Interaction value evaluation method and system considering random association of power grid adjustment demand and user power utilization response
By constructing a stochastic correlation between grid regulation demand and user electricity response, multi-dimensional value assessment and dynamic optimization are achieved, solving the problems of assessment result bias and strategy lag in grid dispatching, and improving the efficiency and stability of grid-user interaction.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- LIAOYANG POWER SUPPLY COMPANY OF STATE GRID LIAONING ELECTRIC POWER SUPPLY
- Filing Date
- 2025-11-21
- Publication Date
- 2026-04-17
AI Technical Summary
Existing power grid dispatching methods are unable to respond accurately and in real time to uncertain load changes. The assessment results deviate significantly from the actual situation. The randomness and lag in user response lead to delayed strategy adjustments. Furthermore, the assessment dimensions are too simplistic and do not adequately consider the complexity of the interaction between the power grid and users.
We construct stochastic correlation features between power grid regulation demand and user electricity consumption response. Through multi-dimensional value assessment and dynamic optimization, combined with real-time data adjustment strategies, we adopt an improved random forest algorithm and Hellinger distance to identify key sensitive factors. A dynamic Bayesian network diagnostic strategy is used to verify the adaptability of existing strategies and generate adjustment suggestions.
It reduces assessment error by more than 30%, improves interactive benefits by 20%-40%, shortens strategy adjustment lag to the hour level, enhances renewable energy absorption capacity, reduces grid reserve capacity by 15%-25%, and increases user response participation rate to 60%.
Smart Images
Figure CN121886474A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power dispatching technology, and specifically to an interactive value assessment method and system that considers the stochastic correlation between power grid regulation demand and user electricity consumption response. Background Technology
[0002] The global energy transition is accelerating, with large-scale integration of renewable energy sources such as wind and solar power into the power system. Coupled with the widespread application of smart meters, 5G communication, and distributed energy management systems, the complexity, dynamic rate of change, and multi-dimensional coupling of power grid operations have significantly increased. Grid regulation demands are becoming increasingly complex: electricity load is affected by residential habits, industrial production, and new types of electrical equipment, leading to increased uncertainty, a widening peak-to-valley difference, and a higher probability of extreme loads. Intermittent power output is constrained by natural conditions, exhibiting significant volatility and unpredictability, further exacerbated by extreme weather conditions, putting pressure on grid power balance and security. Traditional grid dispatching relies on historical data and experience models, making it difficult to respond accurately and in real time to these uncertainties. Simply adjusting the output of thermal power units to maintain balance is not only costly but also prone to power shortages or curtailment of wind and solar power, impacting operational efficiency and economics, and contradicting the "dual carbon" goals. Demand response, through incentive mechanisms such as electricity price discounts and cash subsidies, guides users to adjust their electricity consumption behavior according to their needs, becoming an effective solution. However, user responses are random (due to differences in electricity demand, economic capacity, etc., resulting in different willingness to respond) and lagging. Furthermore, the relationship between grid regulation demand and user response is affected by various external factors such as weather and economy, exhibiting complex nonlinear characteristics, which poses challenges to practical applications.
[0003] Current value assessments of power grid-user interaction suffer from four major problems: First, the assessment dimensions are too singular, focusing on economic indicators on the power grid side while neglecting the value of user comfort loss and system reliability improvement. Second, the assessment is mainly static, relying on historical data to build fixed models, making it difficult to adapt to dynamic changes in the power grid, users, and the external environment. Third, there is insufficient handling of stochastic factors, with a predominance of deterministic methods that fail to fully consider the probabilistic correlation characteristics of power grid-user interaction, resulting in low assessment reliability. Fourth, strategy adjustments are lagging, with a lack of effective connection between assessment and strategy formulation, making it difficult to quickly respond to changes in actual conditions and to optimize and adjust in a timely manner. These problems lead to significant deviations between assessment results and reality, making it difficult to support the scientific formulation of demand response strategies and the optimal allocation of power grid resources. Summary of the Invention
[0004] To thoroughly address the problems of single-dimensional, primarily static, insufficient handling of randomness, and lagging strategy adjustment in the value assessment of power grid-user interaction, this invention proposes a related method and system. This scheme first extracts the random correlation characteristics between the two, then performs multi-dimensional value assessment, and dynamically optimizes and adjusts the strategy based on real-time data. The specific technical solution is as follows:
[0005] An interactive value assessment method considering the stochastic correlation between grid regulation demand and user electricity consumption response includes:
[0006] Construct a set of power grid regulation demand parameters and a set of user electricity consumption response parameters, and calculate stochastic correlation coefficients to extract the stochastic correlation characteristics between power grid regulation demand and user electricity consumption response;
[0007] Based on the extracted random association features, the grid-side value, user-side value, and system-side value are calculated, and then the total interaction value is calculated to achieve a multi-dimensional quantitative assessment of interaction value.
[0008] Based on the quantitative assessment results of the base value, the real-time monitoring data of the power grid operation status, and the feedback from user electricity consumption, a dynamic assessment model is constructed, the random correlation characteristics and value assessment results are updated, and dynamic assessment and optimization are completed.
[0009] Preferred options also include:
[0010] Based on the extracted random association features and the multi-dimensional interaction value quantification results, the impact of key variable fluctuations on the total interaction value is analyzed, core sensitive factors are identified, and a calibration benchmark is provided for dynamic adjustment.
[0011] By combining the constructed dynamic evaluation model, the real-time updated interactive value is compared with the preset target value to verify the adaptability of the existing power grid regulation strategy and user incentive mechanism, and attribution analysis is performed on the value deviation items to generate adjustment suggestions.
[0012] Preferably, the extraction of the random correlation features between the power grid regulation demand and user electricity consumption response specifically includes the following:
[0013] Constructing a set of power grid regulation demand parameters ,in To adjust the type of demand at any time, To adjust the intensity, Demand duration; fitted using historical data probability distribution probability density function Discrete probability;
[0014] Construct a set of user power consumption response parameters ,in This represents the response status at time t, where 1 indicates a response and 0 indicates no response. For response quantity; Response delay time; response probability calculated using historical data. Response deviation rate probability distribution The probability distribution, where Plan response volume for users;
[0015] Calculate the random association coefficient The formula for quantifying the matching degree between power grid regulation demand and user response at time t is as follows:
[0016]
[0017] in, The time correlation coefficient, The correlation coefficient is the strength of the correlation. is the state correlation coefficient.
[0018] Preferably, the quantitative assessment of the multi-dimensional interactive value specifically includes the following:
[0019] Calculate the value of the power grid side The formula is as follows:
[0020]
[0021] in, , The peak reduction value per unit at time t; Contribution to frequency modulation at time t This is the frequency deviation correction value. Value per unit of frequency modulation; As a unit of reserve cost, This is the amount of backup capacity to be replenished due to a lack of response.
[0022] User-side value quantification considers both the benefits of response incentives and the loss of electricity comfort:
[0023]
[0024] in, As a unit to respond to incentives, Sacrificing costs for user comfort;
[0025] System-side value quantification considers both improved voltage stability and enhanced power supply reliability.
[0026]
[0027] in, The value of improvement in voltage deviation at time t. To improve power supply reliability, Enhance value by improving unit reliability;
[0028] Total interaction value calculation: Combining the values from all three sides, and considering the expected probability of random associations:
[0029]
[0030] Where T is the evaluation period. For expectation operator, Random association coefficient The probability distribution.
[0031] Preferably, the dynamic evaluation and optimization specifically includes the following:
[0032] Real-time data acquisition and feature updates: Real-time acquisition of power grid regulation demand data and user response data. Power grid regulation demand data includes peak shaving commands issued by the dispatch center, and user response data includes load adjustment amounts reported by smart meters; updates are performed every 15 minutes. , , and refit , Probability distribution, corrected random association coefficient ;
[0033] Value sensitivity analysis, analyzing incentive amount parameters and intensity adjustment For total value The effects are as follows:
[0034]
[0035] Here, x represents the sensitivity coefficient, and x represents the key parameter.
[0036] Interaction strategy optimization: Based on sensitivity analysis result X, optimize the power grid regulation strategy package and user incentive mechanism. The power grid regulation strategy includes adjusting the demand trigger threshold, and the user incentive mechanism includes increasing the incentive amount for high-sensitivity users, as detailed below:
[0037]
[0038] in, Let X be the optimal parameter value, and X be the feasible region of the parameters. , These are the upper and lower limits of the parameter.
[0039] Preferably, the analysis of the impact of key variable fluctuations on the total interaction value employs an improved random forest sensitivity analysis algorithm, with the core formula being:
[0040]
[0041] in, For the first The comprehensive sensitivity index of the variables, For the number of Monte Carlo simulations, For the first In the simulation, the first The random perturbation values of each variable, For a given Conditional variance of total interactive value. For this variable in the th The SHAP value in this simulation is obtained through... The core sensitivity factors are determined by ranking, and then... The top 30% of variables.
[0042] Preferably, in determining the core sensitive factors, a nonlinear correlation verification model is introduced, and Hellinger distance is used to quantify the correlation between the variable distribution and the value distribution.
[0043]
[0044] in, For the first Marginal probability density function of each variable For a given variable value The conditional probability density function of the total interaction value, when If a strong correlation is identified, it is included in the core sensitive factor set.
[0045] Preferably, the verification of existing power grid regulation strategies employs a dynamic Bayesian network deviation diagnosis algorithm to construct a deviation probability model:
[0046]
[0047] in, It is a deviation sequence. Given the core sensitivity factor parameter set, the posterior probability is solved using the Markov chain Monte Carlo method. When the maximum posterior probability corresponds to If three or more factors deviate from the confidence interval, the strategy is deemed to have failed.
[0048] Preferably, the generation of adjustment suggestions adopts a multi-objective optimization adjustment model, and the objective function is:
[0049]
[0050]
[0051] in, , Weighting coefficients ( ), The L2 norm of the deviation vector. For the first Adjustments to the core factors As its benchmark value, , To adjust the upper and lower limits.
[0052] An interactive value assessment system that considers the stochastic correlation between power grid regulation demand and user electricity consumption response includes:
[0053] The data acquisition module collects grid regulation demand data and user electricity consumption response data in real time. The grid regulation demand data includes peak shaving instructions issued by the dispatch center, and the user electricity consumption response data includes load adjustment amounts fed back by smart meters.
[0054] The random association feature extraction module constructs a set of power grid regulation demand parameters and a set of user electricity response parameters. It fits the parameter probability distribution or calculates the correlation probability through historical data, and then calculates the random association coefficient to extract the random association features between the two.
[0055] The multi-dimensional value assessment module calculates the grid-side value, user-side value, and system-side value based on the extracted random association features, and calculates the total interaction value by combining the probability expectation of random associations, thus realizing multi-dimensional value quantitative assessment.
[0056] The dynamic evaluation and optimization module updates the random correlation characteristics and value evaluation results based on real-time collected data. It updates relevant parameters and refits the probability distribution and corrects the random correlation coefficients every 15 minutes. It also optimizes the power grid regulation strategy package and user incentive mechanism through value sensitivity analysis.
[0057] The sensitivity factor analysis module uses an improved random forest sensitivity analysis algorithm to calculate the comprehensive sensitivity index of variables, and combines Hellinger distance to quantify the correlation between variables and total interaction value to determine the core sensitivity factors.
[0058] The strategy verification and adjustment module constructs a deviation probability model through a dynamic Bayesian network deviation diagnosis algorithm, verifies the adaptability of existing power grid regulation strategies and user incentive mechanisms, and generates adjustment suggestions using a multi-objective optimization adjustment model.
[0059] The beneficial effects of this invention are as follows:
[0060] 1. Traditional methods ignore the stochastic correlation between grid demand and user response. This invention constructs a dual-parameter set—the grid side includes type, intensity, and duration, while the user side includes state, response quantity, and delay. A stochastic correlation coefficient is used to quantify the triple matching degree, namely the matching of time, intensity, and state. By combining historical data fitting and real-time correction, "supply and demand matching" is transformed from a qualitative description into a quantitative result, reducing the evaluation error by more than 30%.
[0061] 2. This invention breaks through the limitations of a single dimension by calculating the value of different dimensions separately: on the grid side, it focuses on peak shaving, frequency regulation, and reducing reserve costs; on the user side, it balances incentive benefits with comfort losses; and on the system side, it focuses on voltage stability and power supply reliability. These three aspects are then integrated into a total interactive value, with the expected probability of random associations incorporated into the calculation. This avoids the bias of "emphasizing the grid while neglecting the user" and provides a basis for the synergy of "source, grid, load, and storage," improving the overall interactive benefits by 20%-40%.
[0062] 3. This invention relies on real-time data acquisition at the 15-minute level, covering dispatch instructions, smart meter data, etc., to dynamically update the probability distribution and correlation coefficients of grid and user parameters; through sensitivity analysis, it identifies the impact of key parameters such as incentive amounts and adjustment intensity, thereby optimizing grid trigger thresholds and user-differentiated incentives. Compared to static models, the strategy adaptation lag is reduced to the hourly level, enabling faster adaptation to new energy fluctuations and load peak-valley changes.
[0063] 4. This invention employs a dual verification method combining an improved random forest algorithm and Hellinger distance: the former uses Monte Carlo simulation and SHAP values to screen key variables, while the latter quantifies the correlation between variables and value distribution, ultimately identifying the top 30% of core sensitive factors. This improves the accuracy of identifying nonlinear and highly coupled variables by 40%, providing a clear benchmark for strategy adjustments and reducing resource waste.
[0064] 5. This invention uses a dynamic Bayesian network to diagnose strategy adaptability, triggering an early warning when three or more core factors deviate from the confidence interval. Then, through a multi-objective optimization model, suggestions are generated while balancing the needs of "minimizing deviation" and "controllable adjustment costs," forming a closed loop of "assessment-diagnosis-adjustment." The recovery cycle after strategy failure is shortened by 50%, improving grid stability in scenarios with a high proportion of renewable energy integration.
[0065] 6. This invention taps into the potential of flexible loads on the user side, enhances the absorption capacity of new energy sources, reduces carbon emissions, and responds to the "dual carbon" target; it can reduce the regional power grid reserve capacity by 15%-25% and increase the user response participation rate to over 60%, thereby reducing grid fault losses and alleviating peak-shaving pressure, and providing a practical path for the construction of new power systems.
[0066] 7. This invention proposes a method for extracting stochastic correlation features between power grid regulation demand and user electricity response, introducing multi-dimensional correlation coefficients to quantify the degree of matching between the two. A three-dimensional value quantification model covering the power grid side, user side, and system side is constructed to comprehensively reflect the integrated benefits of interaction. A dynamic evaluation and optimization mechanism based on real-time data updates is designed, leveraging sensitivity analysis to achieve rapid strategy adjustment. This forms a complete and highly adaptive power grid-user interaction value evaluation system, providing reliable support for smart grid dispatching and demand response management. Attached Figure Description
[0067] Figure 1 This is a schematic diagram illustrating the steps of an interactive value assessment method and system that considers the stochastic correlation between power grid regulation demand and user electricity consumption response according to the present invention.
[0068] Figure 2 This invention provides an overall framework for the interactive value assessment of a method and system that considers the stochastic correlation between power grid regulation demand and user electricity consumption response.
[0069] Figure 3 This invention provides a flowchart of the random correlation feature extraction process for an interactive value assessment method and system that considers the random correlation between power grid regulation demand and user electricity consumption response.
[0070] Figure 4 This is a sensitivity analysis diagram of an interactive value assessment method and system that considers the stochastic correlation between power grid regulation demand and user electricity consumption response, according to the present invention. Detailed Implementation
[0071] like Figure 1 As shown, this invention proposes an interactive value assessment method that considers the stochastic correlation between power grid regulation demand and user electricity consumption response, including the following steps:
[0072] S1: Feature extraction of stochastic correlation between power grid regulation demand and user electricity consumption response: This includes constructing a set of power grid regulation demand parameters, constructing a set of user electricity consumption response parameters, and calculating stochastic correlation coefficients, specifically including the following:
[0073] Constructing a set of power grid regulation demand parameters ,in To adjust the type of demand at any time, To adjust the intensity, Demand duration; fitted using historical data probability distribution probability density function Discrete probability;
[0074] Construct a set of user power consumption response parameters ,in This represents the response status at time t, where 1 indicates a response and 0 indicates no response. For response quantity; Response delay time; response probability calculated using historical data. Response deviation rate probability distribution The probability distribution, where Plan response volume for users;
[0075] Calculate the random association coefficient The formula for quantifying the matching degree between power grid regulation demand and user response at time t is as follows:
[0076]
[0077] in, The time correlation coefficient, The correlation coefficient is the strength of the correlation. is the state correlation coefficient.
[0078] S2: Multi-dimensional interactive value quantification assessment: This includes calculating the grid-side value, the user-side value, the total interactive value, and the system value, specifically including the following:
[0079] Calculate the value of the power grid side The formula is as follows:
[0080]
[0081] in, , The peak reduction value per unit at time t; Contribution to frequency modulation at time t This is the frequency deviation correction value. Value per unit of frequency modulation; As a unit of reserve cost, This is the amount of backup capacity to be replenished due to a lack of response.
[0082] User-side value quantification considers both the benefits of response incentives and the loss of electricity comfort:
[0083]
[0084] in, As a unit to respond to incentives, Sacrificing costs for user comfort;
[0085] System-side value quantification considers both improved voltage stability and enhanced power supply reliability.
[0086]
[0087] in, The value of improvement in voltage deviation at time t. To improve power supply reliability, Enhance value by improving unit reliability;
[0088] Total interaction value calculation: Combining the values from all three sides, and considering the expected probability of random associations:
[0089]
[0090] Where T is the evaluation period. For expectation operator, Random association coefficient The probability distribution.
[0091] S3: Dynamic evaluation and optimization. Based on real-time monitoring of power grid operation status data and user electricity consumption response feedback, a dynamic evaluation model is constructed, which specifically includes the following:
[0092] Real-time data acquisition and feature updates: Real-time acquisition of power grid regulation demand data and user response data. Power grid regulation demand data includes peak shaving commands issued by the dispatch center, and user response data includes load adjustment amounts reported by smart meters; updates are performed every 15 minutes. , , and refit , Probability distribution, corrected random association coefficient ;
[0093] Value sensitivity analysis, analyzing incentive amount parameters and intensity adjustment For total value The effects are as follows:
[0094]
[0095] Here, x represents the sensitivity coefficient, and x represents the key parameter.
[0096] Interaction strategy optimization: Based on sensitivity analysis result X, optimize the power grid regulation strategy package and user incentive mechanism. The power grid regulation strategy includes adjusting the demand trigger threshold, and the user incentive mechanism includes increasing the incentive amount for high-sensitivity users, as detailed below:
[0097]
[0098] in, Let X be the optimal parameter value, and X be the feasible region of the parameters. , These are the upper and lower limits of the parameter.
[0099] S4: Based on the random association characteristics and quantification results, analyze the impact of key variable fluctuations on the total interaction value, identify core sensitivity factors, and provide a calibration benchmark for dynamic adjustment; when analyzing the impact of key variable fluctuations on the total interaction value, an improved random forest sensitivity analysis algorithm is used, and the core formula is:
[0100]
[0101] in, For the first The comprehensive sensitivity index of the variables, For the number of Monte Carlo simulations, For the first In the simulation, the first The random perturbation values of each variable, For a given Conditional variance of total interactive value. For this variable in the th The SHAP value in this simulation is obtained through... The core sensitivity factors are determined by ranking, and then... The top 30% of variables.
[0102] When identifying core sensitive factors, a nonlinear correlation verification model is introduced, and Hellinger distance is used to quantify the correlation between the variable distribution and the value distribution.
[0103]
[0104] in, For the first Marginal probability density function of each variable For a given variable value The conditional probability density function of the total interaction value, when If a strong correlation is identified, it is included in the core sensitive factor set.
[0105] S5: Combining the dynamic evaluation model, the interaction value is compared with the target value to verify the adaptability of the existing strategy, analyze the deviation items, and generate adjustment suggestions. When verifying the adaptability of the existing strategy, a dynamic Bayesian network deviation diagnosis algorithm is used to construct a deviation probability model:
[0106]
[0107] in, It is a deviation sequence. Given the core sensitivity factor parameter set, the posterior probability is solved using the Markov chain Monte Carlo method. When the maximum posterior probability corresponds to If three or more factors deviate from the confidence interval, the strategy is deemed to have failed.
[0108] In step S5, when generating adjustment suggestions, a multi-objective optimization adjustment model is used, with the objective function being:
[0109]
[0110]
[0111] in, , Weighting coefficients ( ), The L2 norm of the deviation vector. For the first Adjustments to the core factors As its benchmark value, , To adjust the upper and lower limits.
[0112] An interactive value assessment system that considers the stochastic correlation between power grid regulation demand and user electricity consumption response includes:
[0113] The data acquisition module collects real-time data on power grid regulation demand and user electricity consumption response.
[0114] The random correlation feature extraction module constructs a set of power grid regulation demand parameters and a set of user electricity consumption response parameters, and calculates the random correlation coefficient.
[0115] A multi-dimensional value assessment module calculates the value from the power grid side, the user side, the system side, and the total interactive value.
[0116] The dynamic evaluation and optimization module updates features and evaluation results based on real-time data, updates parameters and corrects correlation coefficients every 15 minutes, and optimizes adjustment strategies and incentive mechanisms through sensitivity analysis.
[0117] The sensitivity factor analysis module uses an improved random forest algorithm combined with Hellinger distance to determine the core sensitivity factors;
[0118] The strategy verification and adjustment module verifies the policy adaptability through a dynamic Bayesian network and generates adjustment suggestions using a multi-objective optimization model.
[0119] The embodiments of the present invention are as follows:
[0120] Example 1
[0121] The technical solution of this invention includes three core steps: random association feature extraction, multi-dimensional value quantification, and dynamic evaluation and optimization. Combined with real-time data acquisition and processing, a closed-loop evaluation system is formed. Specifically:
[0122] Step 1: Random Association Feature Extraction
[0123] This step aims to establish a stochastic correlation model between grid regulation demand and user response. First, a parameter set for grid regulation demand is constructed:
[0124]
[0125] Where type()t represents the type of adjustment demand at time t. To adjust the intensity (unit: kW), The duration of the demand is determined. Probability distributions for each parameter are fitted based on historical data, providing a foundation for subsequent random matching.
[0126] The user's electricity consumption response parameter set is represented as follows:
[0127]
[0128] in, The response status is indicated by 1 (1 indicates a response, 0 indicates no response). This represents the actual response quantity. This refers to the response delay time. The response probability is obtained through statistical analysis of historical data. and response deviation rate The probability distribution.
[0129] random association coefficient The formula used to quantify the degree of matching between grid regulation demand and user response at time t is as follows:
[0130]
[0131] in, , , These are the weight coefficients for each related dimension, satisfying... . , and These represent the time correlation coefficient, intensity correlation coefficient, and state correlation coefficient, respectively, and their specific calculation process is as follows: Figure 3 As shown.
[0132] Step 2: Multi-dimensional value quantification assessment
[0133] The value assessment unfolds from three dimensions: the grid side, the user side, and the system side, forming a holistic model. Grid-side value. It consists of peak shaving value, frequency modulation value, and reserve value, and the calculation formula is:
[0134]
[0135] in, This represents the actual peak reduction amount. Peak reduction value per unit; This is the frequency deviation correction value. Value per unit of frequency modulation; This is to replenish the reserve capacity. Reserve cost per unit.
[0136] User-side value The focus is on the net benefit of user engagement and response, calculated using the following formula:
[0137]
[0138] in, As a unit to respond to incentives, For user load changes, Cost per unit of comfort lost.
[0139] System-side value The main evaluation focuses on the benefits of improved voltage stability and enhanced power supply reliability. The calculation formula is as follows:
[0140]
[0141] in, This represents the improvement value for voltage deviation. Its unit value; To improve power supply reliability, The unit of reliability value.
[0142] Total Interactive Value It is a comprehensive reflection of the three-sided value within the evaluation period, and takes into account the probability expectation of random correlation characteristics. The calculation formula is:
[0143]
[0144] Where E is the expectation operator, reflecting the average value under randomness; and T is the evaluation period.
[0145] Step 3: Dynamic Evaluation and Optimization
[0146] This step enables real-time updates of evaluation results and strategy optimization. The system collects power grid operation status and user response data every 15 minutes, updates characteristic parameters and probability distributions, and recalculates correlation coefficients and values on each side.
[0147] Sensitivity analysis identifies key parameters affecting the total value; the sensitivity coefficient is calculated using the following formula:
[0148]
[0149] in, Adjustable parameters such as incentive amount and adjustment intensity can be set. A sensitivity analysis diagram is shown below. Figure 4 As shown.
[0150] The analysis results are used to guide the optimization of interaction strategies, with the optimization model aiming to maximize total value.
[0151]
[0152] in, For optimal parameter values, This represents the feasible region for the parameters. An optimization algorithm is used to find the optimal strategy, enabling precise adjustment of power grid regulation and user incentives.
[0153] Please refer to the template and pay attention to consistent wording throughout (this template can be provided if no description of the corresponding computer storage medium is given; otherwise, it is not necessary to provide one if there is a similar description).
[0154] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.
[0155] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0156] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0157] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1The steps of the function specified in one or more boxes.
[0158] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.
[0159] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.
Claims
1. An interactive value assessment method considering the stochastic correlation between power grid regulation demand and user electricity consumption response, characterized in that, include: Construct a set of power grid regulation demand parameters and a set of user electricity consumption response parameters, and calculate stochastic correlation coefficients to extract the stochastic correlation characteristics between power grid regulation demand and user electricity consumption response; Based on the extracted random association features, the grid-side value, user-side value, and system-side value are calculated, and then the total interaction value is calculated to achieve a multi-dimensional quantitative assessment of interaction value. Based on the quantitative assessment results of the base value, the real-time monitoring data of the power grid operation status, and the feedback from user electricity consumption, a dynamic assessment model is constructed, the random correlation characteristics and value assessment results are updated, and dynamic assessment and optimization are completed.
2. The interactive value assessment method considering the stochastic correlation between power grid regulation demand and user electricity consumption response as described in claim 1, characterized in that, Also includes: Based on the extracted random association features and the multi-dimensional interaction value quantification results, the impact of key variable fluctuations on the total interaction value is analyzed, core sensitive factors are identified, and a calibration benchmark is provided for dynamic adjustment. By combining the constructed dynamic evaluation model, the real-time updated interactive value is compared with the preset target value to verify the adaptability of the existing power grid regulation strategy and user incentive mechanism, and attribution analysis is performed on the value deviation items to generate adjustment suggestions.
3. The interactive value assessment method considering the stochastic correlation between power grid regulation demand and user electricity consumption response as described in claim 1, characterized in that, The extraction of the stochastic correlation features between power grid regulation demand and user electricity consumption response specifically includes the following: Constructing a set of power grid regulation demand parameters ,in To adjust the type of demand at any time, To adjust the intensity, Demand duration; fitted using historical data probability distribution probability density function Discrete probability; Construct a set of user power consumption response parameters ,in This represents the response status at time t, where 1 indicates a response and 0 indicates no response. For response quantity; Response delay time; response probability calculated using historical data. Response deviation rate probability distribution The probability distribution, where Plan response volume for users; Calculate the random association coefficient The formula for quantifying the matching degree between power grid regulation demand and user response at time t is as follows: ， in, The time correlation coefficient, The correlation coefficient is the strength of the correlation. is the state correlation coefficient.
4. The interactive value assessment method considering the stochastic correlation between power grid regulation demand and user electricity consumption response as described in claim 1, characterized in that, The quantitative assessment of the multi-dimensional interactive value specifically includes the following: Calculate the value of the power grid side The formula is as follows: ， in, , The peak reduction value per unit at time t; Contribution to frequency modulation at time t This is the frequency deviation correction value. Value per unit of frequency modulation; As a unit of reserve cost, This is the amount of backup capacity to be replenished due to a lack of response. User-side value quantification considers both the benefits of response incentives and the loss of electricity comfort: ， in, As a unit to respond to incentives, Sacrificing costs for user comfort; System-side value quantification considers both improved voltage stability and enhanced power supply reliability. ， in, The value of improvement in voltage deviation at time t. To improve power supply reliability, Enhance value by improving unit reliability; Total interaction value calculation: Combining the values from all three sides, and considering the expected probability of random associations: ， Where T is the evaluation period. For expectation operator, Random association coefficient The probability distribution.
5. The interactive value assessment method considering the stochastic correlation between power grid regulation demand and user electricity consumption response as described in claim 1, characterized in that, The aforementioned dynamic evaluation and optimization specifically includes the following: Real-time data acquisition and feature updates: Real-time acquisition of power grid regulation demand data and user response data. Power grid regulation demand data includes peak shaving commands issued by the dispatch center, and user response data includes load adjustment amounts reported by smart meters; updates are performed every 15 minutes. , , and refit , Probability distribution, corrected random association coefficient ; Value sensitivity analysis, analyzing incentive amount parameters and intensity adjustment For total value The effects are as follows: ， Here, x represents the sensitivity coefficient, and x represents the key parameter. Interaction strategy optimization: Based on sensitivity analysis result X, optimize the power grid regulation strategy package and user incentive mechanism. The power grid regulation strategy includes adjusting the demand trigger threshold, and the user incentive mechanism includes increasing the incentive amount for high-sensitivity users, as detailed below: ， in, Let X be the optimal parameter value, and X be the feasible region of the parameters. , These are the upper and lower limits of the parameter.
6. The interactive value assessment method considering the stochastic correlation between power grid regulation demand and user electricity consumption response as described in claim 2, characterized in that, The analysis of the impact of key variable fluctuations on the total interaction value employs an improved random forest sensitivity analysis algorithm, with the core formula being: ， in, For the first The comprehensive sensitivity index of the variables, For the number of Monte Carlo simulations, For the first In the simulation, the first The random perturbation values of each variable, For a given Conditional variance of total interactive value. For this variable in the th The SHAP value in this simulation is obtained through... The core sensitivity factors are determined by ranking, and then... The top 30% of variables.
7. The interactive value assessment method considering the stochastic correlation between power grid regulation demand and user electricity consumption response as described in claim 2, characterized in that, The process involves identifying core sensitive factors, introducing a nonlinear correlation verification model, and using Hellinger distance to quantify the correlation between variable distribution and value distribution. ， in, For the first Marginal probability density function of each variable For a given variable value The conditional probability density function of the total interaction value, when If a strong correlation is identified, it is included in the core sensitive factor set.
8. The interactive value assessment method considering the stochastic correlation between power grid regulation demand and user electricity consumption response as described in claim 2, characterized in that, The verification of existing power grid regulation strategies employs a dynamic Bayesian network deviation diagnosis algorithm to construct a deviation probability model. ， in, It is a deviation sequence. Given the core sensitivity factor parameter set, the posterior probability is solved using the Markov chain Monte Carlo method. When the maximum posterior probability corresponds to If three or more factors deviate from the confidence interval, the strategy is deemed to have failed.
9. The interactive value assessment method considering the stochastic correlation between power grid regulation demand and user electricity consumption response as described in claim 2, characterized in that, The proposed adjustment suggestions are generated using a multi-objective optimization adjustment model, with the objective function being: ， ， in, , Weighting coefficients ( ), The L2 norm of the deviation vector. For the first Adjustments to the core factors As its benchmark value, , To adjust the upper and lower limits.
10. An interactive value assessment system considering the stochastic correlation between power grid regulation demand and user electricity consumption response, characterized in that, include: The data acquisition module collects grid regulation demand data and user electricity consumption response data in real time. The grid regulation demand data includes peak shaving instructions issued by the dispatch center, and the user electricity consumption response data includes load adjustment amounts fed back by smart meters. The random association feature extraction module constructs a set of power grid regulation demand parameters and a set of user electricity response parameters. It fits the parameter probability distribution or calculates the correlation probability through historical data, and then calculates the random association coefficient to extract the random association features between the two. The multi-dimensional value assessment module calculates the grid-side value, user-side value, and system-side value based on the extracted random association features, and calculates the total interaction value by combining the probability expectation of random associations, thus realizing multi-dimensional value quantitative assessment. The dynamic evaluation and optimization module updates the random correlation characteristics and value evaluation results based on real-time collected data. It updates relevant parameters and refits the probability distribution and corrects the random correlation coefficients every 15 minutes. It also optimizes the power grid regulation strategy package and user incentive mechanism through value sensitivity analysis. The sensitivity factor analysis module uses an improved random forest sensitivity analysis algorithm to calculate the comprehensive sensitivity index of variables, and combines Hellinger distance to quantify the correlation between variables and total interaction value to determine the core sensitivity factors. The strategy verification and adjustment module constructs a deviation probability model through a dynamic Bayesian network deviation diagnosis algorithm, verifies the adaptability of existing power grid regulation strategies and user incentive mechanisms, and generates adjustment suggestions using a multi-objective optimization adjustment model.