An active power distribution network power quality panoramic sensing and abnormal positioning management system

Abstract

The present application relates to the technical field of power transmission line detection, and more particularly to the technical field of active distribution network power quality monitoring and treatment, and more particularly to an active distribution network power quality panoramic sensing and abnormal positioning treatment system, comprising a monitoring terminal adaptation module, a monitoring point optimization configuration module, a power quality sensing and early warning module, a composite power quality tracing module, a power quality comprehensive treatment module and a distribution network dispatching sensing module; the present application realizes whole-process management and control from power quality monitoring terminal optimization configuration, to accurate sensing, positioning and tracing of problems, to comprehensive treatment and intelligent support of distribution network dispatching; effectively solves various power quality problems in the active distribution network, guarantees safe and stable operation of the power grid, improves user power quality and satisfaction, and promotes efficient and reliable development of the active distribution network.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of active power distribution network power quality monitoring and treatment, and particularly relates to an active power distribution network power quality panoramic sensing and abnormal positioning treatment system. BACKGROUND

[0002] The proportion of distributed power sources (such as photovoltaic power, wind power, etc.) in the active power distribution network is increasing; a large number of distributed power sources are connected to the power grid, and relying on local advantages to develop large industrial bases introduces many power quality interference sources, such as electric arc furnaces, intermediate frequency furnaces, etc.; this makes the power supply environment of the active power distribution network complex, the problem difference of different power supply areas increases, the nonlinearity and randomness of the source and load intensify, and thus the power quality problems such as voltage deviation, three-phase imbalance, and harmonics become increasingly prominent.

[0003] For example, the output of the distributed power source is intermittent and volatile due to the influence of natural conditions, the photovoltaic power generation has large power during the day and stops generating power at night, and the wind power generation has unstable wind speed leading to fluctuation of output power, which will all cause changes in power flow of the power grid and cause voltage fluctuation; the unbalanced distribution of single-phase distributed power sources and single-phase loads causes three-phase imbalance; the power electronic converter in the distributed power source and the nonlinear load generate a large amount of harmonics, polluting the power grid.

[0004] There are many types of current power quality monitoring devices, but each has limitations; the handheld device in the portable analyzer can only detect single-phase harmonics, and the function is single; the portable device can detect three-phase multiple indicators, but cannot monitor continuously. The discrete device in the online monitoring instrument has limited measurement channels, and the integrated device has powerful functions but high cost.

[0005] The configuration of the power quality monitoring terminal lacks systematization, and factors such as the structure of the active power distribution network, the distribution of the distributed power source, and the load characteristics are not fully considered, which leads to unreasonable layout of the monitoring points, existence of monitoring blind area, difficulty in comprehensive and accurate monitoring of the power quality problem, and inability to provide reliable data support for treatment.

[0006] The existing treatment equipment is mainly aimed at a single power quality problem and lacks comprehensive treatment ability; for example, the on-load voltage regulating transformer mainly solves the high / low voltage problem, and has poor treatment effect on three-phase imbalance and harmonics; the SVG has limited effect on improving voltage quality and three-phase imbalance, and has high cost; the APF has good treatment effect on harmonics, but has limited capacity and high cost, and is limited in application in large-scale active power distribution networks.

[0007] The treatment decision lacks overall planning, only focuses on the internal problem of a single active power distribution network, ignores the mutual influence between power distribution networks and the synergistic effect of treatment equipment, and leads to unsatisfactory treatment effect and waste of resources.

[0008] In view of this, the present application proposes a kind of active power distribution network electric energy quality panorama perception and abnormal positioning management system. SUMMARY

[0009] In order to make up for the deficiencies of the prior art, the present application aims to provide an active power distribution network electric energy quality panorama perception and abnormal positioning management system to overcome the deficiencies of the prior art and provide strong guarantee for the safe and stable operation of power transmission lines.The technical scheme adopted by the present application to solve its technical problems is as follows: an active power distribution network electric energy quality panorama perception and abnormal positioning management system, comprising a monitoring terminal adaptation module, a monitoring point optimization configuration module, an electric energy quality perception and early warning module, a composite electric energy quality tracing module, an electric energy quality comprehensive management module and a power distribution network dispatching perception module, wherein:

[0010] The monitoring terminal adaptation module is used to analyze the applicability of electric energy quality monitoring terminals to distributed power electric energy quality monitoring to select appropriate monitoring terminals.

[0011] The monitoring point optimization configuration module realizes the optimization layout of monitoring points based on an improved algorithm to ensure that voltage sag domains are observable and current information is complete, thereby assisting in electric energy quality event diagnosis and positioning.

[0012] The electric energy quality perception and early warning module is used to accurately perceive and timely warn of electric energy quality problems through model establishment and feature analysis.

[0013] The composite electric energy quality tracing module is used to analyze the causes, simulation and tracing of electric energy quality problems to provide a basis for management.

[0014] The electric energy quality comprehensive management module realizes multi-objective optimization management of electric energy quality by establishing a model through analysis of equipment influence and proposing strategies.

[0015] The power distribution network dispatching perception module is used to build a platform to realize situation awareness and support power distribution network dispatching decisions.

[0016] Preferably, in the monitoring terminal adaptation module, the electric energy quality monitoring terminal includes a portable analyzer (including handheld and portable, the handheld type is used for detecting single-phase harmonic, and the portable type can detect three-phase multiple indicators but is not suitable for continuous monitoring) and an online monitor (including discrete and integrated types, the discrete type uses single-chip microcomputer or DSP technology, and the integrated type uses industrial computer technology, and the measurement channels and functions are different), and the module determines the applicability of the monitoring terminal according to the characteristics of distributed power and electric energy quality problems.

[0017] Preferably, in the monitoring point optimization configuration module, based on the improved binary particle swarm optimization algorithm (BPSO), the particle iteration formula is modified, a new evaluation function is constructed, and a feasible region feedback checking link is added. The feasible region condition is that the number of installed PQMs is not less than NKCL. The monitoring points are optimized based on the line distribution weight under the condition of ensuring node voltage and current information, meeting the intelligent diagnosis and accurate positioning requirements of power quality events.

[0018] Preferably, in the power quality awareness and early warning module, a model considering different output scenarios of distributed power sources and load time sequence characteristics is established. By piecewise linearization, that is, using an improved important point selection method, trend characteristics are extracted. The abnormal index is determined by time sequence trend similarity and numerical outlier proportion and is weighted and combined. The weight uses the CRITIC weighting method. Combined with historical data, abnormal identification and early warning of power quality problems are realized.

[0019] Preferably, the composite power quality tracing module includes a three-phase harmonic power flow analysis unit, a power quality cause analysis unit, and a tracing and positioning unit.

[0020] The three-phase harmonic power flow analysis unit obtains a three-phase harmonic admittance matrix through matrix operation, calculates harmonic voltage using a harmonic current content rate model and an alternating iteration method, and analyzes power quality problem simulation.

[0021] The power quality cause analysis unit determines the property of the transformer area based on static information, evaluates the problem causes, identifies multi-time scale problems combined with multi-source data, analyzes the causes after extracting features, and predicts the trend of steady-state power quality indicators using an LSTM network model.

[0022] The tracing and positioning unit obtains relevant data, filters features to form a disturbance event set, calculates correlation coefficients through time series DDTW distance calculation, and matches the event set to realize composite power quality problem tracing and positioning.

[0023] Preferably, the power quality comprehensive management module includes:

[0024] The harmonic management unit analyzes the harmonic control ability of the APF management device, establishes an objective function reflecting the harmonic content of each node, total distortion rate, and management cost, and satisfies the voltage harmonic national standard requirements to reduce the impact of harmonics on the power grid and improve power quality.

[0025] The reactive power compensation unit analyzes the compensation effect of the intelligent capacitor device on the reactive power, constructs an objective function involving reactive power compensation effect, system transmission power loss, voltage deviation, and management cost, and realizes reasonable compensation of reactive power and stabilization of power grid voltage combined with power balance, unbalance constraints, power factor limit values, and capacitor capacity limit values.

[0026] The three-phase imbalance treatment unit is used for treating three-phase imbalance of the commutating switch device, sets a target function containing treatment effect, i.e., negative sequence three-phase imbalance degree, and treatment cost, i.e., the number of commutating switch actions, improves the three-phase imbalance condition under the constraint of average voltage drop of the commutatable load node, and ensures three-phase load balance.

[0027] Preferably, the power distribution network dispatching perception module comprises:

[0028] The platform building unit is configured to build a power distribution network dispatching support platform to provide a basic framework for subsequent function implementation.

[0029] The situation perception function unit is configured to add a situation perception function to the original power distribution network dispatching system, build a power distribution network panoramic model through collection of power distribution network data and operation, build a power distribution network situation perception map, and realize panoramic perception of the power distribution network operation state.

[0030] The operation platform unit is configured to build a lightweight power distribution network situation perception operation platform, which is configured to call real-time and historical situation perception maps, has data analysis and hawk-eye magnification functions, and provides strong support for power distribution network dispatching.

[0031] The data processing unit is configured to process power quality cause identification data required by power distribution area treatment decision technology, including calculation of statistical characteristic values in an event time period, including average value, maximum value, minimum value, and distribution transformer tap position and capacitor state information, and cumulative statistics of daily, weekly, monthly, and seasonal statistical characteristic values according to a preset rule (daily statistical characteristic values are based on real-time statistics of 5-minute interval measurement data, and weekly, monthly, and seasonal statistical characteristic values are based on cumulative statistics of higher-level statistical data at specific time points), and delay cause analysis is performed after the corresponding statistics are completed, thereby providing comprehensive data support and decision basis for operation and maintenance personnel.

[0032] Preferably, the self-learning and self-adaptive module comprises:

[0033] The data collection unit is configured to collect historical operation data and real-time monitoring data to provide a data basis for subsequent analysis.

[0034] The model optimization unit is configured to optimize the power quality evaluation model according to the collected data to improve evaluation accuracy.

[0035] The algorithm improvement unit is configured to improve the abnormal positioning algorithm according to data characteristics and operation conditions to improve positioning accuracy.

[0036] The strategy adjustment unit is configured to adjust the treatment decision strategy based on data analysis results to adapt to dynamic changes of the active power distribution network and ensure that the system always maintains good performance.

[0037] Preferably, the data interaction and sharing module comprises:

[0038] interface adaptation unit, responsible for interface adaptation with other power system management platforms (such as power distribution automation system, power marketing system, etc.), ensuring the compatibility of data interaction;

[0039] data transmission unit, realizing the bidirectional transmission of power quality data and equipment operation state data between other platforms;

[0040] data integration unit, integrating the received and sent data to meet the data format and requirements of both platforms, ensuring the interconnection of information, and providing comprehensive data support for the overall operation and management of the power grid.

[0041] Preferably, the function expansion module comprises:

[0042] function evaluation unit, according to user demand and distribution network development demand, the feasibility and necessity of the expandable function (such as new energy access adaptability evaluation, power quality prediction, etc.) are evaluated;

[0043] module adding unit, after the evaluation is passed, the selected function module is added to the existing system to realize the expansion of system function;

[0044] compatibility test unit, the added function module is tested for compatibility to ensure that it can work with other modules in the original system, improving the practicality and forward-looking nature of the system.

[0045] The beneficial effects of the present application are as follows:

[0046] 1、The active power distribution network power quality panoramic perception and abnormal positioning treatment system provided by the present application can comprehensively evaluate the applicability of various monitoring terminals, optimize the monitoring point configuration, and realize comprehensive perception, abnormal identification and accurate tracing of power quality problems by using advanced algorithms and models. In a complex active power distribution network, it can accurately monitor the power quality parameters of each node, timely discover and locate the root causes of problems such as harmonics and voltage deviation, provide reliable basis for subsequent treatment, and improve the reliability of power grid operation.

[0047] 2、The active power distribution network power quality panoramic perception and abnormal positioning treatment system provided by the present application can establish a multi-objective optimization configuration model for different power quality problems and propose a coordinated control strategy. In actual power grid, treatment equipment can be reasonably configured according to demand, network loss can be effectively reduced, power quality indicators such as high / low voltage, three-phase imbalance and harmonics can be improved, power quality stability can be ensured, user and power grid operation requirements can be met, and equipment service life can be extended.

[0048] 3, The active power distribution network power quality panoramic sensing and abnormal positioning management system, the construction of the power distribution network dispatching sensing module and the operation of the related units, the new situational awareness function, the construction of the panoramic model and the sensing graph; the operator can call data, analyze and enlarge view in real time or historically through the operation platform, which provides strong support for dispatching decision. In the operation of the power grid, it can quickly respond to power quality changes, reasonably adjust the operation mode, improve the power grid operation and maintenance efficiency, enhance the ability of the power grid to respond to emergencies, ensure the safe and stable operation of the power grid, and improve user satisfaction. BRIEF DESCRIPTION OF DRAWINGS

[0049] The application will be further described below with reference to the drawings.

[0050] Figure 1 It is a system block diagram of the application;

[0051] Figure 2 It is a schematic diagram of photovoltaic output timing characteristics of the application;

[0052] Figure 3 It is a schematic diagram of the method for selecting important points of the application;

[0053] Figure 4 It is a flow chart of the alternating iteration method of the harmonic part of the application;

[0054] Figure 5 It is a schematic diagram of the LSTM network structure of the application. DETAILED DESCRIPTION

[0055] In order to make the technical means, creative features, purposes and effects realized by the application easy to understand, the specific embodiments of each module will be further described below.

[0056] As Figures 1-5 shown, the application provides an active power distribution network power quality panoramic sensing and abnormal positioning management system, which includes a monitoring terminal adaptation module, a monitoring point optimization configuration module, a power quality sensing and early warning module, a composite power quality tracing module, a power quality comprehensive management module and a power distribution network dispatching sensing module.

[0057] The implementation of the monitoring terminal adaptation module is as follows:

[0058] Monitoring terminal analysis:

[0059] For the handheld power quality monitoring device in portable analyzer, its hardware design focuses on simple single-phase signal acquisition circuit, mainly used for detecting single-phase harmonic. The software algorithm focuses on fast analysis and display of single-phase harmonic data, such as using a simplified version of fast Fourier transform (FFT) algorithm to achieve preliminary detection of harmonic frequency and amplitude with lower computing resource requirements; in practical application, when the harmonic situation of a specific single-phase device or local area needs to be quickly investigated, the handheld device can be easily connected to the circuit for measurement.

[0060] The portable power quality monitoring device is equipped with three-phase signal acquisition interface and higher precision data processing chip in hardware, which can simultaneously collect three-phase voltage and current signals and calculate multiple power quality indicators such as three-phase unbalance and voltage deviation; its software functions include data storage and preliminary analysis, which can be used for short-term continuous monitoring on site, suitable for temporary power quality evaluation of distributed power access points or important load areas; for example, it can be used during the installation and debugging stage of distributed power or for preliminary detection of areas suspected of having power quality problems.

[0061] The discrete power quality monitoring device uses single-chip microcomputer or DSP technology, with 6-8 measurement channels in hardware, which can meet the accurate measurement of single-point three-phase voltage and current. Its software system realizes complex power quality index calculation algorithms, such as three-phase unbalance calculation method based on three-phase instantaneous reactive power theory, which can monitor and analyze power quality in real time. It is widely used in single-point power quality monitoring of substations or large industrial users.

[0062] The integrated power quality monitoring device is based on industrial computer technology, with strong hardware expansion capability, which can be expanded to more than 32 channels, and can realize synchronous measurement of multiple monitoring points; its software platform integrates database management function, which can store and analyze a large amount of power quality monitoring data for a long time, suitable for comprehensive monitoring of power quality of entire active distribution network or large area.

[0063] According to the type (such as photovoltaic, wind power, etc.), capacity, access location of distributed power and the topology of power grid, combined with the functional characteristics of different monitoring terminals, its applicability is judged, such as for the access point of distributed photovoltaic power station, if only the harmonic situation needs to be concerned, the handheld device may meet the preliminary detection requirement; if the overall impact of the power quality of the grid needs to be evaluated, including voltage deviation, three-phase unbalance, etc., discrete or integrated monitoring device needs to be selected; at the same time, considering the importance and scope of the monitoring area, for important load concentration area, high-precision and full-featured monitoring device is preferred; for remote and small load area, portable device can be selected for regular inspection according to cost-effectiveness.

[0064] Monitoring point optimization configuration module embodiment:

[0065] The improved BPSO algorithm is as follows:

[0066] The correction particle position and speed iteration formula: in the traditional BPSO algorithm, the particle position update formula The monitoring point configuration is unreasonable;

[0067] The improvement of the application is The speed update formula is modified accordingly, wherein Inertia weight, And Acceleration coefficient, And Random number, Individual optimal position, Global optimal position, the search efficiency of the particle is improved by introducing individual and global optimal guidance.

[0068] Construct a new evaluation function: the new evaluation function , E is the observation coverage rate of the monitoring point to the voltage sag domain, C is the configuration cost, D is the current information completeness, 、 、 Weight coefficient; reasonable setting of weight balances monitoring effect and cost, such as increasing the weight of E and D in the area with high power quality requirement, increasing the weight of C in the cost sensitive area to optimize the configuration.

[0069] Feasible domain feedback checking link: in the algorithm iteration, after updating the particle position each time, it is checked whether the number of monitoring points meets the feasible domain condition (the number of installed PQM is not less than NKCL), if not, the particle position is adjusted back to the feasible domain, so as to ensure that the final configuration scheme meets the monitoring requirements and is feasible.

[0070] Optimization configuration process: first, determine the node voltage sag domain and line distribution weight according to the active distribution network topological structure and line parameters; then initialize the particle swarm (particle position represents monitoring point selection, 1 for selection and 0 for non-selection); then calculate the particle fitness to determine the individual and global optimal position; in the iteration, update the particle position and speed according to the improved formula and check the feasible domain; when the maximum iteration number is reached or the convergence condition is met, output the optimal monitoring point configuration scheme; for example, in a 100-node active distribution network, after multiple iterations, it is determined that the power quality monitoring terminal is installed at 20 key nodes to realize effective monitoring.

[0071] Electric energy quality perception early warning module embodiment:

[0072] Distributed power generation model is established:

[0073] For the photovoltaic power generation model, according to the actual light intensity and temperature data, the Beta probability density distribution function is used:

[0074] , , and According to the local light conditions;

[0075] As Figure 2 shown, the daily periodic variation of light intensity is described, combined with the solar cell array output power formula , is the total area of the array, is the photoelectric conversion efficiency, is the actual light intensity; real-time calculation of power generation output power, and through historical data to determine the model parameters in different seasons and weather to improve accuracy.

[0076] For other distributed power sources (such as wind power), according to the wind speed, wind direction and other meteorological data to establish the power generation power model, such as using wind speed and power curve model, according to the local wind speed statistical data to determine the curve parameters, input real-time wind speed to calculate the output power, at the same time considering the characteristics of wind turbine cut-in, rated, cut-out wind speed to ensure the effectiveness of the model.

[0077] Power quality anomaly identification and early warning:

[0078] Data acquisition and preprocessing: through the sensor of the monitoring point to collect the voltage, current and other electrical quantity data with high sampling frequency (such as 1kHz), using digital filtering (such as low-pass filtering to remove high-frequency noise) and data normalization (normalizing the data to a certain interval) technology for pretreatment;

[0079] As Figure 3 shown, piecewise linearization and feature extraction: using improved important point selection method for piecewise linearization, setting the compression rate R, such as (50%), selecting important points according to the vertical distance of data to two-point connecting line; for time series

[0080] ,

[0081] Among them, is the acquisition time, is the data value, the vertical distance of each point to the connecting line of the first and last end is calculated, and the longest distance point is segmented, and the new sequence is extracted as the main trend feature;

[0082] Abnormal index calculation and judgment: calculate the trend abnormal index and , Among them, is the trend mode sequence of the test period, For normal period trend mode sequence, For similarity calculation function, dynamic time warping algorithm is used to calculate the principle of constructing time series dynamic mapping relationship to measure the similarity of shape and trend, and the minimum mapping path is found by iterative calculation of corresponding point distance. The total distance of this path is the dynamic time warping distance, which represents the similarity). The numerical anomaly index determines the normal data distribution interval according to the historical data using the interquartile range method (the interquartile range is determined by calculating the difference between the upper and lower quartiles of the data to exclude the influence of extreme values, and the middle 50% distribution range), and the proportion of out-of-interval outliers in the test period data is ; Then use CRITIC weighting method to determine the trend and numerical anomaly index weights and , ), calculate the comprehensive anomaly index ; Set threshold (such as 0.7), when determine the power quality anomaly of the monitoring point, send an early warning signal and record the upload of abnormal data (such as time, index value) to the dispatching center.

[0083] Power quality traceability module implementation:

[0084] Three-phase harmonic power flow analysis unit implementation:

[0085] Three-phase harmonic basic admittance matrix calculation: through the conversion matrix Integrate the three-phase parameters of the same component and the same harmonic, and perform phase sequence conversion operation to obtain the three-phase harmonic basic admittance matrix , which represents the three-phase branch admittance between the first and last nodes of the component, and is called the three-phase harmonic basic admittance matrix; The specific process is as follows:

[0086]

[0087] Matrix The generation process is as follows:

[0088]

[0089]

[0090] In the formula: represents the Kronecker product, which physically represents the expansion of the reference space from a single-phase bus to the corresponding abc three-phase bus; The subscript H represents that the matrix contains harmonic components; The subscript base represents the basic component parameter; The subscript 120 represents the three sequence; is a dimensional unit matrix, is a dimensional column vector with all elements being 1, , is the total number of elements, is the maximum number of harmonics; is a column vector with 1 as the starting point and 1 as the step size, is a column vector with 1 as the starting point and 1 as the step size, and so on; , is the phase sequence conversion matrix; the function is a column vector with 1 as the starting point and 1 as the step size, is a column vector with 1 as the starting point and 1 as the step size, and so on. Sparse matrix is generated with the elements of vector

[0091] Three-phase harmonic sub-admittance matrix derivation:

[0092] Including the self-admittance matrix and mutual admittance matrix of the first and last nodes of the element;

[0093] Line:

[0094]

[0095] In the formula: is the admittance matrix.

[0096] Transformer:

[0097]

[0098] In the formula: , are the primary and secondary tap ratio matrices of the transformer, respectively.

[0099] In the manner of constructing the index matrix, the three-phase harmonic sub-admittance matrix is integrated into the three-phase harmonic admittance matrix ; The specific process is as follows:

[0100]

[0101]

[0102]

[0103] In the formula: the subscript diag indicates that the corresponding matrix is a block diagonal matrix; is the floor function; the function is a column vector of triad information of a certain sparse matrix, that is, the row number, column number, and value of the non-zero element; is the index association matrix.

[0104] On the basis of the above matrix pattern, the three-phase harmonic admittance matrix of various models is generated:

[0105] Three-phase line harmonic admittance matrix:​

[0106]

[0107] Three-phase transformer harmonic admittance matrix:

[0108]

[0109] Three-phase load / generator harmonic admittance matrix:

[0110] Since the interaction between the load and the generator is not considered, the corresponding three-phase harmonic admittance matrix is a diagonal matrix, and the three-phase load harmonic admittance matrix and the three-phase generator harmonic admittance matrix can be directly obtained without using the canonical form.

[0111] As shown in Figure 4 , the harmonic source uses a harmonic current content rate model, and the algorithm of the harmonic part uses an alternating iteration method to alternately solve the fundamental and harmonic power flow, i.e., using the harmonic power flow result of the last iteration as the boundary condition to solve the fundamental power flow, using the fundamental power flow result of the last iteration as the boundary condition to solve the harmonic power flow, and realizing the alternating iteration of the fundamental and harmonic power flow based on the power conservation principle, so as to obtain each harmonic voltage:

[0112]

[0113] In the formula: is a harmonic three-phase voltage; is a harmonic three-phase current.

[0114] Power quality cause analysis unit implementation:

[0115] Distribution area type and attribute determination: based on the static information such as the metering record, the transformer capacity, the line length, the load type and other attributes of the distribution area are obtained. For example, the rated capacity and the transformation ratio are obtained from the transformer metering record, the length and the conductor type are obtained from the line design data, and the load type (such as industrial, residential, etc.) is determined according to the user information. According to these information, the distribution area type (such as urban residential area, industrial development area, etc.) is determined, and the power quality characteristics and potential problems of different types of distribution areas are different.

[0116] Power quality problem assessment and feature extraction: Combining distribution transformer monitoring data (three-phase voltage, current, power factor, etc.) and low-voltage user monitoring data (user-end voltage, current waveforms, etc.), data analysis algorithms such as cluster analysis and association rule mining are used to identify power quality problems at multiple time scales. For example, cluster analysis categorizes voltage fluctuation data to determine abnormal periods, and association rule mining finds the relationship between voltage deviation and load changes. Data features are extracted, including single-event statistical features (such as the duration and amplitude change of a voltage sag, where the duration is the interval from the start of the voltage drop to its return to normal, and the amplitude change is the difference in amplitude before and after the sag), daily statistical features (daily average voltage is the average of voltage samples within a day, and daily maximum voltage deviation is the maximum deviation within a day), weekly, monthly, and quarterly statistical features, and equipment status data (transformer oil temperature is measured by a temperature sensor, and load rate is the ratio of actual output power to rated power).

[0117] Cause Analysis and Prediction: Based on the extracted data characteristics and preset criteria (such as the relationship between voltage deviation and load change, and the relationship between harmonic content and distributed power output), the causes of power quality problems are analyzed. Spatially, the causes are categorized into 10kV feeder levels (e.g., long feeders leading to large voltage drops, according to Ohm's law). The voltage drop can be caused by factors such as: long line length leading to high resistance and large voltage drop when current flows through it; distribution transformer level (e.g., improper transformer tap settings causing output voltage to deviate from rated voltage); and low-voltage line level (e.g., aging lines increasing resistance and reduced conductor cross-sectional area leading to increased resistance and voltage drop); and time scale, it can be categorized into emergency types (e.g., short-circuit faults causing voltage dips, and sudden voltage drops due to a surge in current during a short circuit), intermittent types (e.g., voltage fluctuations caused by intermittent load startup, and instantaneous fluctuations caused by large inrush current during startup), and long-term types (e.g., long-term improper connection of distributed power sources causing harmonic problems, and harmonic injection into the grid due to their control strategies or inverter characteristics). Figure 5 As shown, an LSTM network is used to model the correlation between active power and power quality indicators. The active power data of power users and the historical monitoring data of steady-state power quality are input. The network state is updated through operations such as forget gate (determining whether to discard the unit state information of the previous moment), input gate (controlling the addition of new information) and output gate (determining the output information). The trend of steady-state power quality indicators is predicted, which provides a basis for early management. For example, if the voltage deviation in a certain area is predicted to increase, the reactive power compensation equipment can be adjusted in advance to maintain voltage stability.

[0118] Implementation method of the source tracing and positioning unit:

[0119] Data Acquisition and Screening: Obtain statistical values ​​of power quality monitoring data at the bus common coupling point from the power grid power quality monitoring and analysis system (the content of each harmonic is calculated by Fourier transform of the voltage or current signal at the monitoring point, and the average voltage deviation is the average deviation over a certain period of time). Obtain user power data (active and reactive power are measured by the user-end power measurement device) from the user power consumption information acquisition system. Screen the data characteristics of harmonics, voltage deviation, and negative sequence voltage imbalance with large fluctuations or exceeding the standard according to national standard limits and control charts. If the harmonics exceed the national standard limits or the voltage deviation exceeds the normal range at multiple sampling points, they are screened out. At the same time, analyze the correlation of disturbance indicators (such as the mutual influence between harmonic content and voltage deviation, and the voltage drop caused by harmonic current on the grid impedance affecting voltage deviation) to form a time-segmented set of power quality disturbance events.

[0120] Correlation coefficient calculation and source tracing: Substitute the selected data into the time series DDTW distance calculation. First, preprocess the data derivatives to obtain a new sequence, and record the disturbance period. Time series of monitoring data at the inner busbar , No. Average active power per user: New sequence after preprocessing , Calculation as follows ( similar), , Because the units are different, after calculating the mean and variance, we use standardization methods to convert them into unitless scores, denoted as... , , ;

[0121] Define a size of Distance matrix ,in:

[0122] ;

[0123] Representation of the characteristic matrix The Middle Each feature element and The Middle The cumulative distance is the square of the Euclidean distance between each feature element. The DDTW distance calculation uses an iterative method, and the cumulative distance matrix is ​​denoted as... The calculation method for each element in this matrix is ​​as follows:

[0124] ;

[0125] Until and , The value is the overall minimum cumulative distance, which is the DDTW distance between the two sequences, denoted as . This represents the similarity between the user and the power quality monitoring data in terms of trend and time characteristics. The smaller the distance value, the higher the similarity between the two during the disturbance period.

[0126] By comparing the correlation coefficients between different users and disturbance events, the source users of the disturbances can be identified, enabling the tracing of complex power quality problems. For example, if a user's correlation coefficient is significantly higher than that of other users, that user may be a source of power quality problems. In practical applications, a correlation coefficient threshold can be set. When a user's correlation coefficient with a disturbance event exceeds this threshold, that user is listed as a key target for investigation, and their electrical equipment and operating status are further analyzed to determine the specific source of the disturbance and appropriate mitigation measures.

[0127] Implementation method of the comprehensive power quality management module:

[0128] Harmonic mitigation unit implementation method:

[0129] Objective function establishment:

[0130] Objective function reflecting the harmonic content of each node (in for Node number RMS value of subharmonic current for The effective value of the fundamental current at each node is used to measure the magnitude of the harmonic current at each node relative to the fundamental current.

[0131] Objective function reflecting the total harmonic distortion rate at each node (in for The total harmonic distortion of each node is considered comprehensively.

[0132] Objective function for governance effectiveness: ( , These are weighting coefficients, set according to the focus and requirements of harmonic mitigation. For example, if more emphasis is placed on reducing harmonic content, then... The distortion rate is relatively high; if more emphasis is placed on improving the overall distortion rate, then... (relatively large), taking into account the single harmonic content and total distortion rate of each node.

[0133] Objective function reflecting governance costs: Assuming the capacity of each APF is fixed. , (For the desired number of APF units to be installed, reduce governance costs by minimizing the number of APF units installed).

[0134] The constraint condition is the total voltage distortion rate. To meet the requirements of the power grid, that is ( (The total voltage distortion limit specified by national standards), the content of each voltage harmonic meets the requirements of the power grid, when When it is an odd number, ( for Secondary voltage harmonic content. (for the corresponding limit); when When it is even, ( (Limit for even-order harmonics).

[0135] Control strategy for power generation equipment: Based on the objective function and constraints, optimization algorithms (such as genetic algorithms and particle swarm optimization algorithms) are used to calculate the optimal installation location and capacity of the APF (Active Power Distribution Filter). For example, the active power distribution network is divided into multiple areas, and several possible APF installation locations are set in each area. The optimization algorithm finds the installation scheme that minimizes the objective function under the constraints. In actual operation, the harmonic situation of the power grid is monitored in real time, and the output of the APF is dynamically adjusted according to the harmonic changes to effectively compensate for harmonic currents and reduce harmonic content and total distortion rate. For example, when a sudden increase in harmonic content is detected in a certain area, the output of the APF is increased to quickly suppress harmonics. At the same time, the parameters of the APF are reasonably adjusted in combination with the operating status of the power grid and load changes to improve its control effect and adaptability.

[0136] Implementation method of reactive power compensation unit:

[0137] Objective function establishment: To reflect the effect of reactive power compensation, an objective function for reactive power compensation is established: ,in, , For nodes and The electrical conductance between them and For nodes and The effective value of the voltage, For nodes and The voltage phase difference between them; used to calculate the active power loss of the distribution network and reflect the effect of reactive power compensation on reducing network losses;

[0138] , The reference voltage, usually the rated voltage, is used to calculate the voltage deviation at each node and measure the impact of reactive power compensation on voltage stability.

[0139] , The number of smart capacitor banks to be deployed is determined by minimizing the number of banks deployed to reduce costs, reflecting the governance costs.

[0140] Comprehensive voltage over-limit penalty function: ,when or hour, Penalty coefficient, , These are the upper and lower limits for node voltage; they constrain the voltage to prevent it from exceeding the limits.

[0141] After weighting , and (The setting depends on the focus of reactive power compensation; for example, it can be increased in areas with large network losses.) The weights) aggregate the three objective functions into one objective function: .

[0142] Reactive power compensation equipment adjustment: Based on the objective function, the impact of the connection location and capacity of smart capacitors on the reactive power compensation effect is analyzed. For example, installing smart capacitors near nodes with heavy loads and high reactive power demand can more effectively compensate for reactive power. By monitoring the grid voltage and reactive power, the switching status of smart capacitors is adjusted in real time. When the voltage is low and reactive power is insufficient, more smart capacitor banks are connected; when the voltage is high and reactive power is excessive, some smart capacitor banks are disconnected to maintain voltage stability and reactive power balance. Simultaneously, considering the lifespan and response speed of smart capacitors, frequent switching is avoided to prevent damage to the equipment and ensure the reliability and effectiveness of reactive power compensation. For example, a reasonable switching delay is set to avoid frequent operation during small fluctuations in voltage and reactive power.

[0143] Implementation method of three-phase imbalance control unit:

[0144] Objective function establishment:

[0145] To reflect the effectiveness of three-phase imbalance control, an objective function is established. ,in, For nodes Negative sequence current, For nodes Positive sequence current, From root node to node The per-unit impedance values ​​between phases reflect the impact of the topology and are used to measure the degree of three-phase imbalance.

[0146] Establish an objective function that reflects governance costs: Let the number of switches be . This extends equipment lifespan and reduces costs by reducing the number of commutation switch operations.

[0147] The constraint condition of three-phase imbalance treatment should meet the average voltage drop constraint of phase-changeable load nodes, that is, , is the average voltage drop of phase-changeable load nodes , is the maximum allowable average voltage drop.

[0148] Three-phase imbalance treatment measures: according to the objective function and constraint condition, determine the installation position and control strategy of phase-change switch; for example, install phase-change switch at load nodes with serious three-phase imbalance, judge the degree of three-phase imbalance by monitoring three-phase current and voltage; when the imbalance degree exceeds the set threshold, select the appropriate phase-change time according to the preset control strategy, control the phase-change switch to act, adjust the phase sequence of the load, and make the three-phase load tend to be balanced. At the same time, considering the change of load and the stability of power grid operation, avoid the problem of voltage fluctuation caused by frequent action of phase-change switch, and ensure the effectiveness and stability of three-phase imbalance treatment; for example, combined with load prediction data, perform phase-change operation in periods with small load changes to reduce the impact on the power grid.

[0149] Distribution network dispatching perception module implementation:

[0150] Platform building unit implementation:

[0151] Hardware architecture building: select high-performance servers as the hardware basis of the distribution network dispatching support platform, configure sufficient memory (such as 32GB or above), large-capacity hard disk (such as 1TB or above), and high-speed CPU (such as multi-core processor) to meet the data storage and processing requirements; install reliable operating system (such as Linux system) on the server, and equip with network communication equipment (such as gigabit Ethernet switch) to ensure stable communication with other systems (such as monitoring terminals, intelligent devices, etc.). For example, choose Dell PowerEdge R740 server, which has powerful computing power and scalability, and can meet the requirements of long-term stable operation of the system.

[0152] Software platform construction: The software platform is designed using a layered architecture, including a data acquisition layer, a data processing layer, a model construction layer, and an application service layer. The data acquisition layer is responsible for communicating with various monitoring devices and systems, collecting real-time operation data of the distribution network (such as voltage, current, power, etc.) and device status information (such as switch status, device temperature, etc.), using standardized data communication protocols (such as IEC 61850 protocol) to ensure accurate data transmission. The data processing layer cleanses the collected data (removes outliers and erroneous data), converts it (such as converting data in different formats to a unified format), and stores it (stores it in a database, which can use a relational database such as MySQL or a non-relational database such as MongoDB, and selects the appropriate storage method based on the characteristics of the data). The model construction layer constructs a panoramic model of the distribution network based on the network topology and device parameters, including electrical device models (such as transformers, lines, switches, etc.) and power quality models (used to calculate and analyze power quality indicators); the application service layer provides various application service interfaces, such as data query interfaces, situation awareness service interfaces, etc., providing data support and functional services for other modules and users. For example, the Spring Boot framework is used to develop the application service layer, providing RESTful-style interfaces for easy integration with other systems.

[0153] Situation awareness function unit implementation:

[0154] Data acquisition and integration: Through the data acquisition layer, real-time operation data of the distribution network is collected from multiple data sources such as monitoring terminals, smart meters, SCADA systems, including real-time electrical quantity data (three-phase voltage, current, power factor, etc.), device status data (switch position, device operating status, etc.), and power quality monitoring data (harmonic content, voltage deviation, three-phase imbalance, etc.). Time synchronization and data format unification processing are performed on the collected data to ensure data accuracy and consistency. For example, GPS clock synchronization technology is used to time-stamp data from each monitoring point, aligning data from different sources in time for subsequent analysis and processing; at the same time, a data format conversion module is developed to convert various formats of data output by different devices into a unified format used internally by the system, such as converting DL / T 645 protocol data from smart meters to JSON format.

[0155] Panoramic model construction and update: Based on the model construction layer's power grid topology and device parameter information, combined with real-time collected data, a panoramic model of the distribution network is constructed; the panoramic model includes the physical structure model of the power grid (line connection relationship, device location, etc.), the operating state model (real-time values and historical trends of each node voltage, current, and power), and the power quality model (real-time values and change trends of power quality indicators at each monitoring point); as real-time data is continuously collected, the panoramic model is dynamically updated, enabling the model to accurately reflect the current operating state and change trends of the distribution network; for example, when the state of a switch in the power grid changes, the physical structure model of the power grid is updated in a timely manner, and the power flow distribution is recalculated based on the new topology to update the operating state model and the power quality model. Real-time data pushing and timely model updating are achieved through subscription to a message queue and other means, ensuring that dispatchers can obtain the latest power grid information.

[0156] Situation awareness map drawing: The panoramic model of the distribution network is displayed in an intuitive situation awareness map using visualization technology (such as a Web-based graphical interface); in the situation awareness map, the line and device distribution of the power grid is drawn on the background of a geographic information system (GIS), and different colors and icons are used to represent the operating state (such as normal operation represented by green and fault represented by red) and power quality condition (such as voltage deviation within the normal range represented by blue and out of range represented by yellow or red) of the devices; at the same time, the real-time change trends of electrical quantities such as voltage, current, and power, as well as the historical change curve of power quality indicators, are displayed through dynamic curves, enabling dispatchers to intuitively understand the operating situation and power quality condition of the distribution network; for example, the Echarts visualization library is used to draw the situation awareness map, and through interactive functions such as map zooming and device clicking to query, dispatchers can easily view detailed information; clicking on a substation icon on the map can pop up detailed operating information and power quality indicators of the substation, including bus voltage at each voltage level, in-out line power, harmonic content, etc.

[0157] Operation platform unit implementation:

[0158] Real-time and historical data retrieval: A lightweight web operation platform is built to provide a user-friendly interface. Users (such as dispatchers, maintenance personnel, etc.) can easily retrieve real-time situational awareness graphs through the operation platform to view the current operating state of the distribution network; at the same time, they can query historical situational awareness graphs to view historical data according to time ranges (such as the past day, week, month, etc.) and analyze the operating conditions and power quality trends of the distribution network at different time periods; for example, when querying historical data, users can select a time range through a drop-down menu, and the operation platform retrieves data for the corresponding time period from the database and displays it in the form of charts or curves on the interface, such as drawing the voltage fluctuation curve of a certain line in the past week to help users analyze voltage stability. The operation platform also supports data export functions to facilitate further analysis and report generation.

[0159] Data analysis function implementation: The operation platform integrates data analysis tools to conduct in-depth analysis of real-time and historical data; for example, it provides statistical analysis functions to calculate statistical parameters such as the mean, standard deviation, maximum, and minimum of power quality indicators, helping users evaluate the overall level and dispersion of power quality; at the same time, it has correlation analysis functions to analyze the correlation between different electrical quantities (such as voltage and current, active power and reactive power, etc.) and between power quality indicators and operating parameters (such as harmonic content and load changes), providing a basis for fault diagnosis and optimal operation; for example, through correlation analysis, it is found that there is a strong positive correlation between voltage deviation and load peak period in a certain area, prompting dispatchers to take appropriate voltage adjustment measures during load peaks; trend analysis can also be conducted to predict future changes in power quality indicators using time series analysis methods, providing early warning of potential power quality problems to facilitate preventive measures. The operation platform provides visual analysis result displays such as column charts, line charts, and scatter plots, making the analysis results more intuitive and easy to understand.

[0160] Eagle magnification function application: To facilitate users in viewing local details and overall situation of the distribution network, the operation platform is equipped with an eagle magnification function; when viewing panoramic situational awareness graphs, users can select a local area of interest by clicking on the eagle area, and the operation platform automatically displays the area in a magnified manner, while displaying the global view and the location relationship of the current magnified area in the eagle area, facilitating quick positioning and switching between different areas; for example, when dealing with local faults, dispatchers can use the eagle magnification function to quickly focus on the area where the fault point is located, view detailed device information and operating data, and at the same time understand the location of the fault point in the entire distribution network through the eagle to comprehensively judge the impact of the fault on the overall situation and develop a reasonable handling plan. When users click on different positions in the eagle area, the operation platform can smoothly switch the magnified area, providing a smooth operation experience.

[0161] Data processing unit implementation:

[0162] Power quality cause identification data processing: For the power quality cause identification data required by the power distribution area governance decision technology, the operation platform calculates the statistical characteristic values according to the set time interval (such as 1 hour); when calculating the average value, the voltage, current and other data collected in this time period are summed and divided by the number of sampling points to obtain the average voltage, average current and other average value characteristic values; the maximum and minimum values are directly obtained by traversing all the sampling data in this time period; at the same time, the distribution transformer tap position, capacitor state and other information (if the equipment supports collection) are obtained, which can be obtained by communicating with the distribution transformer monitoring terminal and the capacitor intelligent control device; for example, the current tap position information of the distribution transformer is read from the distribution transformer monitoring terminal through the Modbus communication protocol, and the switching state of the capacitor is obtained from the capacitor intelligent control device. In the data processing process, the collected data is checked for validity, such as checking whether the voltage and current data are within a reasonable range, and if abnormal data is found, it is marked or corrected to ensure the accuracy of the data.

[0163] Different time scale statistical characteristic value calculation and analysis: the daily statistical characteristic value is calculated based on the 5-minute interval measured data in real time; the average voltage, maximum voltage deviation and other characteristic values in the period are calculated every 5 minutes, and the statistical characteristic values of the whole period are accumulated; at the same time, it is judged whether there is voltage overrun and other situations in each 5-minute period, and the overrun period index (such as overrun times, overrun duration, etc.) is counted; when a day ends, the calculation of daily statistical characteristic values is completed, and the data is stored in the database, and at the same time, the delay cause analysis program is started to analyze the possible causes of the power quality problem of the day. The weekly statistical characteristic value is based on the daily statistical data, and the daily statistical characteristic value and the daily overrun period index statistical characteristic value are accumulated from 0 o'clock of each day, and the calculation method is similar to the calculation of daily statistical characteristic value, for example, the weekly statistical characteristic value such as average daily voltage deviation and maximum daily voltage fluctuation is calculated; the monthly statistical characteristic value is based on the weekly statistical data, and the weekly statistical characteristic value and the weekly overrun period index statistical characteristic value are accumulated from 0 o'clock of each week, and the statistical index prefix is “M”, such as “M monthly average voltage” and the like, and the overrun period statistical index prefix is “MM”, such as “MM monthly voltage overrun total duration”. The quarterly statistical characteristic value is based on the monthly statistical data, and the monthly statistical characteristic value and the monthly overrun period index statistical characteristic value are accumulated from 0 o'clock of each month, and the statistical index prefix is “F”, such as “F quarterly maximum load current”, and the monthly overrun period statistical index prefix is “FM”, such as “FM quarterly voltage overrun times”; through the calculation and analysis of statistical characteristic values of different time scales, the power quality status of the distribution network can be comprehensively understood from multiple angles such as long-term trend and short-term fluctuation, which provides strong data support for power quality management and power grid optimization operation. In the statistical process, data caching technology is used to improve the calculation efficiency and reduce the frequent read-write operation of the database, and the data is backed up to prevent data loss.

[0164] Self-learning adaptive module implementation:

[0165] Data collection unit implementation:

[0166] Data source access and integration: Establish stable connections with multiple data sources such as monitoring terminals, power quality monitoring systems, SCADA systems, etc., to ensure real-time access to various operation data of the active distribution network; data includes electrical quantity data (voltage, current, power, etc.), equipment status data (switch status, transformer oil temperature, etc.), power quality monitoring data (harmonic content, voltage deviation, three-phase imbalance, etc.), and distributed power operation data (output power, power factor, etc.); Use data adapter technology to convert data from different sources and different formats into a unified data format for subsequent processing; for example, for some specific format power quality data output by the monitoring terminal, write the corresponding data adapter to convert it into a general format (such as JSON format) for easy data storage and analysis; At the same time, real-time monitoring of data, when data transmission interruption or abnormality is found, timely alarm and attempt to reconnect the data source, to ensure the continuity and integrity of the data.

[0167] Data storage and management: Store the collected data in the database, use distributed database architecture (such as HBase or Cassandra) to cope with large-scale data storage and high-concurrency access requirements; store data by type (such as operation data, monitoring data, etc.), time range (such as daily, weekly, monthly, etc.) and equipment area (such as different substations, lines, etc.) for easy data query and retrieval. At the same time, establish data index to improve data query efficiency; for example, establish time and monitoring point location index for voltage data, when need to query voltage data of specific monitoring point in a certain time period, can quickly locate and obtain data; Regularly clean up and optimize the database, delete expired or useless data, improve the performance of the database.

[0168] Model optimization unit implementation:

[0169] Power quality evaluation model optimization: Based on the collected historical operation data and real-time monitoring data, machine learning algorithms (such as support vector machines, neural networks, etc.) are used to optimize the power quality evaluation model; first, the data is divided into training set and test set, the training set is used to train the model, and the test set is used to evaluate the model performance; in the training process, according to the actual power quality situation (such as whether there is harmonic over-standard, voltage deviation too large, etc.), the parameters of the model are adjusted; for example, for the support vector machine model, by adjusting the kernel function parameters and penalty factor, etc., the classification accuracy of the model for different power quality problems is improved. The optimized model is verified periodically using the test set, and the accuracy, recall rate and other evaluation indicators of the model are calculated, when the indicators reach the expected target or no longer have obvious improvement, stop the optimization process; at the same time, model fusion technology is used to fuse multiple different power quality evaluation models, and the advantages of multiple models are integrated to improve the accuracy and reliability of the evaluation.

[0170] Abnormal positioning algorithm improvement: According to the dynamic change characteristics of active distribution network, the abnormal positioning algorithm is improved. Combined with the change of power grid topology structure, the fluctuation of distributed power output and other factors, the dynamic weighting algorithm is used to adjust the weight in the abnormal positioning algorithm; for example, when the distributed power output changes greatly, increase the weight of the monitoring data related to the distributed power in the abnormal positioning algorithm, improve the positioning accuracy of the algorithm for the power quality anomalies caused by the distributed power; at the same time, deep learning algorithm (such as convolutional neural network) is introduced to improve the abnormal positioning algorithm, which can better identify complex power quality anomaly patterns by using its automatic feature extraction capability; through the learning of a large number of historical abnormal data and normal operation data, the algorithm can more accurately judge the type and location of the anomaly; in the process of algorithm improvement, sufficient experiments and verifications are carried out, and the performance of the algorithm before and after improvement is compared to ensure that the improved algorithm has better effect.

[0171] Strategy adjustment unit implementation:

[0172] Governance decision strategy optimization: According to the data analysis results and model optimization, adjust the power quality governance decision strategy; when it is found that the harmonic problem in a certain area is mainly caused by a specific type of distributed power supply, adjust the governance strategy, and prefer to use the harmonic governance technology for this type of distributed power supply (such as improving the inverter control strategy of distributed power supply), rather than simply relying on traditional harmonic governance equipment (such as APF); in terms of reactive power compensation, according to the load change trend and voltage stability of the power grid, dynamically adjust the switching strategy of the reactive power compensation equipment; for example, when the load is at the peak and the voltage is low, more reactive power compensation equipment is put into operation in advance to improve the voltage stability; when the load is at the trough and the voltage is high, part of the reactive power compensation equipment is appropriately cut off to avoid overcompensation; at the same time, combined with economic cost analysis, the optimal governance scheme is selected to reduce the governance cost on the premise of ensuring power quality.

[0173] System parameter adaptive adjustment: According to the running state change of the active distribution network, the parameters of the system are adaptively adjusted; for example, when the grid frequency fluctuates, the data sampling frequency of the monitoring terminal is automatically adjusted to ensure that the dynamic change information of the power grid can be accurately collected. When it is found that the power quality problem in some areas has a greater impact on system operation, the data acquisition accuracy and monitoring frequency of the monitoring points in that area are increased to strengthen the monitoring of the power quality in that area; at the same time, according to the operation of the governance equipment (such as the harmonic compensation effect of APF, the reactive power compensation ability of intelligent capacitor, etc.), the control parameters of the governance equipment are automatically adjusted to keep it in the best operating state, improve the effect and efficiency of power quality governance; for example, according to the real-time harmonic compensation effect of APF, the size and phase of the compensation current are dynamically adjusted to achieve the best harmonic suppression effect.

[0174] Data interaction sharing module implementation:

[0175] Interface adaptation unit implementation:

[0176] Communication protocol analysis and conversion: For different communication protocols (such as IEC 61850, IEC 60870-5, etc.) adopted by different power system management platforms (such as distribution automation systems, power marketing systems, etc.), develop corresponding protocol analysis modules; when interacting with other platforms, the interface adaptation unit first receives the data sent by the other platform, parses it according to its communication protocol, and extracts the effective information (such as electrical quantity data, device status information, etc.) in the data; then, convert the parsed information into a format that the system can recognize and process (such as a unified data structure body or a database record format); for example, when receiving switch state change information transmitted using the IEC 61850 protocol from the distribution automation system, the protocol analysis module parses the switch number, state change time, and new state information according to the IEC 61850 protocol specification, and converts it into a format used by the system to store switch state information, facilitating subsequent processing. At the same time, log the parsing process for troubleshooting and data analysis.

[0177] Interface development and adaptation: According to the interface specifications provided by other platforms and the needs of the system, develop corresponding interface programs; ensure that the system can seamlessly integrate with other platforms to achieve bidirectional data transmission; for example, for the Web service-based interface provided by the power marketing system, develop a matching client program to achieve data interaction with the power marketing system by calling the Web service interface; at the same time, optimize and adapt the interfaces provided by the system to the outside world to meet industry standards and security specifications, making it easy for other platforms to access; for example, when developing the data query interface of the system, use the RESTful architectural style to provide simple and standardized interface definitions to facilitate other platforms to obtain the required data through HTTP requests. During interface development, conduct thorough testing, including functional testing, performance testing, and security testing, to ensure the stability and reliability of the interface.

[0178] Data transmission unit implementation:

[0179] Data packaging and sending: Package the data that needs to be sent to other platforms, add necessary identification information (such as data source, timestamp, data type, etc.), ensure the integrity and traceability of the data; according to the communication requirements and network environment of the target platform, select the appropriate transmission method (such as TCP / IP protocol, UDP protocol, etc.) for data transmission; for example, for data with high real-time requirements (such as fault alarm information), use UDP protocol for fast transmission to reduce transmission delay; for data with high accuracy requirements (such as power settlement data), use TCP / IP protocol for reliable transmission to ensure data is not lost or tampered with. During data transmission, establish a data transmission log to record the time, target platform, and data volume sent, etc. to facilitate monitoring and troubleshooting of the data transmission process; at the same time, encrypt the sent data to ensure data security.

[0180] Data receiving and analysis: When receiving data sent by other platforms, first perform data integrity and legality verification; verification content includes whether the data format is correct, whether the data length meets the agreement, whether the checksum is correct, etc.; if the verification is passed, the data is parsed to extract the valid information, and it is stored in the database or cache of the system for subsequent processing; for example, when receiving user power data sent by the power marketing system, verify whether the data format meets the agreed power data format, check whether the data length is consistent with the expected, calculate the checksum and compare it with the checksum provided by the sender; if the verification is correct, parse the user number, power consumption, power time, etc. and store them in the database to provide data support for power quality analysis and management decision-making; during data reception, set up a data buffer to prevent system crashes caused by too fast data reception.

[0181] Data integration unit implementation:

[0182] Data format unification and conversion: Since the data formats of different platforms may differ, the data integration unit is responsible for unifying and converting the received data; according to the data model and application requirements of the system, convert the data from other platforms to the format used internally by the system; for example, convert the device status data in the power distribution automation system represented in a specific format to a format consistent with the device status model in the system, facilitating data correlation and analysis. During data conversion, match and convert data types, such as converting string type numerical data to numerical type to ensure data accuracy and consistency; at the same time, evaluate the quality of the converted data, check for missing, abnormal, etc. and repair or mark problem data.

[0183] Data correlation and fusion: correlate and fuse data from different platforms to establish logical relationships between data; for example, correlate power quality monitoring data with user electricity consumption data in the power marketing system, establish contact through user number or geographic location, etc., analyze the relationship between power quality problems and user electricity consumption behavior; fuse data from different sources to generate comprehensive information and provide a more comprehensive perspective for power grid operation and management; for example, combine the power grid topology structure data in the distribution automation system with the power quality monitoring data in the system to analyze the propagation path and influence range of power quality problems in the power grid, and provide more accurate basis for fault location and governance; in the process of data correlation and fusion, data mining and machine learning algorithms are used to discover potential relationships and rules between data and improve the value of data; at the same time, the fused data is visualized to facilitate users to intuitively understand and analyze the data.

[0184] Function extension module implementation:

[0185] Function evaluation unit implementation:

[0186] User demand analysis: communicate with power grid operation and management departments, power users, etc. to collect their demands and expectations for system function extension; for example, organize user forums, issue questionnaires, etc. to understand users' specific demands in new energy access adaptability evaluation, power quality prediction, etc., including evaluation accuracy requirements, prediction time range, required input and output parameters, etc.; at the same time, analyze power grid development planning and industry trends to predict future new demands, such as the demand for coordinated control and optimized operation of distributed power clusters with the large-scale development of distributed energy; sort and classify the collected demands to establish a demand library to provide a basis for function extension.

[0187] Feasibility study: based on user demands and existing system architecture, conduct a feasibility study on the proposed extended functions. Evaluate technical feasibility, analyze whether existing technical conditions (such as hardware resources, software algorithms, communication capabilities, etc.) can support the implementation of new functions; for example, for power quality prediction function, evaluate whether the current system's computing power and data resources are sufficient to run complex prediction models; at the same time, consider economic feasibility, estimate the cost (including labor cost, equipment procurement cost, software development cost, etc.) required for developing and implementing new functions and the economic benefits (such as improving power quality, reducing power grid loss, improving user satisfaction, etc.) that may be brought; based on technical and economic factors, judge the feasibility of function extension and propose corresponding suggestions and solutions; during the feasibility study, refer to relevant industry standards and technical specifications to ensure the rationality and compliance of the extended functions.

[0188] Module addition unit implementation:

[0189] Function module development: Based on the function evaluation results, determine the function modules that need to be added and develop them; for the new energy access adaptability evaluation module, develop the corresponding algorithms and models for evaluating the impact of new energy access on power grid power quality, stability and reliability; for example, establish new energy generation models (such as photovoltaic cell models, wind turbine models, etc.), combine power flow calculation and stability analysis methods to evaluate voltage fluctuation, harmonic injection, frequency stability and other indicators after new energy access. For the power quality prediction module, use advanced machine learning or deep learning algorithms (such as long short-term memory networks, convolutional neural networks, etc.) to establish prediction models based on historical power quality data and related influencing factors (such as load changes, weather conditions, etc.) to achieve short-term and long-term prediction of power quality indicators (such as voltage deviation, harmonic content, etc.); during development, follow the software development life cycle model to perform demand analysis, design, coding, testing and other stages to ensure the quality and stability of the function modules; at the same time, conduct integration testing with existing systems to ensure compatibility and collaborative work capability of new function modules with original systems.

[0190] System integration and testing: integrate the developed function modules into the existing system for system integration testing; test content includes interface compatibility between function modules and existing systems, accuracy and timeliness of data interaction, and collaborative work between new functions and original functions, etc.; for example, when integrating the new energy access adaptability evaluation module, test whether the data interaction between it and the power quality monitoring module and the power grid analysis module is normal, and whether the evaluation results can be accurately fed back to the decision support module of the system; through comprehensive system integration testing, find and solve possible problems to ensure that the system with added function modules can run stably and meet user needs; during system integration testing, use automated testing tools to improve testing efficiency and accuracy.

[0191] Compatibility test unit implementation:

[0192] Hardware compatibility testing: check the compatibility of the added function modules with the system hardware; test the running of the function modules under different hardware configurations (such as different models of servers, monitoring terminals, etc.), including CPU usage, memory occupancy, hard disk read / write speed and other performance indicators; ensure that the function modules can run normally under various hardware environments and will not cause system failure or performance degradation due to hardware differences; for example, test the running speed and stability of the power quality prediction module on servers with different performance to observe whether there is prediction delay or error due to insufficient hardware resources; at the same time, test the compatibility of the driver programs of the hardware devices to ensure that the function modules can correctly call hardware resources.

[0193] Software compatibility testing: verify the compatibility of the functional module with the existing software system (including operating system, database management system, other application programs, etc.); test the installation, startup and running of the functional module under different operating systems (such as Windows, Linux, etc.), ensure that it does not conflict with the kernel, driver and other operating systems; at the same time, check whether the data interaction with the database management system (such as MySQL, Oracle, etc.) is normal, and whether the data storage and query function is correct; in addition, test the cooperative work ability of the functional module with other related application programs (such as power distribution automation software, power marketing system client, etc.), ensure the overall stability and functionality of the system in a multi-application environment; for example, when running with power distribution automation software, test whether the new energy access adaptability evaluation module can correctly obtain real-time operation data of power grid and feed back the evaluation results to the power distribution automation software to realize cooperative optimization control; through comprehensive compatibility testing, it is ensured that the added functional module can be well compatible with the hardware and software environment of the system, and the overall reliability and usability of the system are improved; during the compatibility testing process, a test report system is established to record the problems found in the testing process and the solutions, providing a reference for subsequent system maintenance and upgrading.

[0194] The basic principles, main features and advantages of the present application are shown and described above. Those skilled in the art should understand that the present application is not limited by the above examples, and the above examples and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. An active power distribution grid power quality panoramic sensing and abnormality locating and governing system, characterized in that, The monitoring terminal adaptation module, the monitoring point optimization configuration module, the power quality perception and early warning module, the composite power quality traceability module, the power quality comprehensive management module and the power distribution network dispatching perception module are included, and the power distribution network dispatching perception module is used for building a platform to realize situation awareness and support power distribution network dispatching decision-making. The monitoring terminal adaptation module is used for analyzing the applicability of power quality monitoring terminals to power quality monitoring of distributed power sources, so as to select appropriate monitoring terminals. The monitoring point optimization configuration module realizes the optimization layout of monitoring points based on an improved algorithm, guarantees the observability of voltage sag domains and the completeness of current information, and helps power quality event diagnosis and positioning. The power quality perception and early warning module is used for accurate perception and timely early warning of power quality problems through model establishment and feature analysis. The composite power quality traceability module is used for analyzing power quality problem causes, simulation and traceability, and providing basis for management. The power quality comprehensive management module realizes multi-objective optimization management of power quality by analyzing equipment influence, establishing a model and proposing strategies. The power distribution network dispatching perception module is used for building a platform to realize situation awareness and support power distribution network dispatching decision-making. In the power quality perception and early warning module, a model considering different output scenarios of distributed power sources and load time sequence characteristics is established, trend characteristics are extracted by piecewise linearization, that is, an improved important point selection method is adopted, abnormal indexes are determined and weighted combined according to time sequence trend similarity and numerical outlier proportion, that is, CRITIC weighting method is adopted, and abnormal identification and early warning of power quality problems are realized in combination with historical data. The composite power quality traceability module includes a three-phase harmonic power flow analysis unit, a power quality cause profiling unit and a traceability positioning unit. The three-phase harmonic power flow analysis unit obtains a three-phase harmonic admittance matrix through matrix operation, calculates harmonic voltage by using a harmonic current content rate model and an alternating iteration method, and analyzes power quality problem simulation. The power quality cause profiling unit determines substation attributes based on static information, evaluates problem causes, identifies multi-time scale problems in combination with multi-source data, analyzes causes after feature extraction, and predicts steady-state power quality index trends by using an LSTM network modeling. The traceability positioning unit obtains relevant data, filters features to form a disturbance event set, calculates correlation coefficients by using time sequence DDWT distance, and realizes composite power quality problem traceability positioning by matching the event set.

2. The active distribution network power quality panoramic sensing and abnormality locating and governing system according to claim 1, characterized in that, In the monitoring terminal adaptation module, power quality monitoring terminals include portable analyzers and online monitors, and the monitoring terminal adaptation module determines the applicability of monitoring terminals according to distributed power source characteristics and power quality problems.

3. The active distribution network power quality panoramic sensing and abnormality locating and governing system of claim 1, wherein, In the monitoring point optimization configuration module, an improved binary particle swarm optimization algorithm, that is, BPSO, is used to correct particle iteration formula, construct a new evaluation function and increase a feasible region feedback verification link, and the feasible region condition is that the number of installed PQMs is not less than NKCL.

4. The active distribution network power quality panoramic sensing and abnormality locating and governing system of claim 1, wherein, The power quality comprehensive management module includes: A harmonic management unit, which analyzes the regulation and control capability of APF management equipment on harmonics, establishes an objective function reflecting harmonic content, total distortion rate and management cost of each node, and satisfies voltage harmonic national standard requirements as a constraint condition, so as to reduce the influence of harmonics on the power grid and improve power quality. The reactive power compensation unit analyzes the compensation effect of the intelligent capacitor device on the reactive power, constructs a target function related to the reactive power compensation effect, system transmission power loss, voltage deviation and treatment cost, combines power balance, unbalance constraint, power factor limit and capacitor capacity limit constraint condition, realizes reasonable compensation of the reactive power, and stabilizes the grid voltage; The three-phase imbalance treatment unit is arranged according to the treatment effect of the commutating switch device on the three-phase imbalance, sets up a target function containing the treatment effect, i.e. the negative sequence three-phase imbalance degree and the treatment cost, i.e. the number of commutating switch actions, and improves the three-phase imbalance condition under the constraint of the average voltage drop of the commutable load node, and ensures the three-phase load balance.

5. The active distribution network power quality panoramic sensing and abnormality locating and governing system of claim 1, wherein, The power distribution network dispatching sensing module comprises: A platform building unit is configured to build a power distribution network dispatching support platform to provide a basic framework for subsequent function implementation; A situation awareness function unit is configured to add a situation awareness function to the original power distribution network dispatching system, build a panoramic model of the power distribution network by collecting power distribution network data and performing calculation, build a power distribution network situation awareness map, and realize panoramic awareness of the running state of the power distribution network; An operation platform unit is configured to build a lightweight power distribution network situation awareness operation platform, which is configured to call real-time and historical situation awareness maps, and has data analysis and eagle-eye magnification functions to provide strong support for power distribution network dispatching; A data processing unit is configured to process power quality cause identification data required by power distribution area treatment decision technology, including calculating statistical characteristic values in an event time period, including average value, maximum value, minimum value, and capacitor state information, and performing cumulative statistics on daily, weekly, monthly, and seasonal statistical characteristic values according to a preset rule, and performing time-delay cause analysis after the corresponding statistics are completed to provide comprehensive data support and decision basis for operation and maintenance personnel.

6. The active distribution network power quality panoramic sensing and abnormality locating and governing system of claim 1, wherein, Further comprising a self-learning adaptive module, the self-learning adaptive module comprising: A data collection unit is configured to collect historical operation data and real-time monitoring data to provide a data basis for subsequent analysis; A model optimization unit is configured to optimize the power quality evaluation model according to the collected data to improve evaluation accuracy; An algorithm improvement unit is configured to improve the abnormal positioning algorithm according to data characteristics and operation conditions to improve positioning accuracy; A strategy adjustment unit is configured to adjust the treatment decision strategy based on data analysis results to adapt to the dynamic changes of the active power distribution network and ensure that the system always maintains good performance.

7. The active distribution network power quality panoramic sensing and abnormality locating and governing system of claim 1, wherein, Further comprising a data interaction sharing module, the data interaction sharing module comprising: An interface adaptation unit is configured to perform interface adaptation with other power system management platforms to ensure the compatibility of data interaction; A data transmission unit is configured to realize data transmission between other platforms, including bidirectional transmission of power quality data and device operation state data; A data integration unit is configured to integrate and process the received and sent data to make them conform to the data format and requirements of both platforms, guarantee information interconnection, and provide comprehensive data support for overall operation and management of the power grid.

8. The active distribution network power quality panoramic sensing and abnormality locating and governing system of claim 1, wherein, Further comprising a function expansion module, the function expansion module comprising: A function evaluation unit is configured to evaluate the feasibility and necessity of expandable functions according to user needs and power distribution network development needs; A module adding unit adds the selected functional module to the existing system to realize the expansion of the system function after the evaluation passes; A compatibility testing unit tests the compatibility of the added functional module to ensure that it can work with other modules of the original system and improve the practicality and forward-looking nature of the system.

Images