User-side multi-power-supply, multi-energy-storage and electrical-load real-time power flow prediction method based on immune heredity

Abstract

The invention relates to a user side multi-power supply, multi-energy storage and electrical load real-time power flow prediction method based on immune heredity. The method comprises the following steps: S1, obtaining real-time operation data of a plurality of power supplies, a plurality of energy storage devices and electrical loads on a user side; s2, constructing a real-time power flow prediction model which is based on an immune genetic algorithm and meets node power equation constraints; s3, based on the real-time operation data, identifying a dynamic relationship among a plurality of power supplies, a plurality of stored energy and an electrical load; and S4, based on the real-time operation data, carrying out power flow prediction by using the real-time power flow prediction model. Compared with the prior art, the method has the characteristics of realizing accurate prediction of the power flow, realizing monitoring of the operation state of the equipment and the like.

Description

Technical Field

[0001] The present invention relates to the technical fields of power flow prediction and power system security, and in particular to a real-time power flow prediction method for multiple power sources, multiple energy storages, and electrical loads on the user side based on immune genetics. Background Art

[0002] With the rapid development of smart grid and renewable energy technologies, modern power systems are facing increasingly complex power demand and supply challenges. Especially on the user side, the power system needs to cope with the coordinated scheduling of multiple power sources (such as renewable energies like wind energy and solar energy) and multiple energy storage devices. In addition, the electrical loads of users also exhibit high dynamics and uncertainties. Therefore, accurately predicting the real-time power flow of multiple power sources, multiple energy storages, and electrical loads on the user side is of great significance for the stable and economic operation of the power system.

[0003] The penetration rate of new energy generation on the user side is gradually increasing, and the power flow changes in the distribution network on the user side are complex. How to achieve economic and technical optimal operation that satisfies all parties has become a difficult problem. Therefore, predicting the power flow of the distribution network on the user side is the key to realizing the optimal operation of the distribution network on the user side. At present, the distribution network on the user side has multiple power sources and energy-consuming devices, and there are real-time upward and downward changes in electric energy, resulting in real-time power flow changes in the distribution network.

[0004] Traditional load forecasting methods usually rely on statistical analysis of historical data, but when facing large-scale distributed energy access and load fluctuations, the forecasting accuracy often fails to meet the actual requirements.

[0005] In summary, there is currently a lack of a real-time power flow prediction method for electrical loads on the user side. Summary of the Invention

[0006] The purpose of the present invention is to overcome the deficiencies of the above-mentioned existing technologies and provide a real-time power flow prediction method for multiple power sources, multiple energy storages, and electrical loads on the user side based on immune genetics, so as to solve or partially solve the problem that the prediction accuracy often fails to meet the actual requirements when facing large-scale distributed energy access and load fluctuations.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] One aspect of the present invention provides a real-time power flow prediction method for multiple power sources, multiple energy storages, and electrical loads on the user side based on immune genetics, including the following steps:

[0009] Step S1, obtaining the real-time operation data of multiple power sources, multiple energy storage devices, and electrical loads on the user side;

[0010] Step S2, construct a real-time power flow prediction model that combines the immune genetic algorithm and satisfies the node power equation constraints;

[0011] Step S3, based on the real-time operation data, identify the dynamic relationships among multiple power sources, multiple energy storages, and electrical loads;

[0012] Step S4, based on the real-time operation data, use the real-time power flow prediction model to perform power flow prediction.

[0013] As a preferred technical solution, step S4 includes:

[0014] Step S401, use the real-time operation model to analyze the power flow on the user side, traverse the power data, screen the valid data and compare it with the output of the prediction model, identify the change trend of the power flow, and screen the matching power dispatching and optimization strategies;

[0015] Step S402, compare the real-time power flow data with the output of the real-time operation model, and provide accurate power dispatching suggestions and optimization strategies.

[0016] As a preferred technical solution, step S3 includes:

[0017] Step S301, based on the real-time operation data, screen the valid data through correlation analysis, and identify the dynamic relationships among multiple power sources, multiple energy storage devices, and electrical loads by comparing the valid data with the historical power flow prediction data, and construct a dynamic relationship model;

[0018] Step S302, predict the change trend of the real-time power flow on the user side based on the dynamic relationship model.

[0019] As a preferred technical solution, step S2 includes:

[0020] Step S201, store the collected real-time operation data in an incremental database;

[0021] Step S202, based on the collected real-time operation data, generate the first power flow characteristics through crossover and mutation and store them in the database;

[0022] Step S203, match the second power flow characteristics that match the collected real-time operation data from the database and store them in the database;

[0023] Step S204, generate a real-time power flow prediction model that satisfies the node power equation constraints based on the initial characteristic data, the first power flow characteristics, and the second power flow characteristics in the database.

[0024] As a preferred technical solution, in step S1, after obtaining the real-time operation data, it further includes:

[0025] Step S101, synchronize the real-time operation data through time series alignment and interpolation;

[0026] Step S102, align the real-time operation data to the same time scale based on a preset power data acquisition period;

[0027] Step S103, for missing data points, use linear interpolation, spline interpolation or dynamic interpolation based on historical trends to fill in the missing data points;

[0028] Step S104, perform filtering and denoising processing on the real-time operation data.

[0029] As a preferred technical solution, in step S1, the real-time operation data of multiple power sources includes at least one of real-time voltage, current, power, frequency, power factor, and load of the device; the real-time operation data of multiple energy storage devices includes at least one of charge and discharge power, voltage of the battery, current, state of charge, temperature, historical power load, and operation state of the device; the real-time operation data of the electrical load includes at least one of real-time power load data, historical load data, meteorological data, and time series data.

[0030] As a preferred technical solution, it further includes:

[0031] Step S5, in response to the predicted abnormal user-side power data, or the gap between the real-time operation data and the predicted data exceeding a preset threshold, use the device corresponding to the abnormal power data as a risk device and send a corresponding alarm notification.

[0032] As a preferred technical solution, step S5 further includes:

[0033] In response to the predicted user-side power data being within a preset normal range, generate a power dispatching log.

[0034] Another aspect of the present invention provides a real-time power flow prediction system for multiple power sources, multiple energy storages and electrical loads on the user side based on immune genetics, including:

[0035] A distributed data acquisition module for obtaining the real-time operation data of multiple power sources, multiple energy storage devices and electrical loads on the user side;

[0036] A prediction model construction module for constructing a real-time power flow prediction model combined with an immune genetic algorithm;

[0037] A data analysis module for identifying the dynamic relationship between multiple power sources, multiple energy storages and electrical loads based on the real-time operation data;

[0038] A real-time prediction module for performing power flow prediction based on the real-time operation data by using the real-time power flow prediction model.

[0039] As a preferred technical solution, it further includes:

[0040] A risk warning module for, in response to the prediction of abnormal power data on the user side or the gap between the real-time operation data and the predicted data exceeding a preset threshold, taking the device corresponding to the abnormal power data as a risk device and sending a corresponding warning notice.

[0041] Compared with the prior art, the present invention has at least one of the following beneficial effects:

[0042] (1) Achieving accurate prediction of power flow: The present invention first collects real-time operation data from user-side devices, including voltage, current, power, and status information, etc. After obtaining the real-time data, a real-time power flow prediction model is constructed through an immune genetic algorithm. In the process of establishing the prediction model, the collected real-time data, historical operation data, and possible operation scenarios are combined to generate the real-time power flow prediction model. The dynamic relationship between multiple power sources, multiple energy storage devices, and electrical loads is identified and analyzed, and the real-time power flow prediction model is retrieved to predict the change trend of the power flow, thereby realizing the real-time monitoring and prediction of the power system.

[0043] (2) Achieving the monitoring of the device operation status: Based on the immune genetic algorithm, the present invention predicts whether the operation status of the device is normal according to the generated real-time power flow prediction model. It compares whether the voltage, current, power, and status information match the expected operation parameters. When one or two parameters are detected to have deviations, a warning message is sent to the relevant personnel and the device is adjusted. If the parameters return to normal after adjustment, it is predicted that the device is operating normally; if the adjustment fails, it is predicted that the device is operating abnormally. When three or more parameters are detected to have deviations, it is directly predicted that the device is operating abnormally, a warning message is sent to the relevant personnel and emergency measures are taken. By setting the threshold of the parameter deviation to predict the abnormal operation of the device, the influence of accidental factors on the system can be reduced, the system risk can be lowered, and the safety and stability of the power system can be improved. Description of the Drawings

[0044] Figure 1 It is a flowchart of the real-time power flow prediction method for multi-power sources, multi-energy storage, and electrical load on the user side based on immune genetics in the embodiment;

[0045] Figure 2 It is a schematic diagram of the power flow change of the user-side distribution network with multiple power sources in the embodiment;

[0046] Figure 3 It is a schematic diagram of the real-time power flow prediction system for multi-power sources, multi-energy storage, and electrical load on the user side based on immune genetics in the embodiment. Detailed implementation manners

[0047] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0048] Embodiment 1

[0049] In view of the problem in the prior art that traditional load forecasting methods usually rely on statistical analysis of historical data, but when facing large-scale distributed energy access and load fluctuations, the forecasting accuracy often fails to meet the actual needs, this embodiment provides a real-time power flow forecasting method for multiple power sources, multiple energy storages and electrical loads on the user side based on immune genetics. This method is used for the user-side power system as shown in Figure 2 , and real-time data of multiple power sources, multiple energy storage devices and electrical loads are collected to perform real-time forecasting and monitoring of power flow. The immune genetic algorithm is adopted, which emphasizes improving the global search ability of the algorithm and the stability of the forecasting model through the immune mechanism to adapt to the real-time requirements of the complex system of multiple power sources and multiple energy storages. And this method emphasizes real-time performance. Through real-time forecasting and monitoring, abnormal situations are alerted in a timely manner to reduce system risks.

[0050] See Figure 1 , this method includes the following steps:

[0051] Step S1, information collection: Collect real-time operation data from multiple power sources, multiple energy storage devices and electrical loads on the user side, including voltage, current, power and status information, etc.

[0052] Among them, the operation status information of the user-side multiple power source devices includes but is not limited to real-time voltage, current, power (active power, reactive power, apparent power), frequency, power factor, and the load condition of the devices, etc. These data help to comprehensively understand the working status and operation efficiency of the power source devices. Real-time voltage and current can reflect the load condition of the power equipment, and power data is used to evaluate the energy efficiency and stability of the equipment, while frequency and power factor help to judge whether the power system is in a normal operation state.

[0053] Among them, the charge and discharge status information of multiple energy storage devices on the user side includes, but is not limited to, charge and discharge power, battery voltage, current, SOC (state of charge), temperature, historical power load, and the operating status of the devices, etc. The purpose of these data is to accurately model the charge and discharge behavior of energy storage devices, evaluate the performance of the devices, optimize energy scheduling, and predict future power demand and supply situations. By monitoring these data in real time, it can help the immune genetic algorithm optimize the operation strategy of energy storage devices and improve their operating efficiency and economic benefits.

[0054] Among them, the power consumption status information of the user-side power consumption load includes, but is not limited to, real-time power load data, historical load data, meteorological data, time series data, and other factors that may affect power consumption. Real-time power load data is the core of prediction and can reflect the current change in power demand; historical load data can help analyze the fluctuation law of the load and provide the basis for training the immune genetic algorithm; meteorological data (such as temperature, humidity, wind speed, etc.) has a significant impact on power consumption demand, especially in extreme weather; time series data (such as hours, days, weeks, etc.) can capture the periodic characteristics of power consumption demand; preferably, consider the impact of factors such as holidays and seasonal changes on the power consumption load and collect them as additional features. Through these data, the immune genetic algorithm can identify potential patterns, optimize the prediction model, and provide accurate load prediction results.

[0055] Among them, during the data processing process, when the data collection periods are different (for example, the data collection periods of meteorological station data and power data are different), a fusion method based on time series alignment and interpolation algorithm is adopted to ensure the synchronization and consistency of data. Set a unified time reference according to the power data collection period and align all data to the same time scale. For the meteorological station data with a lower sampling frequency, fill in the missing data points through interpolation methods (such as linear interpolation, spline interpolation, or Fourier interpolation) to ensure that there is corresponding prediction input at each power data time point. Preferably, it also includes data filtering and denoising processing to eliminate the error accumulation caused by different collection periods. This multi-dimensional data fusion method can provide a high-precision input basis for real-time power flow prediction, thereby improving the reliability of prediction results.

[0056] Preferably, the information collection supports multiple communication protocols, including but not limited to Modbus, DNP3, and IEC61850, to realize the real-time collection of multi-source device data.

[0057] Preferably, an information collection is realized by adopting a distributed data collection architecture, and the edge computing device is used to realize the efficient collection and processing of the data of multiple power sources, multiple energy storages, and power consumption loads on the user side.

[0058] Step S2, construct a prediction model: For Figure 2For the power of each branch, an immune genetic algorithm is used to construct a real-time power flow prediction model that satisfies the constraints of the nodal power equation, and the prediction accuracy and stability of the model are improved through an optimization algorithm. Specifically, step S2 includes steps S201 - S204.

[0059] In step S201, the characteristic data of multiple power sources, multiple energy storages, and electricity loads on the user side collected in step S1 are added to the memory storage database, the characteristic data is backed up, the data in the memory storage database is enriched, and its coverage range is improved.

[0060] In step S202, genetic operations are performed on the characteristic data of multiple power sources, multiple energy storages, and electricity loads on the user side, and the first possible power flow characteristics are generated through crossover and mutation, including but not limited to voltage (including amplitude and phase angle), active power, reactive power, etc., and new data files are generated and stored in the memory storage database.

[0061] In step S203, the power flow characteristics similar to those generated in step S202 are searched from the memory storage database as the second possible power flow characteristics, and new data files are generated and stored in the memory storage database.

[0062] In step S204, the initial characteristic data, the first possible power flow characteristics, and the second possible power flow characteristics in the memory storage database are combined to generate a real-time power flow prediction model that satisfies the constraints of the nodal power equation. The power changes of each branch are calculated by combining the nodal power equation, so as to predict the magnitude and direction of the power flow of each branch. Step S204 includes steps S20401 - S20409.

[0063] In step S20401, initialization is performed. Let the size of the antibody population be N, and each antibody xi represents the power of a certain branch of a certain node in the power grid. The initial population can be expressed as:

[0064] X = {x1, x2, …, xN}.

[0065] And the power of each branch conforms to the constraints of the nodal power equation, that is, the sum of the input powers of a certain node is equal to the sum of the output powers, and the power vector sum is 0. The nodal power equation is expressed as:

[0066]

[0067] In step S20402, fitness calculation is performed. A fitness function f(xi) is defined to evaluate the performance of each antibody xi, which is usually related to the operating cost, efficiency, and stability of the power grid. The fitness function can be multi-objective. Preferably:

[0068] f(xi) = α·C(xi) + β·E(xi) + γ·S(xi)

[0069] Among them, C(xi) is the operating cost, E(xi) is the efficiency, S(xi) is the stability, and α, β, γ are the weight coefficients.

[0070] Step S20403, immune selection: According to the fitness function f(xi), select antibodies with better performance for cloning and mutation operations. The selection probability Pi can be expressed as:

[0071]

[0072] Step S20404, cloning: Clone the selected antibody xi. The number of clones ni can be determined according to the fitness f(xi), such as:

[0073]

[0074] Step S20405, mutation: Perform mutation operations on the cloned antibodies to introduce new solutions. The mutation operation can be random or directional to improve the search efficiency. The mutated antibody can be expressed as xi′ = xi + Δxi, where Δxi is the mutation amount.

[0075] Step S20406, clone suppression: Perform clone suppression operations on the cloned antibodies, that is, eliminate antibodies with poor performance and retain antibodies with better performance. The population after clone suppression can be expressed as X′, and the power of each branch meets the node power equation constraint conditions.

[0076] Step S20407, population update: Combine the antibodies after cloning and mutation operations with the original population to form a new population X″, and the power of each branch meets the node power equation constraint conditions.

[0077] Step S20408, iteration: Repeat the selection, cloning, mutation, and clone suppression operations until the iteration times are met or the preset performance criteria are reached.

[0078] Step S20409, output result: After the iteration ends, select the antibody with the highest fitness as the optimal solution, that is, the best power grid voltage and power distribution scheme. Let the optimal solution be x*, then:

[0079]

[0080] Step S3, data analysis: Analyze the collected data to identify the dynamic relationship between multiple power sources, multiple energy storages, and electrical loads. Specifically, Step S3 includes Steps S301 - S302.

[0081] Step S301 analyzes the collected data of multiple power sources, multiple energy storages, and electricity loads, traverses the data set, filters out irrelevant data, reduces data redundancy, compares the valid data with historical prediction data, and identifies the dynamic relationships among multiple power sources, multiple energy storages, and electricity loads;

[0082] Step S302 calls the filtered dynamic relationship model, compares the real-time collected data with the historical data in the dynamic relationship model, and predicts the real-time power flow change trend among multiple power sources, multiple energy storages, and electricity loads.

[0083] Step S4, real-time prediction: Apply the constructed prediction model to perform real-time prediction on the power flow on the user side, and provide accurate power dispatching and optimization strategies. Specifically, Step S4 includes Steps S401 - S402.

[0084] Step S401 applies the constructed prediction model to perform real-time analysis on the power flow on the user side, traverses the power data, filters out irrelevant data, reduces data redundancy, compares the valid data with the output of the prediction model, identifies the power flow change trend, and filters out the power dispatching and optimization strategies applicable to this data;

[0085] Step S402 retrieves the filtered power dispatching and optimization strategies, compares them with the real-time power flow data and the output of the prediction model, and provides accurate power dispatching suggestions and optimization strategies. If the predicted power flow change trend is normal, update the power dispatching suggestions to the memory storage database, generate a new power dispatching log, and improve the coverage of the memory storage database. If the predicted power flow change trend is abnormal, enter Step S5 to alarm and adjust the power system.

[0086] Step S5, risk alarm: When it is predicted that the power data on the user side is abnormal or the actual data differs greatly from the predicted data, mark the power data, determine the device (power source / energy storage / electricity load) corresponding to the marked power data as a risk device, and send an alarm notice to the responsible department of the device.

[0087] Specifically, when the output power fluctuation of the distributed power source exceeds the predicted range, for example, the power generation of a photovoltaic power station suddenly decreases due to sudden weather changes. It is judged that the following risks may exist: ① Insufficient power supply, affecting the electricity demand of users; ② Reduced power grid stability, which may lead to power grid frequency deviation. The provided handling suggestions include: ① Mark abnormal devices: Mark the relevant power source devices as risk devices; ② Alarm notice: Notify the power source operation and maintenance department and provide the detailed operation data of the device and the prediction deviation analysis in real time; ③ Enable backup power: According to the dispatching strategy, start the backup generator or draw power from the power grid; ④ Update the prediction model: Retrain the prediction model with real-time data, optimize the algorithm parameters, and improve the prediction accuracy.

[0088] Specifically, when the charge-discharge power of the energy storage device exceeds the preset range, for example, the battery pack experiences over-discharge. The following risks may be judged: ① The lifespan of the energy storage device is shortened, and even safety accidents (such as thermal runaway) may occur; ② The power supply to the electrical load is interrupted, affecting the user experience. The provided handling suggestions include: ① Mark the abnormal device: Locate the energy storage device and mark the abnormal data; ② Alarm notification: Notify the maintenance department of the energy storage system and attach a report on the device operation status and prediction error analysis; ③ Load transfer: Partially transfer the current load to other energy storage or power supply devices to reduce the dependence on the faulty device; ④ Equipment inspection and repair: Immediately conduct a comprehensive inspection of the energy storage device, including battery temperature, SOC (state of charge), etc.; ⑤ Optimize the algorithm: Improve the state-of-charge prediction model of the energy storage and introduce more real-time monitoring data sources.

[0089] Specifically, when the actual power of the electrical equipment exceeds the predicted value, for example, a large industrial load device experiences abnormal energy consumption increase due to a fault. The following risks may be judged: ① The electrical equipment is damaged or aged; ② The power grid load is overloaded, which may lead to regional power outages. The corresponding handling methods are: ① Mark the abnormal device: Mark the load device and its related data as risk objects; ② Alarm notification: Send an abnormal alarm to the device management party, including the power change trend and prediction deviation analysis; ③ Intelligent scheduling: Activate the demand response mechanism to reduce the power of non-critical loads and relieve the system pressure; ④ Investigate load anomalies: The operation and maintenance team needs to check the device operation status, eliminate faults or optimize the power consumption configuration; ⑤ Improve the prediction algorithm: Strengthen the learning of user behavior and changes in power consumption patterns to enhance the adaptability of the algorithm to load fluctuations.

[0090] Specifically, when multiple devices have anomalies simultaneously, for example, power fluctuations and energy storage faults occur at the same time. The following risks may be judged: ① Regional power interruption or power quality degradation; ② Multiple devices are damaged, resulting in high repair costs. The corresponding handling methods are: ① Hierarchical alarm: Set a hierarchical alarm mechanism according to the risk severity and give priority to handling high-risk devices; ② Coordination and optimization: Optimize the real-time scheduling based on the immune genetic algorithm and dynamically adjust the distribution of power sources, energy storage, and loads; ③ Linkage response: Trigger the multi-department cooperation mechanism to ensure information sharing and rapid processing; ④ System review and improvement: After the abnormal event, review the data prediction and processing plan to improve the algorithm robustness and device cooperation efficiency.

[0091] This method is different from the traditional power flow prediction methods that focus on the "grid side", highlighting the uniqueness and advantages of the user side: by directly collecting real-time data from multiple power sources, multiple energy storage devices, and electricity loads on the user side, and combining with an efficient prediction model constructed by the immune genetic algorithm, it can accurately identify the dynamic relationships between user-side devices and monitor the operating status in real time. Compared with the macro centralized prediction on the grid side, the user-side method of this method is closer to the equipment operation site, can quickly respond to local changes, improve real-time performance and agility, especially performs outstandingly in anomaly detection and alarm, effectively reduces the accidental risks caused by the user side, further improves the safety and stability of the overall power system, and at the same time provides precise support for power dispatching and optimization strategies on the user side.

[0092] Embodiment 2

[0093] Based on Embodiment 1, refer to Figure 3 , this embodiment provides a real-time power flow prediction system for multiple power sources, multiple energy storages, and electricity loads on the user side based on immune genetics, including:

[0094] (1) Distributed data acquisition module, used to obtain the real-time operation data of multiple power sources, multiple energy storage devices, and electricity loads on the user side. Specifically, this module is used to implement the function of step S1 in Embodiment 1.

[0095] (2) Prediction model construction module, used to construct a real-time power flow prediction model combined with the immune genetic algorithm. Specifically, this module is used to implement the function of step S2 in Embodiment 1.

[0096] (3) Data analysis module, used to identify the dynamic relationships between multiple power sources, multiple energy storages, and electricity loads based on the real-time operation data. Specifically, this module is used to implement the function of step S3 in Embodiment 1.

[0097] (4) Real-time prediction module, used to perform power flow prediction based on the real-time operation data using the real-time power flow prediction model. Specifically, this module is used to implement the function of step S4 in Embodiment 1.

[0098] Preferably, this system further includes:

[0099] (5) Risk warning module, used to respond to the abnormal user-side power data predicted, or when the gap between the real-time operation data and the predicted data exceeds the preset threshold, take the device corresponding to the abnormal power data as a risk device and send the corresponding warning notification. Specifically, this module is used to implement the function of step S5 in Embodiment 1.

[0100] When this system predicts the real-time power flow of multiple power sources, multiple energy storages and electrical loads on the user side based on the immune genetic algorithm, it first collects real-time operation data from user-side devices, including voltage, current, power and status information, etc. After obtaining the real-time data, a real-time power flow prediction model is constructed through the immune genetic algorithm. In the process of establishing the prediction model, the collected real-time data, historical operation data and possible operation scenarios are combined to generate the real-time power flow prediction model. Identify and analyze the dynamic relationship between multiple power sources, multiple energy storage devices and electrical loads, retrieve the real-time power flow prediction model, and predict the change trend of the power flow, so as to realize the real-time monitoring and prediction of the power system.

[0101] At the same time, based on the immune genetic algorithm, according to the generated real-time power flow prediction model, predict whether the operation status of the device is normal. Compare whether the voltage, current, power and status information match the expected operation parameters. When one or two parameters are detected to have deviations, warning information is sent to relevant personnel and the device is adjusted. If the parameters return to normal after adjustment, it is predicted that the device is operating normally; if the adjustment fails, it is predicted that the device is operating abnormally. When three or more parameters are detected to have deviations, it is directly predicted that the device is operating abnormally, an alarm message is sent to relevant personnel and emergency measures are taken. By setting the threshold of parameter deviation to predict the abnormal operation of the device, the impact of accidental factors on the system can be reduced, the system risk can be lowered, and the safety and stability of the power system can be improved.

[0102] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for real-time power flow prediction of multiple power sources, multiple energy storage and power load on the user side based on immune genetics, characterized in that: The steps include: Step S1, obtaining real-time operation data of multiple power sources, multiple energy storage devices and power loads on the user side; Step S2, constructing a real-time power flow prediction model based on immune genetic algorithm; Step S3, based on the real-time operation data, identifying the dynamic relationship between multiple power sources, multiple energy storages and power loads; Step S4: Based on the real-time operation data, power flow prediction is performed using the real-time power flow prediction model.

2. According to claim 1, a method for real-time power flow prediction of multiple power sources, multiple energy storage and power load at the user side based on immune genetics, characterized in that: The step S4 comprises: Step S401, using the real-time operation model to analyze the power flow on the user side, traverse the power data, screen the valid data and compare it with the output of the prediction model, identify the power flow change trend, and screen the matching power dispatch and optimization strategy; Step S402 , comparing the real-time power flow data with the output of the real-time operation model, and providing accurate power dispatching suggestions and optimization strategies.

3. According to claim 1, a method for real-time power flow prediction of multiple power sources, multiple energy storage and power load at the user side based on immune genetics, characterized in that: The step S3 comprises: Step S301, based on the real-time operation data, effective data is screened through correlation analysis, and by comparing the effective data with historical power flow forecast data, a dynamic relationship between multiple power sources, multiple energy storage devices and power loads is identified, and a dynamic relationship model is constructed; Step S302: predicting the real-time power flow change trend at the user side based on the dynamic relationship model.

4. According to claim 1, a method for real-time power flow prediction of multiple power sources, multiple energy storage and power load at the user side based on immune genetics, characterized in that: The step S2 comprises: Step S201, storing the collected real-time operation data into an incremental database; Step S202, based on the collected real-time operation data, generate a first power flow feature through crossover and mutation and store it in the database; Step S203, matching the second power flow characteristics that match the collected real-time operation data from the database and storing them in the database; Step S204: generating a real-time power flow prediction model that satisfies the node power equation constraint based on the initial characteristic data, the first power flow characteristic and the second power flow characteristic in the database.

5. According to the method of real-time power flow prediction of multiple power sources, multiple energy storage and power load at the user side based on immune genetics in claim 1, it is characterized in that: In the step S1, after obtaining the real-time operation data, the step further includes: Step S101, synchronization of real-time operation data is achieved through time series alignment and interpolation; Step S102, aligning the real-time operation data to the same time scale based on a preset power data collection cycle; Step S103, for missing data points, linear interpolation, spline interpolation or dynamic interpolation based on historical trends are used to fill in the missing data points; Step S104: filtering and denoising the real-time operation data.

6. According to the method of real-time power flow prediction of multiple power sources, multiple energy storage and power load at the user side based on immune genetics in claim 1, it is characterized by: In step S1, the real-time operating data of multiple power sources include at least one of real-time voltage, current, power, frequency, power factor, and equipment load; the real-time operating data of multiple energy storage devices include at least one of charging and discharging power, battery voltage, current, state of charge, temperature, historical power load, and equipment operating status; the real-time operating data of power load includes at least one of real-time power load data, historical load data, meteorological data, and time series data.

7. According to claim 1, a method for real-time power flow prediction of multiple power sources, multiple energy storage and power load at the user side based on immune genetics, characterized in that: Also includes: Step S5, in response to the predicted user-side power data being abnormal, or the difference between the real-time operation data and the predicted data exceeding a preset threshold, the device corresponding to the abnormal power data is treated as a risk device and a corresponding alarm notification is sent.

8. The method for real-time power flow prediction of multiple power sources, multiple energy storage and power load at the user side based on immune genetics according to claim 7 is characterized in that: Step S5 also includes: In response to the predicted user-side power data being within a preset normal range, a power dispatch log is generated.

9. A real-time power flow prediction system for multiple power sources, multiple energy storage and power loads on the user side based on immune genetics, characterized in that: include: Distributed data acquisition module, used to obtain real-time operation data of multiple power sources, multiple energy storage devices and power loads on the user side; Prediction model building module, used to build a real-time power flow prediction model based on immune genetic algorithm; A data analysis module, used to identify the dynamic relationship between multiple power sources, multiple energy storages and power loads based on the real-time operation data; The real-time prediction module is used to perform power flow prediction based on the real-time operation data using the real-time power flow prediction model.

10. The real-time power flow prediction system of multiple power sources, multiple energy storage and power load at the user side based on immune genetics according to claim 9 is characterized in that: Also includes: The risk warning module is used to respond to the predicted abnormal user-side power data, or the difference between the real-time operation data and the predicted data exceeds a preset threshold, and treat the device corresponding to the abnormal power data as a risk device and send a corresponding alarm notification.

Images