter for power quality fault diagnosis, meeting the needs of high-voltage and high-efficiency power grids for rapid and intelligent operation and maintenance. Attached Figure Description

[0036] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 A flowchart illustrating the adaptive power quality diagnostic method for power grids provided in this application embodiment. Figure One ;

[0038] Figure 2 A flowchart illustrating the adaptive power quality diagnostic method for power grids provided in this application embodiment. Figure Two ;

[0039] Figure 3 A flowchart illustrating the adaptive power quality diagnostic method for power grids provided in this application embodiment. Figure Three ;

[0040] Figure 4Flowchart of power quality adaptive diagnosis method of power grid provided by the embodiment of the present application Figure Four ;

[0041] Figure 5 Structural diagram of power adaptive diagnosis device of power grid provided by the embodiment of the present application

[0042] Figure 6 Structural diagram of B-level three-phase rail type electric energy meter provided by the embodiment of the present application. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme of the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. The components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0044] Before introducing the technical scheme of the present application, first explain the professional terms involved:

[0045] The B-level three-phase rail type electric energy meter includes a data acquisition module, a processor, a memory, a clock module (i.e. a high-precision clock module), and a communication module.

[0046] The data acquisition module includes a high-precision analog-to-digital converter and a digital input / output interface, wherein the high-precision analog-to-digital converter is used to sample the three-phase voltage (Ua, Ub, Uc) and three-phase current (Ia, Ib, Ic) of the power grid at a high sampling rate, and the digital input / output interface is used to access the switching quantity state signals of external key equipment, such as circuit breaker closing and opening position, motor contactor state, etc.

[0047] The processor adopts a heterogeneous architecture of "CPU + DSP + NPU", the central processing unit (CPU) is responsible for general task scheduling and communication, the digital signal processor (DSP) is used for real-time calculation of high-speed waveform data (such as FFT, sequence component), and the neural processing unit (NPU) is used for efficient execution of lightweight artificial intelligence (AI) model.

[0048] Memory: including Random Access Memory (RAM) and Flash. RAM is used as program running memory and ring buffer for high-resolution waveform data, and Flash is used to solidify operating system, algorithm program, dynamic baseline parameters, and store final diagnostic report and waveform file.

[0049] High-precision clock module: by implementing Precision Time Protocol (PTP) or receiving B code, Global Positioning System (GPS) signal, Beidou Pulse Per Second (PPS) signal, to ensure that the device has a unified and accurate to the microsecond level time reference. The clock is used to stamp high-precision timestamps for all collected data (such as waveform sampling points, switching quantity displacement).

[0050] Communication module: provides Ethernet, RS485, 4G / 5G, etc. Communication mode, used to upload the diagnostic report to the cloud platform or Supervisory Control And Data Acquisition (SCADA) system, and receive remote configuration and algorithm update.

[0051] The "double-high" power grid, while optimizing the energy structure, also brings frequent and rapid switching actions of power electronic devices and high sensitivity to disturbances, making the frequency and complexity of short-time, transient power quality events such as voltage sag, swell, and short-time interruption in the power grid significantly increase, and their waveform characteristics are much more complex than similar events caused by motor starting, short-circuit faults, etc. in traditional power grids. Facing this challenge, although the B-class three-phase rail-type electric energy meter has the function of monitoring and recording power quality events, its function level mainly stays at the sensing stage, i.e., it can record basic parameters such as the occurrence time, amplitude, and duration of events according to preset standard thresholds, and form an event list, but it cannot realize power quality fault diagnosis. Therefore, current operation and maintenance personnel still need to perform a large amount of manual comparison and analysis combining power grid topology, protection action signals, SCADA data, and other multi-source information, which is time-consuming and laborious, seriously restricting the rapid positioning and recovery of faults, and is difficult to meet the urgent needs of rapid intelligent operation and maintenance of the "double-high" power grid.

[0052] Based on this, the application adopts a B-level three-phase rail type electric energy meter to perform fault diagnosis on power quality through adaptive capture features combined with an algorithm for intelligently diagnosing event sources, solves the problem of insufficient diagnosis capability of the B-level three-phase rail type electric energy meter under a complex power grid background, improves operation and maintenance efficiency, meets the demand of a 'double-high' power grid for rapid intelligent operation and maintenance, and realizes intelligent upgrading from event alarm to source insight.

[0053] The technical solutions of the application will be described below in combination with several embodiments.

[0054] Figure 1 Flowchart of the power quality adaptive diagnosis method of the power grid provided by the embodiments of the application Figure One The execution subject of the embodiments can be a B-level three-phase rail type electric energy meter.

[0055] As shown in Figure 1 The method can include:

[0056] S101, generating dynamic behavior baselines of a plurality of power quality parameters, and generating adaptive floating threshold values corresponding to the power quality parameters based on the dynamic behavior baselines.

[0057] The power quality parameters are used to describe the power quality of the current power grid, and can include voltage parameters, current parameters, harmonic parameters, etc. of the power grid, for example. The harmonic parameters can include a harmonic frequency spectrum, a total harmonic distortion rate, a single harmonic content rate, a K factor, etc. The harmonic frequency spectrum can be used to analyze the amplitude and phase of each harmonic. The total harmonic distortion rate is the ratio of the effective value of all harmonic components to the effective value of the fundamental wave. The single harmonic content rate is the percentage of the amplitude of a certain harmonic to the amplitude of the fundamental wave. The K factor is used to measure the influence degree of harmonics on the heating effect of a transformer.

[0058] One power quality parameter corresponds to one dynamic behavior baseline. The dynamic behavior baseline of the power quality parameter is used to indicate a reference value of the power quality parameter, such as a voltage reference value or a current reference value. In some embodiments, the reference value can be a range including an upper limit reference value and a lower limit reference value.

[0059] Based on the dynamic behavior baseline, an adaptive floating threshold value corresponding to each power quality parameter is generated. If the dynamic behavior baseline corresponds to one reference value, the adaptive floating threshold value is one and is the reference value. If the dynamic behavior baseline corresponds to the upper limit reference value and the lower limit reference value, the adaptive floating threshold value includes an adaptive floating upper limit value and an adaptive floating lower limit value. The adaptive floating upper limit value is the upper limit reference value, and the adaptive floating lower limit value is the lower limit reference value. That is, the adaptive floating threshold value is a range, thereby improving the capture sensitivity to weak disturbances and event precursors based on the adaptive floating threshold value.

[0060] It should be noted that the number of reference values corresponding to the dynamic behavior baseline can be determined by the type of power quality parameter itself, and the present embodiment does not particularly limit this.

[0061] S102, real-time monitoring of power quality parameters, if any power quality parameter exceeds the corresponding adaptive floating threshold, a diagnostic event is triggered.

[0062] The current parameter value of the power quality parameter of the power grid is collected in real time, and it is judged whether the current parameter value of the power quality parameter exceeds the respective corresponding adaptive floating threshold, for example, the adaptive floating threshold includes the adaptive floating upper limit value and the adaptive floating lower limit value, and the current parameter value is greater than the adaptive floating upper limit value or less than the adaptive floating lower limit value, then it is determined that the respective corresponding adaptive floating threshold is exceeded, and for another example, the adaptive floating threshold is one, and the current parameter value is greater than the adaptive floating threshold, then it is determined that the respective corresponding adaptive floating threshold is exceeded, that is, any power quality parameter exceeds the corresponding adaptive floating threshold, then a diagnostic event is triggered.

[0063] That is, the B-level three-phase rail type electric energy meter is switched from the daily monitoring mode to the deep diagnosis mode, for example, the current voltage value of the voltage parameter exceeds the adaptive floating threshold corresponding to the voltage parameter, then a diagnostic event is triggered.

[0064] S103, based on the diagnostic event, capturing the multi-dimensional data set corresponding to any power quality parameter trigger time point.

[0065] Among them, the power quality parameter trigger time point is the time point of collecting the current parameter value of any power quality parameter, and the multi-dimensional data set corresponding to any power quality parameter trigger time point can be a multi-dimensional data set within a set time length containing the trigger time point, for example, a multi-dimensional data set within a set time length before the trigger time point.

[0066] After triggering the diagnostic event, based on the diagnostic event, the multi-dimensional data set is obtained from the B-level three-phase rail type electric energy meter, and the multi-dimensional data set includes: three-phase current (a phase, b phase, c phase) waveform data, three-phase voltage waveform data, and waveform data time synchronization switch state data.

[0067] Among them, the three-phase current waveform data is the three-phase current waveform data of the power grid, which is used to indicate the waveform of the three-phase current changing with time, the three-phase voltage waveform data is the three-phase voltage waveform data of the power grid, which is used to indicate the waveform of the three-phase voltage changing with time, and the waveform data time synchronization switch state data is used to indicate the state of the switch which is completely aligned in time with the three-phase current waveform data and the three-phase voltage waveform data. The electrical switch may, for example, include circuit breakers, protection relays, disconnectors, contactors, etc., and the state of the switch includes the split state (open state) and the closed state (closed state).

[0068] It should be noted that in this step, the clock module of the Class B three-phase rail-mounted energy meter uses a high-precision clock synchronization protocol to ensure that the waveform data and switch status data have a unified and microsecond-accurate timestamp, so as to achieve accurate causal correlation analysis.

[0069] S104. Use a preset root cause inference engine to analyze and process the multi-dimensional dataset to obtain the power quality diagnosis results of the current power grid.

[0070] The preset root cause inference engine can integrate preset power quality diagnosis rules. It uses these rules to analyze and process multi-dimensional datasets to extract corresponding data features from each dimension, such as voltage sag, duration, phase jump, harmonic spectrum, and oscillation frequency. Based on these data features, it performs power quality diagnosis on the current power grid. For example, if a voltage sag is accompanied by a surge in phase b current and the circuit breaker is open, then a phase b short circuit fault is determined.

[0071] In some embodiments, the preset root cause inference engine may also integrate a large language model to generate prompt words based on multi-dimensional datasets and power quality diagnostic information, and use the large language model to perform power quality diagnosis on the current power grid based on the prompt words to obtain the power quality diagnosis result of the current power grid.

[0072] In this embodiment, the built-in root cause inference engine is used to comprehensively analyze multi-dimensional datasets to infer the root cause of the current power grid fault, which solves the problem of insufficient diagnostic capabilities in the context of complex power grids and realizes the intelligent upgrade from event alarm to root cause insight.

[0073] In an optional implementation, step S104 above, which uses a preset root cause inference engine to analyze and process a multi-dimensional dataset to obtain the current power quality diagnosis results of the power grid, may include:

[0074] The system uses a pre-defined diagnostic model in the root cause inference engine to analyze and process multi-dimensional datasets, and outputs the power quality diagnostic results of the current power grid.

[0075] The preset diagnostic model is trained using a power grid sample dataset. The power grid sample data includes the mapping relationship between the multi-dimensional data features of known power grid event samples and the root causes of the events. The known power grid event samples are samples of known power grid fault event root causes. The multi-dimensional data features of the known power grid event samples can include the three-phase current waveform data, the three-phase voltage waveform data, and the switching status data of the waveform data time synchronization of the known power grid event samples.

[0076] Given that the multi-dimensional data features of known power grid event samples are labeled with corresponding event root causes, a pre-defined diagnostic model is obtained by training the power grid sample dataset.

[0077] The preset root inference engine integrates a preset diagnosis model, takes the multi-dimensional data benchmark as input of the preset diagnosis model, analyzes and processes the multi-dimensional data set by using the preset diagnosis model, and outputs the power quality diagnosis result of the power grid.

[0078] In the embodiment, the preset diagnosis model is trained by learning a large number of samples, so that more detailed quality diagnosis results such as “control oscillation during low-voltage ride-through of a photovoltaic inverter” and “DC bus voltage drop caused by fault of a rectifier side of a frequency converter” are given by using the preset diagnosis model, thereby improving the power quality diagnosis accuracy.

[0079] Figure 2 Flowchart of the power quality adaptive diagnosis method of the power grid provided in the embodiments of the present application Figure Two As shown in the optional embodiment, the step S101 of generating the dynamic behavior baseline of the plurality of power quality parameters and generating the adaptive floating threshold corresponding to each power quality parameter based on the dynamic behavior baseline can include: Figure 2

[0080] S201, for each power quality parameter, the corresponding historical data is counted by using a sliding time window.

[0081] The sliding time window refers to a preset time window before the current time, for example, the past 24 hours or 48 hours.

[0082] For each power quality parameter, the historical data corresponding to each power quality parameter in the sliding time window is counted, wherein the historical data is the parameter value of each power quality parameter collected in the sliding time window.

[0083] S202, according to the historical data corresponding to each power quality parameter, the dynamic average value and the dynamic standard deviation of each power quality parameter are calculated by using an exponential weighted moving average algorithm.

[0084] The core of the exponential weighted moving average algorithm is to give the historical data a gradually decaying weight. The closer to the current time, the higher the weight and the slower the decay, the historical data has a more lasting impact. The farther from the current time, the lower the weight and the faster the decay, the historical data has a less lasting impact.

[0085] The dynamic average value of the power quality parameter is used to indicate the average situation of the power quality parameter, and the dynamic standard deviation of the power quality parameter is used to indicate the fluctuation intensity of the power quality parameter around the dynamic average value. The dynamic average value and the dynamic standard deviation of the power quality parameter are calculated by using the exponential weighted moving average algorithm, so that the dynamic average value and the dynamic standard deviation of the power quality parameter can better represent the current state. ​

[0086] For example, the historical data includes a parameter value 1 collected at a historical time 1 and a parameter value 2 collected at a historical time 2, where the historical time 1 is closer to the current time than the historical time 2, and a and 1-a respectively represent corresponding weights, and a is greater than 1-a. Then, the dynamic average value μ t of the power quality parameter can be represented as:

[0087] μ t = a x parameter value 1 + (1-a) x parameter value 2

[0088] Similarly, the historical time 1 is closer to the current time than the historical time 2, and β and 1-β respectively represent corresponding weights, and β is greater than 1-β. Then, the dynamic standard deviation σ t is represented as:

[0089] σ t = sqrt(β x (parameter value 1-μ t) + (1-β) x (parameter value 2-μ t))

[0090] Wherein, sqrt represents the square root operation, and β can be equal to a or not equal to a. The present embodiment does not particularly limit this.

[0091] S203, generating a dynamic behavior baseline of each power quality parameter according to the dynamic average value and the dynamic standard deviation of each power quality parameter.

[0092] In some embodiments, the dynamic behavior baseline of the power quality parameter is a range, which can be represented as [dynamic average value-dynamic standard deviation, dynamic average value+dynamic standard deviation], for example. The upper limit value of the baseline is dynamic average value+dynamic standard deviation, and the lower limit value of the baseline is dynamic average value-dynamic standard deviation.

[0093] In some embodiments, the dynamic behavior baseline can also be represented as [dynamic average value-k x dynamic standard deviation, dynamic average value+k x dynamic standard deviation], where the upper limit value of the baseline is dynamic average value+k x dynamic standard deviation, and the lower limit value of the baseline is dynamic average value-k x dynamic standard deviation. Wherein, k is a constant, which is selected according to actual conditions, for example, 3, 4 or 5.

[0094] In some embodiments, the dynamic behavior baseline of the power quality parameter can also correspond to a baseline value, and the sum of the dynamic average value and the dynamic standard deviation of the power quality parameter can be determined as the baseline value.

[0095] S204, generating an adaptive floating threshold corresponding to each power quality parameter based on the dynamic behavior baseline.

[0096] In some embodiments, the upper limit value of the baseline of the dynamic behavior baseline is determined as the adaptive floating upper limit value, and the lower limit value of the baseline of the dynamic behavior baseline is determined as the adaptive floating lower limit value. Wherein, the adaptive floating threshold corresponding to each power quality parameter includes the adaptive floating upper limit value and the adaptive floating lower limit value.

[0097] In some embodiments, the one reference value corresponding to the dynamic behavior baseline is determined as an adaptive floating threshold.

[0098] In the present embodiment, since the dynamic average value and the dynamic standard deviation are constantly updated over time, the dynamic behavior baseline of each power quality parameter is also constantly moving and scaling smoothly. When the power grid is running smoothly, the dynamic standard deviation is small, the dynamic behavior baseline range is narrow, and the monitoring is sensitive. When the power grid itself fluctuates greatly, the dynamic standard deviation will become larger, and the dynamic behavior baseline range will become wider, thereby improving the capture sensitivity of weak disturbances and precursors based on the adaptive floating threshold, and preventing inaccurate power quality diagnosis. In addition, by limiting the sliding time window and the exponential weighted moving average algorithm, it is ensured that the dynamic behavior baseline is determined by the recent behavior and is not affected by the behavior far from the current, and the calculation load is reduced, without the need to store a large amount of historical data, only the data within the window needs to be processed, and the calculation efficiency is high.

[0099] In an optional implementation, the step S204 of generating the adaptive floating threshold corresponding to each power quality parameter based on the dynamic behavior baseline can include:

[0100] The adaptive floating threshold corresponding to each power quality parameter is obtained by subtracting a preset multiple of the dynamic standard deviation from the dynamic average value of the power quality parameter.

[0101] If the power quality parameter is a voltage parameter, the adaptive floating threshold corresponding to the power quality parameter can be the dynamic average value of the power quality parameter minus a preset multiple of the dynamic standard deviation, where the preset multiple can be, for example, 2, 3, 4, -5.

[0102] For example, if the power quality parameter is a voltage parameter, the adaptive floating threshold corresponding to the voltage parameter in the sag state can be expressed as the dynamic average value minus 4 times the dynamic standard deviation. For example, if the power quality parameter is a current parameter, the adaptive floating threshold corresponding to the current parameter in the harmonic state can be expressed as the dynamic average value plus 5 times the dynamic standard deviation.

[0103] In some embodiments, the preset multiple can be determined according to the current power grid operating condition. For example, under peak load conditions, a large number of impact loads such as air conditioners, elevators, water pumps, and industrial equipment are started at the same time, resulting in large fluctuations in power quality parameters, so the preset multiple is relatively large, such as 5. Under base load conditions, the fluctuations of power quality parameters are small, so the preset multiple is relatively small, such as 2.

[0104] In the present embodiment, the adaptive floating threshold can closely follow the changes in the power grid operating mode, and realize true adaptation.

[0105] Figure 3Flowchart of power quality adaptive diagnosis method of power grid provided by the embodiment of the application Figure Three As shown in the optional embodiment, the method can further include: Figure 3

[0106] S301, using a preset ring buffer to temporarily store high-resolution three-phase current waveform data, three-phase voltage waveform data and waveform data time-synchronized switching value state data within a preset time period.

[0107] The preset time period can be a pre-set fixed time period, and the preset ring buffer only retains the most recent preset time period, for example, 24 hours or 48 hours.

[0108] The data acquisition module is used to collect high-resolution three-phase current waveform data, three-phase voltage waveform data and waveform data time-synchronized switching value state data in real time, and the collected data is constantly written into the preset ring buffer, so that the preset ring buffer only retains the most recent preset time period of three-phase current waveform data, three-phase voltage waveform data and waveform data time-synchronized switching value state data, that is, the data exceeding the preset time period is covered by newly collected data.

[0109] The step S103 above, based on the diagnosis event, captures a multi-dimensional data set corresponding to any power quality parameter trigger time point, which can include:

[0110] S302, based on the diagnosis event, locking the preset ring buffer and capturing a multi-dimensional data set of each preset time period before and after the trigger time point of any power quality parameter.

[0111] The preset time period before and after the trigger time point can be, for example, 1 hour or 2 hours. Based on the diagnosis event, the preset ring buffer is locked to prevent newly collected data from covering the multi-dimensional data set of each preset time period before and after the trigger time point of any power quality parameter, and to obtain the multi-dimensional data set of each preset time period before and after the trigger time point of any power quality parameter from the preset buffer.

[0112] In this embodiment, the preset ring buffer is used to provide the necessary data for analyzing the root cause of the fault, while avoiding massive data storage, thereby improving storage efficiency and analysis pertinence.

[0113] In an optional embodiment, the multi-dimensional data set further includes a positive sequence, negative sequence or zero sequence current time sequence curve calculated and obtained according to the three-phase current waveform data.

[0114] Or, a positive sequence, negative sequence or zero sequence voltage time sequence curve calculated and obtained according to the three-phase voltage waveform data.

[0115] ​The positive sequence current time sequence curve is a curve composed of the positive sequence currents at each sampling moment extracted from the three-phase current waveform data, that is, the positive sequence currents at each sampling moment are connected in time sequence to obtain the positive sequence current time sequence curve. The positive sequence current represents the normal energy transmission part of the power grid, which drives all three-phase motors, transformers and other equipment to rotate and work normally, and only the positive sequence current exists in an ideal balanced system.

[0116] The negative sequence current time sequence curve is a curve composed of the negative sequence currents at each sampling moment extracted from the three-phase current waveform data, that is, the negative sequence currents at each sampling moment are connected in time sequence to obtain the negative sequence current time sequence curve. The negative sequence current is mainly caused by three-phase load imbalance, broken line, asymmetric fault (such as two-phase short circuit), and the negative sequence current is extremely harmful to the motor because it can generate a reverse rotating magnetic field, causing the motor to overheat, vibrate and reduce efficiency.

[0117] The zero sequence current time sequence curve is a curve composed of the zero sequence currents at each sampling moment extracted from the three-phase current waveform data, that is, the zero sequence currents at each sampling moment are connected in time sequence to obtain the zero sequence current time sequence curve. The zero sequence current refers to the current formed when a single-phase ground fault occurs, which flows back through the ground or neutral line.

[0118] The three-phase currents (Ia, Ib, Ic) at each sampling moment are determined according to the three-phase current waveform data, and a preset transformation matrix is used to calculate the positive sequence current, negative sequence current and zero sequence current at each sampling moment according to the three-phase currents at each sampling moment, and then the positive sequence current time sequence curve, negative sequence current time sequence curve and zero sequence current time sequence curve are generated.

[0119] Similarly, the positive sequence voltage time sequence curve is a curve composed of the positive sequence voltages at each sampling moment extracted from the three-phase voltage waveform data, that is, the positive sequence voltages at each sampling moment are connected in time sequence to obtain the positive sequence voltage time sequence curve. The positive sequence voltage represents the normal energy transmission part of the power grid, which drives all three-phase motors, transformers and other equipment to rotate and work normally, and only the positive sequence voltage exists in an ideal balanced system.

[0120] The negative sequence voltage time sequence curve is a curve composed of the negative sequence voltages at each sampling moment extracted from the three-phase voltage waveform data, that is, the negative sequence voltages at each sampling moment are connected in time sequence to obtain the negative sequence voltage time sequence curve. The negative sequence voltage refers to the voltage caused by load imbalance, asymmetric fault, etc.

[0121] The zero-sequence voltage time-series curve is a curve composed of the zero-sequence voltage at each sampling time extracted from the three-phase voltage waveform data, that is, the zero-sequence voltage at each sampling time is connected in time sequence to obtain the zero-sequence voltage time-series curve. The zero-sequence voltage refers to the voltage caused by single-phase ground fault and the like.

[0122] According to the three-phase voltage waveform data, the three-phase voltage (Ua, Ub, Uc) at each sampling time is determined, and a preset transformation matrix is used to calculate the positive-sequence voltage, the negative-sequence voltage and the zero-sequence voltage at each sampling time according to the three-phase voltage at each sampling time, and then the positive-sequence voltage time-series curve, the negative-sequence voltage time-series curve and the zero-sequence voltage time-series curve are generated.

[0123] Figure 4 Flowchart of the power quality adaptive diagnosis method of the power grid provided by the embodiments of the present application Figure Four As shown in Figure 4 In an optional embodiment, the step S104 uses the preset root cause inference engine to analyze and process the multi-dimensional data set to obtain the power quality diagnosis result of the current power grid, including:

[0124] S401, using a preset root cause inference engine to check the current time-series curve and the voltage time-series curve, if there is a current or voltage increase amplitude greater than a preset amplitude, and the phase difference is in a preset interval near 180 degrees, then diagnose the current power grid as an upstream power grid fault; if there is a current or voltage increase amplitude greater than a preset amplitude, and the phase difference is in a preset interval near 0 degrees, then diagnose the current power grid as a downstream user side fault.

[0125] The preset interval near 180 degrees can be a phase interval centered at 180 degrees, for example, an interval of 170 degrees to 190 degrees, and the preset interval near 0 degrees can be a phase interval centered at 0 degrees, for example, an interval of -10 degrees to 10 degrees.

[0126] The positive current of the power grid is defined as the current from the upstream power grid to the downstream user. If there is a downstream user side fault, the current and the positive current are in the same direction, and the phase difference between the current and the positive current is close to 0 degrees. If there is an upstream power grid side fault, the current and the positive current are in opposite directions, from the downstream user to the upstream power grid, and the phase difference between the current and the positive current is close to 180 degrees.

[0127] The preset root cause inference engine is integrated with power quality diagnosis rules. By using the preset root cause inference engine, the positive sequence current time sequence curve, the negative sequence current time sequence curve, the zero sequence current time sequence curve, the positive sequence voltage time sequence curve, the negative sequence voltage time sequence curve and the zero sequence voltage time sequence curve are checked. If the current at any sampling time on any current time sequence curve increases by more than a preset amplitude (compared to the current value at the previous sampling time, the fluctuation is large), and the phase difference with the positive current is within the preset interval near 180 degrees, it is diagnosed that the current power grid is an upstream power grid fault.

[0128] If the voltage at any sampling time on any voltage time sequence curve increases by more than a preset amplitude (compared to the voltage value at the previous sampling time, the fluctuation is large), and the current at the sampling time and the positive current have a phase difference within the preset interval near 180 degrees, it is diagnosed that the current power grid is an upstream power grid fault.

[0129] Similarly, if the current at any sampling time on any current time sequence curve increases by more than a preset amplitude (compared to the current value at the previous sampling time, the fluctuation is large), and the phase difference with the positive current is within the preset interval near 0 degrees, it is diagnosed that the current power grid is a downstream user side fault.

[0130] If the voltage at any sampling time on any voltage time sequence curve increases by more than a preset amplitude (compared to the voltage value at the previous sampling time, the fluctuation is large), and the current at the sampling time and the positive current have a phase difference within the preset interval near 0 degrees, it is diagnosed that the current power grid is a downstream user side fault.

[0131] Wherein, the upstream power grid fault may be, for example, C-phase upstream new energy station inverter control oscillation, and the downstream user side fault may be, for example, user side 2# transformer feeder branch high-power motor starting.

[0132] In the embodiment, normal voltage or current fluctuations are filtered out by the preset interval, and the current power grid is accurately diagnosed by the phase relationship, solving the problem of insufficient diagnosis capability of the traditional monitoring device under the complex power grid background, and realizing the intelligent upgrade from event alarm to root cause insight.

[0133] Figure 5 The structure diagram of the power grid power self-adaptive diagnosis device provided in the embodiment of the application, which can be integrated in a B-level three-phase guide rail type electric energy meter.

[0134] As shown in Figure 5 The device can include:

[0135] The generating module 501 is configured to generate a dynamic behavior baseline of a plurality of power quality parameters, and generate an adaptive floating threshold corresponding to each power quality parameter based on the dynamic behavior baseline.

[0136] The monitoring module 502 is configured to monitor the power quality parameters in real time, and trigger a diagnosis event if any power quality parameter exceeds the corresponding adaptive floating threshold value;

[0137] The acquisition module 503 is configured to capture a multi-dimensional data set corresponding to a triggering time point of any power quality parameter based on the diagnosis event, and the multi-dimensional data set includes three-phase current waveform data, three-phase voltage waveform data, and switch state data synchronized with the waveform data in time;

[0138] The acquisition module 503 is further configured to analyze and process the multi-dimensional data set by using a preset root inference engine, and obtain a power quality diagnosis result of the current power grid.

[0139] In an optional implementation, the generation module 501 is specifically configured to:

[0140] For each power quality parameter, a sliding time window is used to statistically analyze the corresponding historical data;

[0141] According to the historical data of each power quality parameter, an exponential weighted moving average algorithm is used to calculate the dynamic average value and the dynamic standard deviation of each power quality parameter;

[0142] According to the dynamic average value and the dynamic standard deviation of each power quality parameter, a dynamic behavior baseline of each power quality parameter is generated;

[0143] Based on the dynamic behavior baseline, an adaptive floating threshold value corresponding to each power quality parameter is generated.

[0144] In an optional implementation, the generation module 501 is specifically configured to:

[0145] The dynamic average value of each power quality parameter is subtracted by a preset multiple of the dynamic standard deviation to obtain the adaptive floating threshold value corresponding to the power quality parameter.

[0146] In an optional implementation, the apparatus further includes:

[0147] The storage module 504 is configured to use a preset ring buffer to temporarily store high-resolution three-phase current waveform data, three-phase voltage waveform data, and switch state data synchronized with the waveform data in time within a preset time length in real time;

[0148] The acquisition module 503 is specifically configured to lock the preset ring buffer based on the diagnosis event, and capture a multi-dimensional data set within each preset time period before and after the triggering time point of any power quality parameter.

[0149] In an optional implementation, the multi-dimensional data set further includes a positive sequence, a negative sequence, or a zero sequence current time sequence curve calculated based on the three-phase current waveform data;

[0150] Alternatively, the positive sequence, negative sequence or zero sequence voltage time series curve is calculated according to the three-phase voltage waveform data.

[0151] In an optional implementation, the obtaining module 503 is specifically configured to:

[0152] The preset root inference engine is used to check the current time series curve and the voltage time series curve, if there is current or voltage increase amplitude greater than the preset amplitude, and the phase difference is in the preset interval near 180 degrees, it is diagnosed that the current power grid is an upstream power grid fault; if there is current or voltage increase amplitude greater than the preset amplitude, and the phase difference is in the preset interval near 0 degrees, it is diagnosed that the current power grid is a downstream user side fault.

[0153] In an optional implementation, the obtaining module 503 is specifically configured to:

[0154] The preset diagnostic model in the preset root inference engine is used to analyze and process the multi-dimensional data set, and output the power quality diagnosis result of the current power grid, wherein the preset diagnostic model is obtained by training using power grid sample data set, and the mapping relationship between the multi-dimensional data features of the known power grid event sample and the event root source.

[0155] It should be noted that the generating module 501 and the obtaining module 503 can be implemented by the processor of the B-level three-phase guide rail type electric energy meter, the monitoring module 502 can be implemented by the data acquisition module of the B-level three-phase guide rail type electric energy meter, and the storage module 504 can be implemented by the storage of the B-level three-phase guide rail type electric energy meter.

[0156] The description of the processing flow of each module in the device and the interaction flow between the modules can refer to the related description in the above method embodiments, and will not be described in detail here.

[0157] Figure 6 The structure diagram of the B-level three-phase guide rail type electric energy meter provided by the embodiments of the present application is shown in FIG. 1. Figure 6 As shown in FIG. 1, the B-level three-phase guide rail type electric energy meter includes a data acquisition module 601, a processor 602, a memory 603, a clock module 604 and a communication module 605. When the B-level three-phase guide rail type electric energy meter is running, the data acquisition module 601, the processor 602, the clock module 604 and the communication module 605 execute the machine readable instructions stored in the memory to execute the above method.

[0158] The embodiments of the present application also provide a computer readable storage medium, and the computer readable storage medium stores a computer program, and the computer program is run by a processor to execute the above method.

[0159] In the embodiments of the present application, the computer program, when executed by the processor, can also execute other machine readable instructions to perform the methods as described in the embodiments, for the specific method steps and principles performed, refer to the description of the embodiments, which will not be described in detail here.

[0160] In the embodiments provided in the present application, it should be understood that the disclosed device and method can be implemented by other means. The device embodiments described above are only illustrative, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, and for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units or components shown or discussed can be indirect coupling or communication connection through some communication interfaces, devices or units, which can be electrical, mechanical or other forms.

[0161] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or they can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0162] In addition, the functional units in the embodiments provided in the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit.

[0163] The functions, if realized in the form of software functional units and sold or used as independent products, can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the part of the present application which essentially contributes to the prior art or the part of the technical solutions can be embodied in the form of software products. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various program code storage media.

[0164] It should be noted that like reference numerals and letters refer to like items throughout the accompanying drawings, and once an item is defined in one drawing, it should not be further defined and explained in subsequent drawings, and further, the terms "first", "second", "third" and the like are used only to distinguish descriptions, and cannot be understood as indicating or implying relative importance.

[0165] Finally, it should be noted that the above-described embodiments are merely specific embodiments of the present application, and are used to illustrate the technical solutions of the present application, but are not limiting, and the protection scope of the present application is not limited thereto, although the present application is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can modify or easily think of changes to the technical solutions recorded in the foregoing embodiments, or make equivalent replacements to some technical features, and these modifications, changes or replacements do not make the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application. should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An adaptive power quality diagnostic method for a power grid, characterized in that, The method, applied to a Class B three-phase rail-mounted energy meter, includes: A dynamic behavior baseline for multiple power quality parameters is generated, and an adaptive floating threshold corresponding to each power quality parameter is generated based on the dynamic behavior baseline. The multiple power quality parameters include: voltage parameters, current parameters, and harmonic parameters. The power quality parameters are monitored in real time. If any power quality parameter exceeds the corresponding adaptive floating threshold, a diagnostic event is triggered. Based on the diagnostic event, a multi-dimensional dataset corresponding to the trigger time point of any power quality parameter is captured. The multi-dimensional dataset includes: three-phase current waveform data, three-phase voltage waveform data, and switch status data synchronized with the waveform data time. The multi-dimensional dataset is analyzed and processed using a preset root cause inference engine to obtain the power quality diagnosis results of the current power grid. The multi-dimensional dataset also includes: positive-sequence, negative-sequence, or zero-sequence current time-sequence curves calculated based on the three-phase current waveform data; Alternatively, positive-sequence, negative-sequence, or zero-sequence voltage timing curves can be calculated based on the three-phase voltage waveform data. The process of using a preset root cause inference engine to analyze and process the multi-dimensional dataset to obtain the current power quality diagnostic results of the power grid includes: The current timing curve and the voltage timing curve are checked using the preset root cause inference engine. If the increase in current or voltage is greater than a preset magnitude and the phase difference is within a preset range of 180 degrees, the current power grid is diagnosed as an upstream power grid fault. If the increase in current or voltage is greater than the preset magnitude and the phase difference is within a preset range of 0 degrees, the current power grid is diagnosed as a downstream user-side fault.

2. The method according to claim 1, characterized in that, The process of generating a dynamic behavior baseline for multiple power quality parameters and generating an adaptive floating threshold for each power quality parameter based on the dynamic behavior baseline includes: For each of the aforementioned power quality parameters, a sliding time window is used to statistically analyze the corresponding historical data. Based on the historical data corresponding to each of the power quality parameters, the dynamic average value and dynamic standard deviation of each power quality parameter are calculated using an exponentially weighted moving average algorithm. Based on the dynamic average value and dynamic standard deviation of each power quality parameter, a dynamic behavior baseline for each power quality parameter is generated. Based on the dynamic behavior baseline, adaptive floating thresholds are generated for each of the power quality parameters.

3. The method according to claim 2, characterized in that, The step of generating adaptive floating thresholds corresponding to each of the power quality parameters based on the dynamic behavior baseline includes: The adaptive floating threshold corresponding to each power quality parameter is obtained by subtracting the dynamic standard deviation of a preset multiple from the dynamic average value of each power quality parameter.

4. The method according to claim 1, characterized in that, The method further includes: A preset circular buffer is used to temporarily store high-resolution three-phase current waveform data, three-phase voltage waveform data, and switch status data synchronized with the waveform data in real time within a preset duration. The step of capturing a multi-dimensional dataset corresponding to the trigger time point of any power quality parameter based on the diagnostic event includes: Based on the diagnostic event, the preset circular buffer is locked, and a multi-dimensional dataset of one preset time period before and after the trigger time of any power quality parameter is captured.

5. The method according to claim 1, characterized in that, The process of using a preset root cause inference engine to analyze and process the multi-dimensional dataset to obtain the current power quality diagnostic results of the power grid includes: The multi-dimensional dataset is analyzed and processed using the preset diagnostic model in the preset root cause inference engine, and the power quality diagnosis result of the current power grid is output. The preset diagnostic model is trained and obtained using a power grid sample dataset, which includes the mapping relationship between the multi-dimensional data features of known power grid event samples and the event root causes.

6. A power quality adaptive diagnostic device for a power grid, characterized in that, include: A generation module is used to generate dynamic behavior baselines for multiple power quality parameters, and generate adaptive floating thresholds for each power quality parameter based on the dynamic behavior baselines. The multiple power quality parameters include: voltage parameters, current parameters, and harmonic parameters. The monitoring module is used to monitor the power quality parameters in real time. If any power quality parameter is detected to exceed the corresponding adaptive floating threshold, a diagnostic event is triggered. The acquisition module is used to capture a multi-dimensional dataset corresponding to the trigger time point of any power quality parameter based on the diagnostic event. The multi-dimensional dataset includes: three-phase current waveform data, three-phase voltage waveform data, and switch status data synchronized with the waveform data time. The acquisition module is also used to analyze and process the multi-dimensional dataset using a preset root cause inference engine to obtain the power quality diagnosis results of the current power grid. The multi-dimensional dataset also includes: positive-sequence, negative-sequence, or zero-sequence current time-sequence curves calculated based on the three-phase current waveform data; Alternatively, positive-sequence, negative-sequence, or zero-sequence voltage timing curves can be calculated based on the three-phase voltage waveform data. The acquisition module is specifically used to check the current timing curve and the voltage timing curve using the preset root cause inference engine. If the increase in current or voltage is greater than a preset magnitude and the phase difference is within a preset range of 180 degrees, the current power grid is diagnosed as an upstream power grid fault. If the increase in current or voltage is greater than the preset magnitude and the phase difference is within a preset range of 0 degrees, the current power grid is diagnosed as a downstream user-side fault.

7. A Class B three-phase rail-mounted energy meter, characterized in that, include: The data acquisition module, processor, memory, clock module, and communication module, when the Class B three-phase rail-mounted energy meter is running, execute machine-readable instructions stored in the memory to perform the method described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the method according to any one of claims 1 to 5.

Images