Microgrid energy management system with abnormal data detection function

Abstract

The invention relates to a micro-grid energy management system with an abnormal data detection function, and the system also stores an energy management scheduling optimization program which is used for matching a day-ahead scheduling plan with intra-day scheduling, and guarantees the reliable operation of a micro-grid. The energy management system executes the energy management scheduling optimization program to realize energy management. The method comprises the following steps: acquiring real-time measurement data, future prediction data and past prediction data of a plurality of distributed energy units, and judging the distributed energy units with abnormal data through a normal data range standard built in an EMS database; judging whether real-time intra-day EMS scheduling, future EMS scheduling and past EMS scheduling are abnormal or not through a comparison algorithm of real-time, future and past residual errors and residual threshold values; and according to an abnormal data detection result, regulating and scheduling to ensure reliable operation of the micro-grid. The method has the beneficial effects that an abnormal data detection function is provided to verify the optimization scheduling effectiveness.

Description

[Technical field]

[0001] The present invention relates to the technical field of power supply systems, and in particular to a microgrid energy management system with an abnormal data detection function. [Background technology]

[0002] A microgrid is a power generation and distribution system that integrates distributed power sources, loads, energy storage devices, converters, and monitoring and protection devices. As an organizational form for effectively integrating and efficiently utilizing renewable energy power generation, it achieves grid-connected or islanded operation through technologies such as operation control and energy management.

[0003] As an emerging microgrid system, the integrated photovoltaic storage and charging system has great research and development potential. However, most of the current research on photovoltaic storage and charging microgrid EMS (energy management system) systems tends to focus on the overall scheduling strategy, while ignoring whether the day-ahead scheduling strategy can operate reliably. Distributed energy output is affected by factors such as weather and temperature and is volatile. Although the EMS system can predict it through historical data, its accuracy cannot be guaranteed. The day-ahead scheduling strategy formulated by the EMS system operating within the day is likely to deviate from the established target, so detecting anomalies in past data, current real-time data, and future predicted data, and recovering from the anomalies based on the detection results, is of great significance to the operation of the EMS (energy management system) and minimizing the risk of potential business losses.

[0004] The present invention aims at the technical problem that the prior art lacks an effective detection data function to verify the effectiveness of its optimized scheduling, and makes technical improvements to the microgrid energy management system. [Summary of the invention]

[0005] The purpose of the present invention is to propose a microgrid energy management system with an abnormal data detection function to verify the effectiveness of optimized scheduling.

[0006] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is a microgrid energy management system with an abnormal data detection function, wherein the microgrid includes a plurality of distributed energy units, the energy management system includes a management engine and an EMS database, the distributed energy units, the management engine and the EMS database are connected via an internet communication protocol, a database connection protocol or a local network protocol, the EMS database includes historical data of scheduling results and normal data range standards under normal operating conditions of the distributed energy units; the energy management system also stores an energy management scheduling optimization program for matching the day-ahead scheduling plan with the intraday scheduling and ensuring the reliable operation of the microgrid; the energy management system executes the energy management scheduling optimization program to realize energy management, including the following steps:

[0007] S1. Collect real-time measurement data of several distributed energy units in each period X i , and upload it to the management engine, which determines the distributed energy unit with abnormal real-time measurement data through the normal data range standard built into the EMS database, and starts the backup unit;

[0008] S2. The management engine detects the current scheduling results of the distributed energy units through a real-time residual and residual threshold size comparison algorithm to determine whether the real-time intraday EMS scheduling is abnormal;

[0009] S3. The management engine uses the historical data in the EMS database to obtain future forecast data of several distributed energy units. The management engine determines the distributed energy units with abnormal future prediction data through the normal data range standard built into the EMS database;

[0010] S4, the management engine detects the future scheduling results of the distributed energy units by predicting the future residual and the residual threshold size comparison algorithm, and determines whether the future EMS scheduling is abnormal;

[0011] S5. The management engine uses the historical data in the EMS database to calculate the past forecast data of several distributed energy units. T is a past time frame, and the management engine determines the distributed energy units with abnormal past forecast data through the normal data range standard built into the EMS database;

[0012] S6. The management engine detects the past scheduling results of the distributed energy units through a comparison algorithm of the past residual and the residual threshold, and determines whether the past EMS scheduling is abnormal and needs to be optimized;

[0013] S7. According to the abnormal data detection results of steps S1, S2, and / or steps S3, S4, and / or steps S5, S6, regulation and scheduling are performed to ensure reliable operation of the microgrid.

[0014] Preferably, the energy management system executes the energy management scheduling optimization program to implement energy management step S2, which specifically includes the following sub-steps:

[0015] S21. The day-ahead dispatch strategy is derived through the management engine's optimized dispatch algorithm, and the predicted output of each distributed energy unit is calculated. The residual error of the microgrid during period t is calculated Among them, X i (t) is the actual output measurement value of distributed energy unit i in the microgrid during period t, is the predicted output value of distributed energy unit i in microgrid during period t;

[0016] S22. Calculate the size of the remaining threshold Among them, E i is the residual threshold, α i is the magnification, arg u To solve the collective function of u, Prob is the probability function, β is the confidence level, and α and β are determined according to the requirements of the day-ahead scheduling strategy;

[0017] S23, by calculating the residual and the residual threshold, determine whether the real-time intraday EMS, dispatch is abnormal, and convert the real-time measurement data of several distributed energy units into i It is divided into green normal data, yellow warning data and red abnormal data; the green normal data meets W i ≤E 1 , the yellow warning data meets E 1 <W i <E 2 , the red abnormal data satisfies W i ≥E 2 ,in,

[0018]

[0019] α 1 is the normal range magnification, α 2 The magnification factor for abnormal data.

[0020] Preferably, the energy management system executes the energy management scheduling optimization program to implement energy management step S23: α 1 =1.1,α 2 =1.4.

[0021] Preferably, the energy management system executes the energy management scheduling optimization program to implement energy management step S4, which specifically includes the following sub-steps:

[0022] S41, the past residual W of distributed energy unit i through historical data i，k (t k ) to predict the future residual W of distributed energy units i，f (t f ), after filtering out the future residuals of the abnormal future forecast data in step S3, the available future residuals Among them, t k is a past time point, W i，k (t k ) is the past residual of distributed energy unit i in the historical data of EMS database, t f For future time points, is the available future residual;

[0023] S42. Calculate the size of the future remaining threshold Among them, α 3 is the future threshold magnification, β f Confidence for future thresholds;

[0024] S43, judging whether the future EMS scheduling is abnormal by the size of the available future residual and the future remaining threshold, if Then determine the future t f The EMS dispatch data during this period is abnormal.

[0025] Preferably, the energy management system executes the energy management scheduling optimization program to implement energy management step S6, which specifically includes the following sub-steps:

[0026] S61. Obtain the past residual W of distributed energy unit i through historical data i，k (t k ), after filtering out the future residuals of the abnormal past prediction data in step S5, the available past residuals Among them, t k is a past time point, W i，k (t k ) is the past residual of distributed energy unit i in the historical data of EMS database, is the available past residual;

[0027] S62: Calculate the size of the remaining threshold in the past Among them, α 4 is the past threshold magnification, β k is the confidence level of the past threshold;

[0028] S63, judging whether the past EMS scheduling is abnormal by the available past residual and the past residual threshold. Then determine the past t k The EMS dispatch data during this period is abnormal.

[0029] Preferably, the microgrid energy management system with abnormal data detection function further includes an operation platform, which is used for operators to perform secondary judgment and correction on the abnormal data of steps S1, S2, and / or steps S3, S4, and / or steps S5, S6.

[0030] Preferably, the microgrid energy management system with abnormal data detection function further includes a recovery controller, which is used to control the output of several distributed energy units.

[0031] Preferably, the distributed energy unit is a photovoltaic, energy storage, and / or charging station energy unit.

[0032] The microgrid energy management system with abnormal data detection function of the present invention has the following beneficial effects: it can intelligently detect data anomalies of each unit in the microgrid, and the detection function includes detection of current scheduling results, detection of future and past scheduling results, and mutual transmission and feedback of modification results through the platform system of the present invention. Finally, the operator performs secondary confirmation of the data results, which can perform real-time detection of the energy efficiency and reliability of the energy management system, and ensure the reliability of the EMS scheduling results.

Brief Description of the Drawings

[0033] Figure 1 The invention relates to an abnormal data detection flow chart of a microgrid energy management system with an abnormal data detection function. [Specific implementation method]

[0034] The present invention will be further described below in conjunction with embodiments and with reference to the accompanying drawings.

[0035] Example

[0036] This embodiment implements a microgrid energy management system with an abnormal data detection function.

[0037] In view of the lack of effective detection data function in the current EMS (energy management) system to verify the effectiveness of its optimized scheduling, this embodiment proposes a microgrid system energy management system with abnormal detection function including photovoltaic, energy storage, and charging station. Specifically, an autonomous distributed energy management platform is introduced. The system of this embodiment has the function of detecting abnormal data, including a method and control process for detecting abnormal data and an energy management system operation process including the function of detecting abnormal data. The energy management system of this embodiment includes an operating platform, an EMS (energy management system) management engine, a recovery controller, and an EMS (energy management system) database. These modules are connected through an internet communication protocol, a database connection protocol, or a local network protocol. The operating platform of this embodiment can make the EMS operation more flexible and dynamic, and the operator makes a secondary judgment and correction on the EMS system data through the operating platform. The EMS management engine of this embodiment separates distributed energy resources (DER) for economic optimization, and sends management commands to each unit in the microgrid. The recovery controller receives the correction command of the upper layer and feeds back the result to the EMS management engine for optimization. The EMS database of this embodiment includes the historical data of distributed energy in the microgrid and the data range standard of each unit under normal operating conditions.

[0038] Figure 1 This is a flow chart of abnormal data detection in a microgrid energy management system with abnormal data detection function. Figure 1 As shown, the abnormal data detection of the system in this embodiment includes the following steps:

[0039] 1. The day-ahead dispatch result is obtained through the built-in EMS algorithm of the EMS management engine, and the day-ahead dispatch result is compared and screened with the actual measured value and historical data; the method of the abnormal detection function includes the detection of real-time EMS dispatch data within the day and the detection of future EMS dispatch data; the corresponding algorithm of the detection function is judged by comparing the calculation residual with the residual threshold, and is divided into three data types for the operator to make a secondary judgment. The three data types include green normal data, yellow warning data, and red abnormal data.

[0040] 2. Upload the data classification results to the operation platform for the operator to perform secondary operations to determine the data type. The operator can modify parameters such as magnification and confidence to change the system's data screening rules.

[0041] 3. The dispatching results can be regulated by the recovery controller. The recovery controller receives commands from the control platform, issues correction instructions to the corresponding abnormal units and updates the EMS database. The operation process of the energy management system with detection function is to establish error analysis in the EMS database, compare it with the prediction data generated in the management engine, and feed back the results to the operation platform for the operator to confirm twice before issuing instructions to the recovery controller, which controls the output of each unit in the microgrid.

[0042] The system of this embodiment can effectively guarantee the normal operation of EMS (energy management system), match the EMS day-ahead dispatch plan with the intra-day dispatch, and ensure the reliable operation of the photovoltaic storage and charging microgrid.

[0043] Specifically, the system of this embodiment collects real-time measurement data Xi of each unit such as load, photovoltaic and energy storage in each period, and uploads the data to the EMS management engine, which makes a preliminary judgment based on the normal data range standard built into the EMS database. If the uploaded data exceeds its normal range, the backup device of the abnormal data unit will be automatically started, and the abnormal data will be uploaded to the operation platform for secondary operation by the operator.

[0044] The detection of the real-time EMS dispatch result by the system of this embodiment can be achieved by the following steps:

[0045] Real-time intraday detection of whether EMS data is abnormal can be determined by calculating the residual and the residual threshold in real time.

[0046] The system in this embodiment uses the optimization scheduling algorithm of the EMS management engine to obtain the day-ahead scheduling strategy and calculate the predicted output of each unit. It can be concluded that the calculated residual of the system in period t is as follows:

[0047]

[0048] Where: X i (t) is the actual measured value of unit i in the microgrid during period t, is the predicted value of microgrid unit i during period t;

[0049] In this embodiment, the size of the remaining threshold is calculated by the following formula:

[0050]

[0051] Where: E i is the residual threshold, α i is the magnification, arg u To solve the aggregate function of u, Prob is the probability function, β is the confidence level, and α and β can be formulated according to the needs of the strategy maker.

[0052] The system of this embodiment divides the data into three situations by calculating the residual and the residual threshold, namely: green normal data, yellow warning data and red abnormal data. The specific data discrimination rules are shown in the following formula:

[0053] Green normal data: W i ≤E 1

[0054] Yellow warning data: E 1 <W i <E 2

[0055] Red abnormal data: W i ≥E 2

[0056] in:

[0057]

[0058]

[0059] Where: α 1 For the normal range magnification, the initial value is 1.1, α 2 is the abnormal data magnification factor, and the initial value here is 1.4.

[0060] The actual data is classified into three situations and uploaded to the operating platform for operators to make a secondary determination of which data are outliers.

[0061] The system of this embodiment determines whether the future scheduling of the EMS is correct by predicting the size of the future residual and the residual threshold, which is achieved by the following steps:

[0062] 1. Calculate the future residual based on the historical residual and related data (such as temperature data, power data, current data, etc. recorded in the EMS). This process analyzes whether the future residual is within the threshold range. If it exceeds the threshold, the EMS system will extract the future abnormal time and notify the operation platform of the future abnormal value and time. The detailed process is as follows:

[0063] Request the EMS management engine to use the EMS historical database to predict the status and data of each unit in the future And check future forecast data through the data range library Is it normal data? If the data exceeds the range, it will be uploaded to the operation platform to remind the operator of the abnormal data time period and status in the future.

[0064] 2. This embodiment detects abnormalities in future EMS scheduling by analyzing the past residual data W of the i unit in the microgrid system. i，k (t k ) to predict the future residual W of each unit i，f (t f ), after filtering out the future residuals with abnormal data, the future residual data can be used It can be expressed as follows:

[0065]

[0066] In the above formula: t k is a past time point, W i，k (t k ) is the past residual of the i-th unit in the microgrid in the database, which can be obtained through the training data set, t f For future time points, is the available future residual.

[0067] 3. Then the correctness of the EMS scheduling strategy in the future period can be judged by the available future residual data and the size of the future residual threshold. The size of the future residual threshold can be expressed as follows:

[0068]

[0069] Where: α 3 is the future threshold magnification, β f is the confidence level for the future threshold.

[0070] like Then the system determines the future t f The time period scheduling data is abnormal and the records are uploaded f The data of the time period is sent to the operation platform, and the operator determines whether to replace or restore the system before the expected failure time.

[0071] The system of this embodiment uses the historical residuals of the measured energy data to predict future residuals, and compares all available predicted residuals with future thresholds. If the predicted residual exceeds the threshold, the operator can see possible faults that may occur in the future.

[0072] The system of this embodiment checks the EMS system data and compares it with the range table to determine whether the data is normal.

[0073] The system of this embodiment detects past data through the following steps:

[0074] According to the system optimization request, the energy data configuration file of the device i unit is imported into the EMS management engine or the energy data is downloaded from the management engine.

[0075] Taking 1h interval scheduling as an example, for each historical time frame T = {t1, t2, ... t24}, the EMS management engine is used to calculate the past forecast data and the past residual W i，k (t k ). Preliminary screening of past forecast data by comparing with the normal data range of EMS If it exceeds the normal value range, it will be uploaded to the management platform.

[0076] Filter the past residual W i，k (t k ), list possible data anomalies, and filter to obtain available past residuals It can be expressed as follows:

[0077]

[0078] The available past residuals Compare with the past threshold to determine whether the past scheduling data is abnormal. The past threshold can be expressed as follows:

[0079]

[0080] Where: α 4 is the past threshold magnification, β k is the past threshold confidence.

[0081] like The past optimization data is judged to be abnormal. The data are uploaded to the operation platform uniformly, and the operator confirms the data for the second time and modifies the erroneous historical data set.

[0082] Those skilled in the art will appreciate that all or part of the steps to implement the above embodiments may be accomplished by hardware, or by a program to instruct related hardware, and the program may be stored in a computer-readable storage medium, wherein the storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0083] The above is only a preferred embodiment of the present invention. It should be pointed out that ordinary technicians in this technical field can make several improvements and supplements without departing from the principle of the present invention. These improvements and supplements should also be regarded as the scope of protection of the present invention.

Claims

1. A microgrid energy management system with an abnormal data detection function, the microgrid comprising a plurality of distributed energy units, the energy management system comprising a management engine and an EMS database, the distributed energy units, the management engine and the EMS database being connected via an internet communication protocol, a database connection protocol or a local network protocol, Features: The EMS database includes historical data of scheduling results and normal data range standards under normal operating conditions of distributed energy units; the energy management system also stores an energy management scheduling optimization program for matching the day-ahead scheduling plan with the intraday scheduling to ensure the reliable operation of the microgrid; the energy management system executes the energy management scheduling optimization program to implement energy management, including the following steps: S1. Collect real-time measurement data of several distributed energy units in each period X i , and upload it to the management engine, which determines the distributed energy unit with abnormal real-time measurement data through the normal data range standard built into the EMS database, and starts the backup unit; S2. The management engine detects the current scheduling results of the distributed energy units through a real-time residual and residual threshold size comparison algorithm to determine whether the real-time intraday EMS scheduling is abnormal; S3. The management engine uses the historical data in the EMS database to obtain future forecast data of several distributed energy units. The management engine determines the distributed energy units with abnormal future prediction data through the normal data range standard built into the EMS database; S4, the management engine detects the future scheduling results of the distributed energy units by predicting the future residual and the residual threshold size comparison algorithm, and determines whether the future EMS scheduling is abnormal; S5. The management engine uses the historical data in the EMS database to calculate the past forecast data of several distributed energy units. T is a past time period, and the management engine determines the distributed energy units with abnormal past forecast data through the normal data range standard built into the EMS database; S6. The management engine detects the past scheduling results of the distributed energy units through a comparison algorithm of the past residual and the residual threshold, and determines whether the past EMS scheduling is abnormal and needs to be optimized; S7. According to the abnormal data detection results of steps S1, S2, and / or steps S3, S4, and / or steps S5, S6, regulation and scheduling are performed to ensure reliable operation of the microgrid.

2. A microgrid energy management system with abnormal data detection function according to claim 1, Features The energy management system executes the energy management scheduling optimization program to implement energy management step S2, which specifically includes the following sub-steps: S21. The day-ahead dispatch strategy is derived through the management engine's optimized dispatch algorithm, and the predicted output of each distributed energy unit is calculated. The residual error of the microgrid during period t is calculated Among them, X i (t) is the actual output measurement value of distributed energy unit i in the microgrid during period t, is the predicted output value of distributed energy unit i in microgrid during period t; S22. Calculate the size of the remaining threshold Among them, E i is the residual threshold, α i is the magnification, arg u To solve the collective function of u, Prob is the probability function, β is the confidence level, and α and β are determined according to the requirements of the day-ahead scheduling strategy; S23, by calculating the residual and the residual threshold value, determine whether the real-time intraday EMS scheduling is abnormal, and convert the real-time measurement data of several distributed energy units into i It is divided into green normal data, yellow warning data and red abnormal data; the green normal data meets W i ≤E 1 , the yellow warning data meets E 1 <W i <E 2 , the red abnormal data satisfies W i ≥E 2 ,in, α 1 is the normal range magnification, α 2 The magnification factor for abnormal data.

3. A microgrid energy management system with abnormal data detection function according to claim 2, Features The energy management system executes the energy management scheduling optimization program to implement energy management step S23: α 1 =1.1,α 2 =1.

4.

4. A microgrid energy management system with abnormal data detection function according to claim 2, Features The energy management system executes the energy management scheduling optimization program to implement energy management step S4, which specifically includes the following sub-steps: S41, the past residual W of distributed energy unit i through historical data i,k (t k ) to predict the future residual W of distributed energy units i,f (t f ), after filtering out the future residuals of the abnormal future forecast data in step S3, the available future residuals Among them, t k is a past time point, W i,k (t k ) is the past residual of distributed energy unit i in the historical data of EMS database, t f For future time points, is the available future residual; S42. Calculate the size of the remaining future threshold where α 3 is the future threshold magnification factor, and β f is the future threshold confidence level; S43, judging whether the future EMS scheduling is abnormal by the size of the available future residual and the future remaining threshold, if Then determine the future t f The EMS dispatch data during this period is abnormal.

5. A microgrid energy management system with abnormal data detection function according to claim 4, Features The energy management system executes the energy management scheduling optimization program to implement energy management step S6, which specifically includes the following sub-steps: S61. Obtain the past residual W of distributed energy unit i through historical data i,k (t k ), after filtering out the future residuals of the abnormal past prediction data in step S5, the available past residuals Among them, t k is a past time point, W i,k (t k ) is the past residual of distributed energy unit i in the historical data of EMS database, is the available past residual; S62: Calculate the size of the remaining threshold in the past Among them, α 4 is the past threshold magnification, β k is the confidence level of the past threshold; S63, judging whether the past EMS scheduling is abnormal by the available past residual and the past residual threshold. Then determine the past t k The EMS dispatch data during this period is abnormal.

6. A microgrid energy management system with abnormal data detection function according to claim 5, Features: It also includes an operating platform, which is used for operators to perform secondary determination and correction on abnormal data of steps S1, S2, and / or steps S3, S4, and / or steps S5, S6.

7. A microgrid energy management system with abnormal data detection function according to claim 6, Features: It also includes a recovery controller, which is used to control the output of several distributed energy units.

8. A microgrid energy management system with abnormal data detection function according to any one of claims 1 to 7, Features: The distributed energy units are photovoltaic, energy storage, and / or charging station energy units.

Images