User side energy storage hybrid regulation capability cooperative control method and system

Abstract

The invention provides a user side energy storage hybrid regulation capability cooperative control method, which comprises the following steps: collecting multi-source data, matching a scene label, and associating a scene database based on the scene label; generating a power grid change trend through a power grid edge trend pre-judgment strategy based on the multi-source data, generating an emergency degree parameter, calling the adaptation degree of a corresponding model from a scene database according to a scene label, calculating a priority coefficient through a priority model, obtaining a corresponding response model according to the priority coefficient, and obtaining a response result according to the response model; the response mode comprises a rigid priority mode, a flexible priority mode and a fusion control mode; according to the response mode, a corresponding control model is called to execute energy storage charging and discharging adjustment, the control model comprises a rigid control model adaptive to a scene adaptation degree threshold value and a multi-constraint flexible adjustment model capable of adaptively selecting an algorithm, and smooth transition of the rigid model and the flexible model is achieved by fusing the control mode based on pre-judgment parameters.

Description

Technical Field

[0001] This invention relates to energy storage user management, and in particular to a method and system for coordinated control of user-side energy storage hybrid regulation capabilities. Background Technology

[0002] With the increasing penetration of new energy power generation in the power system, the intermittency and volatility of its output pose a severe challenge to the stable operation of the power grid. User-side energy storage systems, as key devices for smoothing load fluctuations and responding to grid dispatch commands, have become an important component of the new power system. However, existing user-side energy storage regulation and control methods still suffer from the following core problems: Most current control methods only collect grid dispatch commands and basic data from the energy storage system, failing to fully integrate user electricity consumption data with real-time user load data, and lacking association with user type, electricity consumption time, and other scenario-specific tags. For example, industrial users have extremely high requirements for power supply continuity during production periods, while residents are sensitive to electricity comfort at night. However, existing methods do not develop differentiated control strategies for different scenarios, leading to potential conflicts between the regulation process and actual user needs, affecting user experience. Existing methods largely rely on real-time grid commands to trigger regulation, lacking the ability to predict grid change trends and failing to prepare for responses in advance. This results in delays in the energy storage system's response when millisecond-level or second-level emergency dispatch commands occur. Meanwhile, the calculation of urgency parameters often only considers single factors such as grid frequency deviation and command response time, without taking into account the characteristics of the scenario for comprehensive judgment. This can easily lead to over-response or under-response, making it difficult to balance grid security and regulation economy. Summary of the Invention

[0003] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a user-side energy storage hybrid regulation capability coordinated control method and system to overcome the above-mentioned defects in the existing technology.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A method for coordinated control of user-side energy storage hybrid regulation capabilities includes:

[0006] The data acquisition steps include collecting multi-source data and matching scene tags. The multi-source data includes power grid dispatch command data, user-side energy storage system data, user electricity consumption data, and user-side real-time load data. The scene tags include user type tags and electricity consumption time period tags. The scene database is associated with the scene tags.

[0007] The dynamic response step generates a power grid change trend based on multi-source data through a power grid edge trend prediction strategy, generates an urgency parameter, and retrieves the adaptation degree of the corresponding model from the scenario database according to the scenario label. Through the priority model, the priority coefficient is calculated, and the corresponding response model is obtained according to the priority coefficient. The response mode includes rigid priority mode, flexible priority mode and fusion control mode.

[0008] The execution steps involve calling the corresponding control model to perform energy storage charging and discharging regulation according to the response mode. The control model includes a rigid control model with an adaptation threshold for the scenario and a multi-constraint flexible adjustment model with an adaptive algorithm. The fusion control mode is based on the predicted parameters to achieve a smooth transition between the rigid model and the flexible model.

[0009] Preferably, the power grid change trend prediction strategy includes:

[0010] The data input step is used to collect data from multiple sources and generate a standardized dataset through standardization processing.

[0011] The scenario analysis step involves extracting scenario-specific feature vectors from a standardized dataset based on scenario labels. These feature vectors include the urgency of instructions, the frequency of instruction occurrence, and the variance of load fluctuations within the corresponding scenario.

[0012] The urgency generation step involves inputting a scenario-based feature vector into the prediction model, outputting a predicted urgency value for a fixed short period of time for the power grid command, and then using the predicted urgency value and real-time load data to calculate the prediction error compensation value to generate urgency parameters.

[0013] Preferably, the step of calling the corresponding model's adaptability from the scene database based on scene tags includes calling the basic scene adaptability from the scene parameter library and dynamically adjusting it based on user type tags and user time period tags. The priority model calculates the priority coefficient based on the urgency parameter and scene adaptability through a weighting algorithm.

[0014] Preferably, the rigid control model includes a method for acquiring user-side energy storage system data. The user-side energy storage system data includes grid frequency parameters and energy storage SOC parameters, and grid frequency thresholds and energy storage threshold ranges are set respectively. When the deviation value of the grid frequency parameter exceeds the grid frequency threshold or the energy storage SOC parameter is not within the energy storage threshold range, the rigid control model is triggered, and charging and discharging rules are preset. The charging and discharging rules include SOC threshold rules and frequency deviation power adjustment rules. If only a single condition is triggered, the charging and discharging operation is performed according to the corresponding rule. If two conditions are triggered simultaneously, the power upper limit is limited first according to the SOC threshold rule, and then the frequency deviation power adjustment rule is superimposed.

[0015] Preferably, the SOC threshold rule includes triggering charging power limitation when the energy storage SOC parameter is less than the energy storage threshold range, and triggering discharging power limitation when the energy storage SOC parameter is greater than the energy storage threshold range.

[0016] The frequency deviation power regulation rule includes generating multi-level deviations based on the deviation values ​​of the grid frequency parameters, and generating differentiated power regulation amplitudes based on the multi-level deviations.

[0017] As a preferred embodiment, the preferred model of the multi-constraint flexible adjustment model of the adaptive selection algorithm includes setting multi-dimensional optimization objectives, including maximizing the grid regulation compliance rate, minimizing user electricity costs, and minimizing energy storage lifespan loss costs. Optimization values ​​are obtained according to the preset weights of each optimization objective, and preset constraints are set. The optimal charging and discharging power is obtained based on the constraints, including transformer capacity, energy storage SOC, and user electricity consumption.

[0018] As a preferred option, a dynamic response feedback step is also included. After the priority model outputs the optimal charging and discharging power and controls the energy storage execution steps, data monitoring is initiated to obtain execution data. Based on the execution data, the deviation value of key indicators is calculated, and the matching degree between the priority model and the actual execution effect is determined. Based on the matching degree, deviation analysis results are generated, and the priority model is dynamically adjusted.

[0019] A user-side energy storage hybrid regulation capability coordinated control system includes:

[0020] The data acquisition module collects multi-source data and matches it with scene tags. The multi-source data includes power grid dispatch command data, user-side energy storage system data, user electricity consumption data, and user-side real-time load data. The scene tags include user type tags and electricity consumption time period tags. The module associates the scene database based on the scene tags.

[0021] The dynamic response module generates power grid change trends based on multi-source data through a power grid edge trend prediction strategy, generates urgency parameters, obtains scenario adaptability under different response modes based on scenario tags, calculates priority coefficients through a priority model, and obtains the corresponding response model based on the priority coefficients. The response modes include rigid priority mode, flexible priority mode, and fusion control mode.

[0022] The execution module calls the corresponding control model to perform energy storage charging and discharging regulation according to the response mode. The control model includes a rigid control model with an adaptation threshold for the scenario and a multi-constraint flexible adjustment model with an adaptive algorithm. The fusion control mode is based on the predicted parameters to achieve a smooth transition between the rigid model and the flexible model.

[0023] The beneficial effects of this invention are as follows: The grid edge trend prediction strategy extracts scenario-specific feature vectors and combines them with prediction models and real-time load data to calculate error compensation values, generating urgency parameters that better match scenario requirements. For example, in commercial scenarios, urgency is corrected by the frequency of command occurrences and load fluctuation variance, and the response model is preheated 10-15 minutes in advance, shortening the response delay of the energy storage system to emergency commands while avoiding over-response in non-emergency scenarios and reducing energy storage lifespan loss costs. The dynamic response step calculates priority coefficients through a priority model and combines them with prediction parameters to achieve proactive adaptation of the response mode; in the execution step, the fusion control mode, based on prediction parameters, first uses a rigid model to ensure basic response, and then uses a flexible model to gradually optimize power output, achieving a smooth transition between rigid and flexible modes; simultaneously, the three response modes are dynamically matched through priority coefficients to meet the needs of scenarios with dynamic changes in grid commands. The rigid control model dynamically adjusts the adaptation threshold based on scene labels to avoid adjustment deviations caused by fixed thresholds; the flexible control model can adaptively select algorithms to achieve accurate matching of algorithm scene timeliness, which avoids waste of computing power and ensures optimization accuracy, so that the response time of the flexible model meets the adjustment requirements at the 2-5 minute level, while reducing the cost of energy storage charging and discharging life loss. Attached Figure Description

[0024] Figure 1 This is an overall structural diagram of the present invention;

[0025] Figure 2 This is a priority response flowchart of the present invention;

[0026] Figure 3 This is a flowchart of the multi-constraint flexible adjustment process of the present invention. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] It should be noted that when a component is said to be fixed to another component, it can be directly on the other component or it may have a component in between. When a component is said to be connected to another component, it can be directly connected to the other component or it may have a component in between. When a component is said to be set to another component, it can be directly set to the other component or it may have a component in between. The terms vertical, horizontal, left, right, and similar expressions used in this document are for illustrative purposes only.

[0029] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The terminology used herein includes, and / or encompasses, any and all combinations of one or more of the associated listed items.

[0030] The embodiments of the present invention will be further described in detail below with reference to the accompanying drawings:

[0031] A method for coordinated control of user-side energy storage hybrid regulation capabilities includes:

[0032] The data acquisition process involves collecting multi-source data and matching it with scenario tags. The multi-source data includes grid dispatch command data (real-time and historical), with real-time data including command type, urgency level, regulation power demand, regulation compensation standard, and command validity period; historical data including command records for the past 30 days, command execution result feedback; user-side energy storage system data including energy storage SOC, SOC change rate, real-time charging and discharging power, energy storage battery cell voltage, temperature, transformer real-time load rate, energy storage charge and discharge cycle count, and lifespan loss coefficient; and user electricity consumption data including real-time power consumption, branch circuit power consumption, and historical power consumption. The system includes load curves, electricity comfort thresholds, electricity cost data, and real-time load data for users, such as total load of distribution transformer areas, load growth rate, load fluctuation variance, renewable energy output data, and load forecast data. Scenario tags include user type tags and electricity consumption period tags. Based on these tags, a scenario database is linked. User type tags are divided into three categories according to user electricity consumption characteristics, each corresponding to differentiated electricity demand. These include residential user tags: characterized by low load fluctuations (e.g., nighttime use primarily for lighting and home appliances), sensitivity to electricity comfort (e.g., continuous power supply for air conditioners and water heaters), and no continuous production needs. Typical users include residential communities. Residential users; Commercial users are categorized by their electricity consumption characteristics: large load fluctuations, such as high load during shopping mall hours and low load after closing; concentrated electricity consumption periods, such as 9:00-22:00; sensitivity to economic costs, requiring peak-valley arbitrage to reduce electricity bills; typical users include shopping malls, office buildings, and hotels. Industrial users are categorized by their electricity consumption characteristics: stable loads, such as continuous production line operation; extremely high requirements for power supply continuity, as interruptions may lead to production losses; numerous constraints, such as transformer capacity and power consumption restrictions in production processes; typical users include factories and industrial parks. Users are classified into three categories based on their electricity load intensity and grid peak-valley periods, and combined with user type tags to form combined scenarios: Peak hour tags correspond to peak hours in the grid's peak-valley electricity pricing, such as concentrated appliance use for residential users from 18:00 to 22:00, peak customer traffic for commercial users from 10:00 to 12:00, and peak production for industrial users from 8:00 to 12:00. Flat hour tags correspond to flat hour electricity pricing periods, such as 9:00 to 18:00 for residential users, 12:00 to 17:00 for commercial users, and 12:00 to 18:00 for industrial users. Off-peak hour tags correspond to off-peak electricity pricing periods, such as 22:00 to 6:00 for residential users, 22:00 to 9:00 for commercial users, and 22:00 to 8:00 for industrial users. The scenario database is the core data carrier storing control parameters, constraints, and algorithm rules under different scenarios. It adopts a hierarchical architecture of main database and sub-databases. The main database stores the mapping relationship between scenario tags and sub-databases. The sub-databases are divided into three sub-databases according to user type tags, and each sub-database is further subdivided into parameter tables according to electricity consumption time tags.

[0033] The dynamic response process involves generating power grid change trends based on multi-source data and a power grid edge trend prediction strategy, as well as generating urgency parameters. It then retrieves the appropriate model's fit from the scenario database based on scenario labels. Priority coefficients are calculated using a priority model, and the corresponding response model is obtained based on these coefficients. Response modes include rigid priority mode, flexible priority mode, and fusion control mode. The priority coefficient is a quantitative indicator of the system state's overall urgency parameter and scenario fit. Priority coefficients are calculated for each of the three response models. Based on the priority coefficients and scenario-based dynamic thresholds, the optimal response model is matched. The thresholds are retrieved from the scenario database according to scenario labels.

[0034] Strategies for predicting power grid change trends include:

[0035] The data input step is used to collect multi-source data and generate a standardized dataset through standardization processing. The multi-source data includes historical data of power grid dispatch instructions in the past 48 hours, real-time power grid operation parameters, power grid supply and demand gap prediction data, total load power of user-side distribution transformer areas, real-time power consumption of users themselves, real-time output power of new energy sources in the transformer area, load growth rate, load fluctuation variance; real-time value of energy storage SOC, SOC change rate, real-time load rate of transformers, and number of energy storage charge and discharge cycles. Outliers and incomplete data are removed, and normalization transformation is achieved to ensure data quality and format consistency.

[0036] The scenario analysis steps involve extracting scenario-specific feature vectors from a standardized dataset based on scenario labels. These vectors include the urgency of instructions, the frequency of instruction occurrence, and the variance of load fluctuations within the corresponding scenario. Based on the combination of user type labels and electricity consumption period labels, pre-defined differential feature extraction rules are used. The urgency of instructions refers to the urgency level of current and recent power grid dispatch instructions, the frequency of instruction occurrence is the number of times power grid dispatch instructions occur per unit time, and the variance of load fluctuations reflects the degree of load data fluctuation in a recent period, indicating the stability of the user-side load.

[0037] The urgency generation step involves inputting a scenario-based feature vector into a prediction model, which outputs a predicted urgency value for a fixed short period of time. This predicted urgency value is then compared with real-time load data to calculate a prediction error compensation value, generating an urgency parameter. This step is crucial for transforming the scenario-based feature vector into a quantifiable urgency parameter. By using the prediction model to output a base value and calculating a compensation value from real-time data, the accuracy of the urgency parameter is improved, ensuring it accurately reflects the urgency of the grid's adjustment needs in the next 10-15 minutes. The scenario-based feature vector is input into an LSTM model adapted to the current scenario parameters. The model outputs a base value for predicting the urgency of the grid command in the next 10-15 minutes, along with a prediction confidence level. Based on real-time load data, a compensation value is calculated to correct the base value, ultimately yielding the urgency parameter.

[0038] The system retrieves the appropriate model's fit from the scenario database based on scenario tags. This includes retrieving basic scenario fit from the scenario parameter library and dynamically adjusting it based on user type tags and user time period tags. Priority models calculate their priority coefficients using a weighting algorithm based on urgency parameters and scenario fit. The scenario database employs a two-level classification and three-dimensional parameter architecture. The first-level classification is user type tags, the second-level classification is electricity consumption time period tags, and the three-dimensional parameters are response model type, fit evaluation dimension, basic weight, and threshold. The system obtains the current scenario tag and locates the corresponding user type and electricity consumption time period subclass from the scenario database. For the three types of response models—rigid priority, flexible priority, and fusion control—the system extracts the fit evaluation dimension and basic weight for each category, forming a basic fit evaluation framework. Based on scenario thresholds in the scenario database and combined with real-time data, the system triggers fit weight adjustments to ensure the fit can adapt to dynamic scenario changes. The adjustment rules are divided into three categories: weight increase, weight decrease, and dimension addition.

[0039] Weighting adjustments are made when scenario risks increase: When real-time data reaches the upper limit of the scenario threshold, the basic weight of the corresponding evaluation dimension is increased, while the weight of other non-core dimensions is reduced.

[0040] When the weight is reduced and the scenario risk is reduced: when the real-time data is below the lower limit of the scenario threshold; the weight of the core dimension is reduced and the weight of the economic or comfort-related dimensions is increased.

[0041] Dimension additions and adjustments: In the event of sudden scenario factors: When factors that are not preset in the scenario database but affect adaptability occur, a sudden load adaptability assessment dimension is temporarily added, and the weights are distributed from other dimensions. This dimension is deleted after the scenario recovers.

[0042] By integrating urgency parameters, scenario adaptability, and system state parameters, priority coefficients for the three types of response models are calculated using a scenario-based weighting algorithm, and the optimal response model is finally matched.

[0043] The execution steps involve calling the corresponding control model to perform energy storage charging and discharging regulation based on the response mode. The control models include a rigid control model with an adaptation threshold to the scenario and a multi-constraint flexible regulation model with adaptive algorithm selection. The fusion control mode, based on predicted parameters, achieves a smooth transition between the rigid and flexible models. The fusion control model, based on the predicted parameters output from the dynamic response steps, achieves a smooth connection between the rigid model's basic response and the flexible model's optimized regulation, avoiding power fluctuations and obtaining grid trend prediction results and the transition power pre-calculated by the flexible model. The rigid model's basic response initially executes the basic power according to the rigid control model, while simultaneously switching the flexible model to a fast calculation mode to prepare for the transition. The rigid power is gradually transitioned to flexible power according to a time-power step curve, with the transition duration adapting to scenario requirements.

[0044] The rigid control model includes data acquisition for user-side energy storage systems. This data includes grid frequency parameters and energy storage SOC parameters, each with a set grid frequency threshold and energy storage threshold range. When the grid frequency parameter deviation exceeds the grid frequency threshold or the energy storage SOC parameter is outside the energy storage threshold range, the rigid control model is triggered, and pre-set charging and discharging rules are implemented. These rules include SOC threshold rules and frequency deviation power adjustment rules. If only a single condition is triggered, the charging and discharging operation is performed according to the corresponding rule. If both conditions are triggered simultaneously, the power limit is limited first according to the SOC threshold rule, and then the frequency deviation power adjustment rule is applied. Based on the scene tags output by the dynamic response steps, the system retrieves the appropriate grid frequency threshold and energy storage SOC threshold from the scene database to avoid scene mismatch caused by fixed thresholds. It collects user-side energy storage system data in real time, including grid frequency parameters and energy storage SOC parameters. Frequency triggering: When the grid frequency parameter deviation exceeds the scene-appropriate grid frequency threshold, frequency deviation rigid control is triggered. SOC triggering: When the SOC is less than the lower threshold or greater than the upper threshold, SOC threshold rigid control is triggered. Dual-condition triggering: When both of the above conditions are met simultaneously, dual-condition rigid control is triggered. If both conditions are triggered simultaneously, the power upper limit is limited first according to the SOC threshold rule, and then the frequency deviation power adjustment rule is superimposed. The power upper limit is calculated first according to the SOC threshold rule, and then the amplitude of the frequency deviation adjustment rule is superimposed to ensure equipment safety is prioritized.

[0045] The SOC threshold rules include triggering charging power limitation when the energy storage SOC parameter is less than the energy storage threshold range, and triggering discharging power limitation when the energy storage SOC parameter is greater than the energy storage threshold range; charging power limitation is triggered when the SOC is less than the lower threshold limit.

[0046] The calculation formula is: Upper limit of charging power P_charging limit = Rated charging power of energy storage × (SOC / lower limit of threshold) × 0.8;

[0047] Discharge power limit: SOC greater than the upper threshold:

[0048] The calculation formula is: Upper limit of discharge power P = Rated discharge power of energy storage × (1 - (SOC - upper limit of threshold) / (100% - upper limit of threshold)) × 0.8;

[0049] The frequency deviation power regulation rule includes generating multi-level deviations based on the grid frequency parameter deviation value, generating differentiated power regulation amplitudes based on the multi-level deviations, and dividing the deviations into three levels based on the magnitude of the grid frequency parameter deviation value, corresponding to differentiated power regulation amplitudes.

[0050] The adaptive algorithm selection multi-constraint flexible regulation model prioritizes a multi-dimensional optimization objective, including maximizing grid regulation compliance rate, minimizing user electricity costs, and minimizing energy storage lifetime loss costs. Optimization values ​​are obtained based on preset weights for each objective, with pre-set constraints. The optimal charging and discharging power is then obtained based on these constraints, including transformer capacity, energy storage SOC, and user electricity consumption. By setting scenario-based optimization objective weights and constraints, an efficient algorithm is adaptively selected to calculate the optimal charging and discharging power, achieving a balance between economic benefits and user needs. Optimization objective weights are set based on scenario labels to ensure consistency between objectives and core scenario requirements. Grid regulation compliance rate: the proportion of actual charging and discharging power falling within the grid commanded power range; User electricity costs: real-time electricity price × user electricity consumption + energy storage charging and discharging loss costs; Energy storage lifetime loss costs: loss rate per charging and discharging cycle × total energy storage lifetime cost / total number of cycles. Optimization values ​​are obtained through a weighted algorithm. Constraints suitable for the scenario are retrieved from the scenario database to ensure optimal charging and discharging power. Electrical operation meets equipment safety and user needs, with the following core constraints: Transformer capacity constraint: Real-time transformer load rate ≤ scenario adaptation threshold, i.e., Ptransformer = user power consumption + energy storage charging / discharging power ≤ 0.9 × Ptransformer rating; Energy storage SOC constraint: SOC ∈ scenario adaptation threshold range to avoid overcharging and over-discharging; User power consumption constraint: Real-time user power consumption ≥ scenario adaptation ratio × historical lowest power consumption to ensure basic power needs; Based on the complexity of the constraints and response time requirements, an adaptive algorithm is selected to ensure a balance between computational efficiency and optimization accuracy. Algorithms include linear programming, force optimization, and an improved genetic algorithm.

[0051] It also includes a dynamic response feedback step. After the priority model outputs the optimal charging and discharging power and controls the energy storage execution steps, data monitoring is initiated to acquire execution data. Based on the execution data, the deviation values ​​of key indicators are calculated, and the matching degree between the priority model and the actual execution effect is determined. Based on the matching degree, deviation analysis results are generated, and the priority model is dynamically adjusted. Data is collected around four dimensions: power, equipment, cost, and grid, with frequencies set according to scenario differences. Based on the monitoring data and the prediction values ​​of the priority model, four types of core deviations are calculated: power tracking deviation, equipment status deviation, economic cost deviation, and grid compliance deviation. The matching degree is calculated according to scenario-based weights, and deviation analysis results are generated based on the matching degree, and the priority model is dynamically adjusted.

[0052] A user-side energy storage hybrid regulation capability coordinated control system includes:

[0053] The data acquisition module collects multi-source data and matches it with scene tags. The multi-source data includes power grid dispatch command data, user-side energy storage system data, user electricity consumption data, and user-side real-time load data. The scene tags include user type tags and electricity consumption time period tags. The module is then linked to the scene database based on the scene tags.

[0054] The dynamic response module generates power grid change trends based on multi-source data and a power grid edge trend prediction strategy, generates urgency parameters, obtains scenario adaptability under different response modes based on scenario tags, calculates priority coefficients through a priority model, and obtains the corresponding response model based on the priority coefficients. The response modes include rigid priority mode, flexible priority mode, and fusion control mode.

[0055] The execution module calls the corresponding control model to perform energy storage charging and discharging regulation according to the response mode. The control model includes a rigid control model with an adaptation threshold for the scenario and a multi-constraint flexible adjustment model with an adaptive algorithm. The fusion control mode is based on the predicted parameters to achieve a smooth transition between the rigid model and the flexible model.

[0056] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principle of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for collaborative control of user-side energy storage hybrid regulation capability, characterized in that, The method comprises the following steps: Data collection step, collecting multi-source data and matching scene label, the multi-source data includes power grid scheduling instruction data, user side energy storage system data, user power consumption data, user side real-time load data, the scene label includes user type label and power consumption period label, based on the scene label association scene database; Dynamic response step, based on multi-source data through power grid edge trend prediction strategy to generate power grid change trend, and generate emergency parameter, and according to scene label from scene database calling corresponding model adaptation degree, through priority model, calculate priority coefficient, according to the priority coefficient obtains corresponding response model, the response mode includes rigid priority mode, flexible priority mode and fusion control mode; Execution step, according to the response mode calling corresponding control model executes energy storage charge and discharge adjustment, the control model includes rigid control model suitable for scene adaptation degree threshold and multi-constraint flexible adjustment model which can select algorithm adaptively, fusion control mode based on prediction parameter, realize the smooth transition of rigid model and flexible model.

2. The user-side energy storage hybrid regulation capability coordinated control method according to claim 1, characterized in that, The power grid change trend prediction strategy comprises: Data input step, for collecting multi-source data, and generating standardized data set through standardization processing; Scene analysis step, based on scene label, extract scene feature vector from standardized data set, the scene feature vector includes instruction emergency degree, instruction frequency and load fluctuation variance under corresponding scene; Emergency generation step, input scene feature vector into prediction model, output power grid instruction emergency prediction value in fixed time, and generate emergency parameter through calculating prediction error compensation value between emergency prediction value and real-time load data. 3.The user-side energy storage hybrid regulation capability coordinated control method according to claim 1, characterized in that, According to the scene label from the scene database calling corresponding model adaptation degree, including according to user type label and user period label, calling basic scene adaptation degree from scene parameter library and dynamically adjusting, the priority model calculates the priority coefficient according to the emergency parameter, scene adaptation degree through weight algorithm. 4.The user-side energy storage hybrid regulation capability coordinated control method according to claim 1, characterized in that, The rigid control model includes user side energy storage system data, the user side energy storage system data includes power grid frequency parameter and energy storage SOC parameter, and respectively set power grid frequency threshold and energy storage threshold range, when the power grid frequency parameter deviation value exceeds the power grid frequency threshold or the energy storage SOC parameter is not in the energy storage threshold range, then trigger rigid control model, and preset charge and discharge rule, the charge and discharge rule includes SOC threshold rule and frequency deviation power regulation rule, if only a single condition is triggered, execute charge and discharge operation according to corresponding rule, if two kinds of conditions are triggered at the same time, preferentially according to SOC threshold rule limit power upper limit, then superimpose the frequency deviation power regulation rule.

5. The user-side energy storage hybrid regulation capability coordinated control method according to claim 4, characterized in that, The SOC threshold rule includes when the energy storage SOC parameter is less than the energy storage threshold range, triggering charging power limit, when the energy storage SOC parameter is greater than the energy storage threshold range, triggering discharging power limit; The frequency deviation power regulation rule includes generating multi-level deviation according to the power grid frequency parameter deviation value, generating differentiated power regulation amplitude according to the multi-level deviation. 6.The user-side energy storage hybrid regulation capability coordinated control method according to claim 1, characterized in that, The multi-constraint flexible adjustment model of the adaptive selection algorithm includes a priority model, which sets a multi-dimensional optimization target, including maximizing the grid regulation compliance rate, minimizing the user electricity cost, and minimizing the energy storage life loss cost, obtains an optimization value according to a preset weight of each optimization target, and has a preset constraint condition, obtains optimal charging and discharging power based on the constraint condition, and the constraint condition includes transformer capacity, energy storage SOC, and user electricity power.

7. The user-side energy storage hybrid regulation capability coordinated control method according to claim 1, characterized in that, The dynamic response feedback step further includes starting data monitoring to obtain execution data after the priority model outputs the optimal charging and discharging power and controls the energy storage execution step, calculating a key indicator deviation value based on the execution data, judging the matching degree of the priority model and the actual execution effect, generating a deviation analysis result according to the matching degree, and dynamically adjusting the priority model.

8. A user-side energy storage hybrid regulation capability coordinated control system, characterized in that, The method comprises The data acquisition module acquires multi-source data and matches scene labels, the multi-source data includes grid scheduling instruction data, user-side energy storage system data, user electricity data, and user-side real-time load data, the scene labels include user type labels and electricity time period labels, and the scene database is associated based on the scene labels; The dynamic response module generates a grid change trend based on the multi-source data through a grid edge trend prediction strategy, generates an emergency degree parameter, obtains a scene adaptation degree in different response modes according to the scene labels, calculates a priority coefficient through the priority model, obtains a corresponding response model according to the priority coefficient, and the response mode includes a rigid priority mode, a flexible priority mode, and a fusion control mode; The execution module calls a corresponding control model to execute energy storage charging and discharging adjustment according to the response mode, the control model includes a rigid control model that adapts to the scene adaptation degree threshold and a multi-constraint flexible adjustment model of an adaptive selection algorithm, and the fusion control mode realizes smooth transition of the rigid model and the flexible model based on the prediction parameter.

Images