An intelligent demand response optimization system for power management

Abstract

The application discloses an intelligent demand response optimization system for power management, and particularly relates to the technical field of power dispatching, and constructs a system architecture including a data acquisition and storage module, a power reduction performance analysis module, an elastic window identification module and a power dispatching decision feedback module, dynamically acquires user static attributes, dynamic power load and environmental data, calculates the equivalent power reduction capacity of dispatchable equipment based on an energy balance model, and generates a demand response resource table in combination with user historical power consumption data and business period data, so that the optimization of demand response is realized; by introducing a rebound risk penalty factor and an excessive discharge penalty factor, a dispatching cost correction is applied to high-risk objects, the user callable capacity and the discharge upper limit of the energy storage equipment are dynamically adjusted, and the demand response resource table is updated in real time, so that the problems of the reduction of the demand response operation efficiency and the shortening of the equipment life in the prior art are solved.

Description

Technical Field

[0001] This invention relates to the field of power dispatching technology, and more specifically, to an intelligent demand response optimization system for power management. Background Technology

[0002] In the current operation and management of the power system, the widespread integration of distributed energy resources, the continuous growth of electricity load, and the increasing diversification of electricity consumption patterns have placed higher demands on the flexibility and refined management of power grid dispatch. Traditional demand response models that rely on simple time-segmentation or single-price incentives are unable to fully identify the characteristics of user-end equipment and load shifting potential, resulting in significant power supply pressure during peak hours. To effectively address peak-valley differences and improve the stability of the power system, all parties urgently need more targeted and intelligent demand response solutions.

[0003] Existing technologies in demand response generally suffer from a lack of comprehensive integration of user electricity consumption information and accurate equipment-level analysis. Due to the inability to effectively correlate information from multiple systems, scheduling plans are often only formulated based on averaging or extensive load forecasting, resulting in the underutilization of equipment-level potential.

[0004] Furthermore, as the contradiction between electricity demand and dispatch costs intensifies, traditional methods struggle to quickly match and assess the reduction capacity of each device when faced with complex and ever-changing user operation models, multiple utilization scenarios of dispatchable equipment, and potential load transfer windows, resulting in poor demand response execution. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, this invention provides an intelligent demand response optimization system for power management. By analyzing user electricity consumption data and equipment information, it dynamically identifies dispatchable equipment and flexible windows, generates power peak shaving decisions on demand, and continuously optimizes the system during execution and monitoring to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an intelligent demand response optimization system for power management, comprising: The data acquisition and storage module acquires the user's static attributes, dynamic electricity load data, and corresponding environmental data. The power reduction performance analysis module analyzes static attributes and dynamic power load data to identify dispatchable equipment; based on the performance parameters of the dispatchable equipment, it obtains the available power of the dispatchable equipment for each time period; based on the energy balance model, it converts the available power into equivalent power reduction capacity according to the conversion factor and outputs a list of dispatchable equipment; the available power refers to the amount of power reduction that the dispatchable equipment can provide within a time period. The flexible window identification module analyzes users' historical electricity consumption data and operating period data, identifies the list of schedulable flexible windows, and aligns them with the schedulable equipment catalog by time period to generate a demand response resource table. The demand response resource table records: the user's callable capacity for each time period, the schedulable flexible windows for different users in each time period, and the equivalent power reduction capacity of the schedulable equipment. The power dispatch decision feedback module generates power dispatch instructions based on the demand response resource table in response to the power grid's peak shaving needs, monitors the execution process of the power dispatch instructions, and optimizes the dispatch process based on the execution status of the power dispatch instructions.

[0007] Preferably, the available power of the schedulable device in each time period is obtained as follows: Step 101: Collect the performance parameters of the dispatchable equipment, including at least the rated power, energy storage capacity, and operating efficiency; Step 102: Combine the historical operating data of the schedulable equipment in different time periods, analyze the historical operating data, extract the schedulable capacity, and output the initial available power; Step 103: Analyze environmental data, construct a correlation model between the environment and operating periods, dynamically correct the initial available power, and then quantify the available power. Step 104: Calculate the available power of the schedulable equipment in each discrete time period using the energy balance model.

[0008] Preferably, for dynamic power load data that cannot be directly collected, edge computing and digital twin technologies are used to estimate the equipment, predict operating conditions, and calculate the equivalent power reduction capacity.

[0009] Preferably, the dispatchable equipment includes at least cold storage and photovoltaic equipment; the dispatchable equipment catalog includes dispatchable equipment and equivalent power reduction capacity that varies over time; the equivalent power reduction capacity refers to a quantitative indicator of the amount of electricity that the dispatchable equipment can reduce.

[0010] Preferably, the power dispatching decision feedback module includes a dispatching decision unit, an execution monitoring unit, and a feedback optimization unit; The scheduling decision unit generates several candidate scheduling schemes based on the power grid peak shaving demand and demand response resource table, calculates the scheduling cost index of each selected scheduling scheme, selects the candidate scheme with the lowest scheduling cost index, and formulates the power dispatch instruction. The execution monitoring unit is used to transmit power dispatch instructions to the corresponding dispatchable equipment, monitor the execution status in real time, and collect actual dispatch response data, including the power reduction amount and response time of the dispatchable equipment. The feedback optimization unit analyzes the actual dispatch response data and compares it with the predicted quantitative values ​​of the evaluation indicators, outputting a dispatch quality evaluation index. When the dispatch quality evaluation index is lower than the threshold, it indicates that the power dispatch is abnormal and the dispatch process needs to be optimized.

[0011] Preferably, the scheduling cost index is obtained in the following way: The evaluation indicators of the candidate scheduling schemes are predicted. There are m evaluation indicators, and j represents the sequential number of the evaluation indicators. The weight of the j-th evaluation indicator is denoted as wj; the predicted quantitative value of the j-th evaluation indicator is denoted as ycj, where the quantitative value is the absolute or relative value of the evaluation indicator and is obtained after linear normalization. The scheduling cost index is calculated using the following formula. ; ; Where pj represents the correlation of the j-th evaluation index. If it is positively correlated with the scheduling cost index, the value is pj=1; if it is negatively correlated with the scheduling cost index, the value is pj=-1.

[0012] Preferably, the scheduling quality evaluation index is obtained in the following way: Analyze the actual scheduling response data and denote the actual quantitative value of the j-th evaluation indicator as scj; The predicted and actual quantitative values ​​of the j-th evaluation indicator are combined using the following formula: ; The scheduling quality evaluation index was calculated. .

[0013] Preferably, the power dispatching decision feedback module further includes a short-sighted penalty dispatching decision unit, which includes: introducing a rebound risk penalty factor and an over-discharge penalty factor into the dispatching cost index, and outputting a modified dispatching cost index. Candidate schemes are selected based on the revised scheduling cost index.

[0014] Preferably, the scheduling cost index The method of obtaining it is: Rebound risk penalty factor based on user's historical performance : ; in, Used to normalize the values ​​in parentheses. This represents the difference between the user load and the baseline level after the demand response ends; Indicates the user load baseline level; Indicates the duration of the rebound; Indicates the maximum evaluation time window; α and β represent the adjustment coefficients for each item, used to balance the effects of load increase and duration on the penalty factor; Over-discharge penalty factor is obtained based on the historical performance of schedulable devices. : ; in, Indicates the actual depth of discharge (the current discharge amount as a percentage of the total energy storage capacity). Indicates the safe depth of discharge (the maximum discharge percentage recommended by the manufacturer of the schedulable equipment). This indicates the cumulative time that schedulable equipment has exceeded the safe depth of discharge; This represents the time constant, used to smooth out the effects of the time factor. γ and δ are the adjustment coefficients for each term, used to control the weights of discharge depth and duration on the penalty factor; By jointly analyzing the scheduling cost index, rebound risk penalty factor, and over-discharge penalty factor, the corrected scheduling cost index is output using the following formula. ; ; Reduce the probability that the corresponding candidate solution will be selected.

[0015] Preferably, the system further includes a demand response resource table update module: It is used to receive rebound risk penalty factors and over-discharge penalty factors, dynamically adjust the user's callable capacity and the discharge limit of schedulable equipment based on rebound risk penalty factors and over-discharge penalty factors, and update the demand response resource table in a timely manner.

[0016] The technical effects and advantages of this invention are as follows: (1) The present invention provides an intelligent demand response optimization system for power management. By constructing a system architecture including a data acquisition and storage module, a power reduction performance analysis module, a flexible window identification module and a power dispatch decision feedback module, the system dynamically collects user static attributes, dynamic power load and environmental data, calculates the equivalent power reduction capacity of dispatchable equipment based on the energy balance model, and generates a demand response resource table by combining user historical power consumption data and operating period data. This solves the problems of inaccurate identification of demand response resources and reliance on static parameters in the background technology.

[0017] (2) The present invention provides an intelligent demand response optimization system for power management. By introducing a rebound risk penalty factor and an over-discharge penalty factor, the system applies scheduling cost correction to high-risk objects in the optimization algorithm, dynamically adjusts the user callable capacity and the discharge limit of energy storage equipment, and updates the demand response resource table in real time. This solves the problem in the background technology that user load rebound and over-discharge of energy storage equipment lead to a decrease in system operating efficiency and a shortening of equipment life. Attached Figure Description

[0018] Figure 1 This is a block diagram of the intelligent demand response optimization system of the present invention.

[0019] Figure 2 This is a block diagram of the power dispatching decision feedback module of the present invention.

[0020] Figure 3 This is a flowchart of the demand response resource table update module of the present invention. Detailed Implementation

[0021] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0022] At the same time, it should be understood that, for ease of description, the dimensions of the various parts shown in the accompanying drawings are not drawn according to actual scale.

[0023] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the scope of this application and its application or use.

[0024] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0025] Example 1, see Figure 1 The intelligent demand response optimization system structure diagram of the present invention provides, as shown in the embodiments of the present invention. Figure 1 The intelligent demand response optimization system for power management shown includes: Data acquisition and storage module: Through the intranet or middleware interface, it samples at fixed time intervals (e.g., 15-minute sampling) to obtain users' static attributes (such as historical electricity consumption data, operating period data, equipment performance parameters), dynamic electricity load data, and corresponding environmental data (such as temperature, humidity, and holiday information); and stores the collected data in the database for later use. The data acquisition method involves batch exporting data via SQL / RPA or API. Inputs include intranet user profiles and load data collection. Output is a database for subsequent database construction and analysis. The power reduction performance analysis module analyzes static attributes and dynamic power load data to identify dispatchable equipment. Based on the performance parameters of the dispatchable equipment (such as rated power, energy storage capacity, and operating efficiency), it obtains the available power of the dispatchable equipment for each time period. Based on the energy balance model, it converts the available power into equivalent power reduction capacity according to the conversion factor and outputs a list of dispatchable equipment. The available power refers to the amount of power reduction that the dispatchable equipment can provide within a time period. The explanation states that the dispatchable equipment includes at least cold storage and photovoltaic equipment; the dispatchable equipment catalog includes dispatchable equipment and equivalent power reduction capacity that varies over time; the equivalent power reduction capacity refers to a quantitative indicator of the amount of electricity that the dispatchable equipment can reduce.

[0026] The flexible window identification module analyzes users' historical electricity consumption data and operating period data to identify a list of dispatchable flexible windows (such as the window for transferring air conditioning load in shopping malls at night; a flexible window represents the adjustable range or potential for transfer of user electricity load in different time periods). This is then aligned with the dispatchable equipment catalog by time period to generate a demand response resource table. The demand response resource table records: the user's callable capacity for each time period, the dispatchable flexible windows for different users in each time period, and the equivalent power reduction capacity of dispatchable equipment. The dispatchable flexible window refers to the time window during which electricity load can be transferred when facing grid peak shaving demands. The power dispatch decision feedback module generates power dispatch instructions based on the demand response resource table in response to the power grid's peak shaving needs, monitors the execution process of the power dispatch instructions, and optimizes the dispatch process based on the execution status of the power dispatch instructions.

[0027] In this embodiment of the invention, it needs to be further explained that the method for obtaining the available power of the schedulable device in each time period is as follows: Step 101: Collect the performance parameters of the dispatchable equipment, including at least the rated power, energy storage capacity, and operating efficiency; Explanation: Performance parameters are the basis for determining the maximum power that a device can output or reduce under ideal conditions. For example, the peak power of a photovoltaic device can be used to assess the upper limit of output during periods of highest solar radiation; the capacity and charge / discharge efficiency of an energy storage battery determine its available peak clipping depth and response rate. Step 102: Combine the historical operating data of the schedulable equipment at different time periods, including the actual load curve, start / stop duration, and energy storage battery status; analyze the historical operating data, extract the schedulable capacity, and output the initial available power; Step 103: Analyze environmental data, construct a correlation model between the environment and operating periods, dynamically correct the initial available power, and quantify the available power; for example, dynamically assess the degree of reduction in the peak shaving capacity of the photovoltaic system under cloudy and rainy weather, or the increase in the available capacity of air conditioning equipment when the mall load decreases during holidays; if the photovoltaic radiation level is far below the rated standard, the available power of the photovoltaic system needs to be reduced accordingly. Explanation: The environmental data includes: temperature, humidity, solar radiation intensity, user operating hours, holidays, etc. For equipment that is greatly affected by weather, such as photovoltaic equipment, dynamic corrections are required based on current or predicted environmental data (such as temperature, humidity, solar radiation intensity). For equipment that is closely related to user operating hours, such as shopping mall air conditioners and industrial production equipment, a comprehensive evaluation is required based on current operating schedules, holiday information, etc. Step 104: Calculate the available power of the schedulable equipment in each discrete time period using the energy balance model; The explanation is that the previously revised available power, energy storage status, and external environment are integrated into a unified energy balance equation. Through iterative calculations or rolling forecasting methods, the final callable power reduction capacity value is generated. If, during a certain period, the air conditioning load surges due to dispatchable equipment being discharged to a safe threshold or a sudden rise in external temperature, the reduction allocation to these devices is automatically reduced to ensure that the entire demand response process proceeds smoothly within the limits of safety and comfort.

[0028] In this embodiment of the invention, it is necessary to further explain that for dynamic power load data that cannot be directly collected, edge computing and digital twin technology are used to estimate the equipment, predict the operating conditions, calculate the equivalent power reduction capacity, and improve the flexibility curve.

[0029] Further explanation is needed in the embodiments of the present invention, see the following: Figure 2 The diagram shows the structural block diagram of the power dispatching decision feedback module, which includes a dispatching decision unit, an execution monitoring unit, and a feedback optimization unit. The dispatch decision unit generates several candidate dispatch schemes based on the power grid peak shaving demand and demand response resource table, calculates the dispatch cost index of each selected dispatch scheme, selects the candidate scheme with the lowest dispatch cost index, and formulates a power dispatch instruction; the dispatch scheme includes the user number participating in the dispatch, the dispatchable equipment number, and the power reduction amount preset by the dispatchable equipment; The execution monitoring unit is used to transmit power dispatch instructions to the corresponding dispatchable equipment, monitor the execution status in real time, and collect actual dispatch response data, including the power reduction amount and response time of the dispatchable equipment. The feedback optimization unit analyzes the actual dispatch response data and compares it with the predicted quantitative values ​​of the evaluation indicators, outputting a dispatch quality evaluation index. When the dispatch quality evaluation index is lower than the threshold, it indicates that the power dispatch is abnormal and the dispatch process needs to be optimized.

[0030] In this embodiment of the invention, it needs to be further explained that the plurality of candidate scheduling schemes are generated in the following manner: One possible implementation involves inputting schedulable devices, schedulable elastic windows, and constraints (such as device power limits, user preferences, environmental data, etc.) into integer linear programming or mixed integer linear programming to obtain a series of feasible solutions, with each feasible solution being regarded as a candidate scheduling scheme. In one possible implementation, under multi-objective or multi-constraint scenarios (such as simultaneously considering peak shaving, scheduling cost, user satisfaction, etc.), heuristic search methods such as genetic algorithms and particle swarm optimization are used to iteratively generate and optimize several scheduling schemes; the several excellent solutions generated in different iteration rounds become the set of candidate scheduling schemes. The explanation is that when the power grid needs to reduce power, decisions are made based on the dispatch cost index, prioritizing reduction schemes with lower dispatch cost index values. For example, energy storage batteries are discharged first, and then the power consumption of other equipment is reduced. The aim is to ensure that while meeting the grid's peak shaving needs, dispatch costs are minimized and dispatch efficiency is improved. Optimizing power generation plans and strengthening line monitoring and maintenance: timely detection and handling of line faults improves grid stability.

[0031] The explanation is that by continuously tracking the power reduction of dispatchable equipment, operations are ensured to be carried out according to instructions, thereby effectively regulating the power grid load. During the execution process, if any abnormal situation is found, such as equipment failure or user non-acceptance of dispatch, an alarm mechanism is immediately triggered, and corresponding measures are taken to ensure the smooth operation of power dispatch.

[0032] In this embodiment of the invention, it needs to be further explained that the scheduling cost index is obtained in the following way: The evaluation indicators of the candidate scheduling schemes are predicted. There are m evaluation indicators, and j represents the sequential number of the evaluation indicators. The evaluation indicators include one or more of the following: economic cost indicators (such as electricity purchase cost and equipment maintenance cost), energy conversion efficiency, stability indicators (such as the fluctuation range of grid frequency, the degree and duration of voltage fluctuation), response speed, and the proportion of renewable energy (such as solar and wind power). The weight of the j-th evaluation indicator is denoted as wj, reflecting the importance of the corresponding evaluation indicator in the overall evaluation; the predicted quantitative value of the j-th evaluation indicator is denoted as ycj, where the quantitative value is the absolute value (such as electricity purchase cost) or relative value (such as the ratio of actual value to preset value) of the evaluation indicator, and is obtained after linear normalization. The scheduling cost index is calculated using the following formula. ; ; Where pj represents the correlation of the j-th evaluation index. If it is positively correlated with the scheduling cost index, the value is pj=1; if it is negatively correlated with the scheduling cost index, the value is pj=-1.

[0033] In this embodiment of the invention, it needs to be further explained that the scheduling quality evaluation index is obtained in the following way: Analyze the actual scheduling response data and denote the actual quantitative value of the j-th evaluation indicator as scj; The predicted and actual quantitative values ​​of the j-th evaluation indicator are combined using the following formula: ; The scheduling quality evaluation index was calculated. .

[0034] Example 2: The difference between this embodiment and Example 1 is that the power dispatching decision feedback module further includes a short-sighted penalty dispatching decision unit. Background information: For each candidate scheduling scheme, its scheduling cost index (including economic costs and operating losses) is calculated and incorporated into a multi-objective optimization model for quantitative evaluation. If a scheme can quickly meet the current peak shaving target, but will cause subsequent energy storage or key equipment to be unable to meet the demand in the next period, short-sighted scheduling needs to be avoided. Based on this, a short-sighted penalty scheduling decision unit is provided to achieve a balance between short-term peak shaving effect and long-term system stability effect. By comprehensively evaluating the user rebound risk and the energy storage over-discharge risk, the discharge limit of the user's callable capacity and the scheduling equipment in the next cycle is dynamically updated, so as to make the scheduling decision more reasonable and reliable. The short-sighted penalty scheduling decision unit includes: introducing a rebound risk penalty factor and an over-discharge penalty factor into the scheduling cost index, and outputting a modified scheduling cost index. Candidate scheduling schemes are selected based on the revised scheduling cost index to avoid selecting high-risk scheduling schemes; the high-risk scheduling schemes refer to: High probability of rebound in user electricity load: After the execution of power dispatch instructions, the user's electricity load rebounds and exceeds the level after the expected reduction, which may cause power grid load fluctuations and stability problems. High probability of over-discharge of dispatchable equipment: During the execution of power dispatch instructions, the depth of discharge of dispatchable equipment exceeds its design or safety threshold, which leads to a shortened equipment life or performance degradation, affecting subsequent peak shaving capabilities.

[0035] In this embodiment of the invention, it is necessary to further explain the scheduling cost index. The method of obtaining it is: Rebound risk penalty factor based on user's historical performance : ; in, Used to normalize the values ​​in parentheses. This represents the difference between the user load and the baseline level after the demand response ends; Indicates the user load baseline level; Indicates the duration of the rebound (the total duration for which the load exceeds the baseline level); Indicates the maximum evaluation time window; α and β represent the adjustment coefficients for each item, used to balance the effects of load increase and duration on the penalty factor; Over-discharge penalty factor is obtained based on the historical performance of schedulable devices. : ; in, Indicates the actual depth of discharge (the current discharge amount as a percentage of the total energy storage capacity). Indicates the safe depth of discharge (the maximum discharge percentage recommended by the manufacturer of the schedulable equipment). This indicates the cumulative time that schedulable equipment has exceeded the safe depth of discharge; This represents the time constant, used to smooth out the effects of the time factor. γ and δ are the adjustment coefficients for each term, used to control the weights of discharge depth and duration on the penalty factor; The scheduling cost index is corrected by combining the joint analysis of the scheduling cost index, rebound risk penalty factor, and over-discharge penalty factor, and the corrected scheduling cost index is output. ; In one possible implementation, the corrected scheduling cost index is output using the following formula. ; ; In the multi-objective optimization process, a higher scheduling cost index is imposed on high-risk scheduling schemes to reduce their probability of being selected; by prioritizing low-risk scheduling schemes, the stable operation of the power system and the long-term availability of dispatchable equipment are ensured.

[0036] In this embodiment of the invention, it should be further explained that the system also includes a demand response resource table update module: Receive rebound risk penalty factor and over-discharge penalty factor, dynamically adjust the user's callable capacity and the discharge limit of the schedulable equipment based on the rebound risk penalty factor and over-discharge penalty factor, and update the demand response resource table in a timely manner to support subsequent rolling optimization and scheduling instruction generation; In this embodiment of the invention, it should be further explained that the rebound risk penalty factor is positively correlated with the reduction of the user's callable capacity; the over-discharge penalty factor is positively correlated with the reduction of the upper limit of the discharge of the schedulable device; the reduction range is based on empirical settings, and this invention does not impose specific limitations on it; For ease of understanding, in the embodiments of the present invention: Based on the value of the rebound risk penalty factor, the user's load rebound risk is divided into three levels: low, medium, and high. Based on the value of the over-discharge penalty factor, the over-discharge risk of schedulable equipment is divided into three levels: low, medium, and high. For medium-risk users, their callable capacity will be reduced to 80% of the original value; for high-risk users, their callable capacity will be reduced to 50% of the original value; for low-risk users, their callable capacity will remain unchanged. For medium-risk dispatchable equipment, its discharge limit will be tightened to 90% of the original value. For high-risk dispatchable equipment, its discharge limit is tightened to 70% of the original value; For low-risk, schedulable equipment, maintain its discharge limit unchanged.

[0037] Explanation: The demand response resource table update module dynamically adjusts the data to reduce the negative impact of high-risk users and dispatchable equipment on subsequent scheduling, preventing user load rebound and excessive discharge of dispatchable equipment; ensuring that the data in the demand response resource table reflects the latest system status and risk assessment results, improving the accuracy and efficiency of scheduling decisions; and maintaining sufficient safety margins to reduce uncertainties in system operation and ensure the stability of the power system and the long-term health of equipment.

[0038] See Figure 3 The flowchart of the demand response resource table update module is provided in this embodiment of the invention. The demand response resource table update module includes the following steps: Step 201, Period Division: Divide the power data into several periods according to fixed time intervals, and record the power dispatch execution status of each period; Step 202: Identify risks and generate risk markers: Based on the power dispatch execution status of the previous cycle, collect the load reduction curves generated during the demand response process, as well as the discharge depth information of dispatchable equipment (such as the power consumption ratio from full charge state to the current state); compare the collected data with the baseline operating parameters (such as historical average load, normal discharge depth, etc.). Analyze whether there is a load rebound phenomenon in the backend of user demand response (such as a sudden and significant increase in load). Determine whether the discharge depth of the schedulable equipment exceeds the safety threshold, and then generate a bounce risk marker and an over-discharge marker; Step 203: Based on the rebound risk penalty factor and the over-discharge penalty factor, dynamically adjust the user's callable capacity and the energy storage discharge limit, and write the updated user's callable capacity and discharge limit into the resource table for subsequent rolling scheduling. For example, if a user is marked as having a rebound risk, check whether their available load exceeds the safe range and reduce the user's available capacity for the next cycle; if a schedulable device is marked as having an over-discharge, reduce the discharge limit for the next cycle according to its risk level to prevent the device from overdrawing its subsequent peak shaving capacity. The explanation is as follows: the available load refers to the maximum user load that the scheduling system can reduce or shift within the current cycle; the discharge limit reflects the maximum discharge power or depth that the scheduling equipment can withstand in the next cycle; the purpose of dynamic adjustment is to reserve an appropriate safety margin for high-risk users and equipment, and reduce the risk of subsequent rebound or equipment wear. Step 204, Rolling Optimization and Scheduling Instruction Generation: Under the power grid peak shaving requirements, multi-objective rolling optimization is performed in conjunction with the updated resource table.

[0039] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An intelligent demand response optimization system for power management, characterized in that, include: The data acquisition and storage module acquires the user's static attributes, dynamic electricity load data, and corresponding environmental data. The power reduction performance analysis module analyzes static attributes and dynamic power load data to identify dispatchable equipment. Based on the performance parameters of the schedulable devices, the available power of the schedulable devices in each time period is obtained. Based on the energy balance model, available power is converted into equivalent power reduction capacity according to the conversion factor, and a list of dispatchable equipment is output; the available power refers to the amount of power reduction that dispatchable equipment can provide within a time period. The flexible window identification module analyzes users' historical electricity consumption data and operating period data, identifies the list of schedulable flexible windows, aligns them with the schedulable equipment catalog by time period, and generates a demand response resource table. The demand response resource table records: the user's available capacity for each time period, the schedulable elastic window for different users in each time period, and the equivalent power reduction capacity of schedulable equipment. The power dispatch decision feedback module generates power dispatch instructions based on the demand response resource table in response to the power grid's peak shaving needs, monitors the execution process of the power dispatch instructions, and optimizes the dispatch process based on the execution status of the power dispatch instructions.

2. The intelligent demand response optimization system for power management according to claim 1, characterized in that, The method for obtaining the available power of the schedulable devices in each time period is as follows: Step 101: Collect the performance parameters of the dispatchable equipment, including at least the rated power, energy storage capacity, and operating efficiency; Step 102: Combine the historical operating data of the schedulable equipment in different time periods, analyze the historical operating data, extract the schedulable capacity, and output the initial available power; Step 103: Analyze environmental data, construct a correlation model between the environment and the operating period, dynamically correct the initial available power, and then quantify the available power; Step 104: Calculate the available power of the schedulable equipment in each discrete time period using the energy balance model.

3. The intelligent demand response optimization system for power management according to claim 1, characterized in that, For dynamic electricity load data that cannot be directly collected, edge computing and digital twin technologies are used to estimate equipment, predict operating conditions, and calculate equivalent power reduction capacity.

4. The intelligent demand response optimization system for power management according to claim 1, characterized in that, The dispatchable equipment includes at least cold storage and photovoltaic equipment; the dispatchable equipment catalog includes dispatchable equipment and equivalent power reduction capacity that varies over time; the equivalent power reduction capacity refers to a quantitative indicator of the amount of electricity that the dispatchable equipment can reduce.

5. The intelligent demand response optimization system for power management according to claim 1, characterized in that, The power dispatch decision feedback module includes a dispatch decision unit, an execution monitoring unit, and a feedback optimization unit; The scheduling decision unit generates several candidate scheduling schemes based on the power grid peak shaving demand and demand response resource table, calculates the scheduling cost index of each selected scheduling scheme, selects the candidate scheme with the lowest scheduling cost index, and formulates the power dispatch instruction. The execution monitoring unit is used to transmit power dispatch instructions to the corresponding dispatchable equipment, monitor the execution status in real time, and collect actual dispatch response data, including the power reduction amount and response time of the dispatchable equipment. The feedback optimization unit analyzes the actual dispatch response data and compares it with the predicted quantitative values ​​of the evaluation indicators, outputting a dispatch quality evaluation index. When the dispatch quality evaluation index is lower than the threshold, it indicates that the power dispatch is abnormal and the dispatch process needs to be optimized.

6. The intelligent demand response optimization system for power management according to claim 5, characterized in that, The scheduling cost index is obtained as follows: The evaluation indicators of the candidate scheduling schemes are predicted. There are m evaluation indicators, and j represents the sequential number of the evaluation indicators. The weight of the j-th evaluation indicator is denoted as wj; the predicted quantitative value of the j-th evaluation indicator is denoted as ycj, where the quantitative value is the absolute or relative value of the evaluation indicator and is obtained after linear normalization. The scheduling cost index is calculated using the following formula. ; ； Where pj represents the correlation of the j-th evaluation index. If it is positively correlated with the scheduling cost index, the value is pj=1; if it is negatively correlated with the scheduling cost index, the value is pj=-1.

7. The intelligent demand response optimization system for power management according to claim 6, characterized in that, The scheduling quality assessment index is obtained as follows: Analyze the actual scheduling response data and denote the actual quantitative value of the j-th evaluation indicator as scj; The predicted and actual quantitative values ​​of the j-th evaluation indicator are combined using the following formula: ； The scheduling quality evaluation index was calculated. .

8. The intelligent demand response optimization system for power management according to claim 7, characterized in that, The power dispatching decision feedback module also includes a short-sighted penalty dispatching decision unit, which includes: introducing a rebound risk penalty factor and an over-discharge penalty factor into the dispatching cost index, and outputting a modified dispatching cost index. Candidate schemes are selected based on the revised scheduling cost index.

9. The intelligent demand response optimization system for power management according to claim 8, characterized in that, The scheduling cost index The method of obtaining it is: Rebound risk penalty factor based on user's historical performance : ； in, Used to normalize the values ​​in parentheses. This represents the difference between the user load and the baseline level after the demand response ends; Indicates the user load baseline level; Indicates the duration of the rebound; Indicates the maximum evaluation time window; α and β represent the adjustment coefficients for each item, used to balance the effects of load increase and duration on the penalty factor; Over-discharge penalty factor is obtained based on the historical performance of schedulable devices. : ； in, Indicates the actual depth of discharge; Indicates the safe depth of discharge; This indicates the cumulative time that schedulable equipment has exceeded the safe depth of discharge; This represents the time constant, used to smooth out the effects of the time factor. γ and δ are the adjustment coefficients for each term, used to control the weights of discharge depth and duration on the penalty factor; By jointly analyzing the scheduling cost index, rebound risk penalty factor, and over-discharge penalty factor, the corrected scheduling cost index is output using the following formula. ; ； Reduce the probability that the corresponding candidate solution will be selected.

10. The intelligent demand response optimization system for power management according to claim 9, characterized in that, It also includes a demand response resource table update module: It is used to receive rebound risk penalty factors and over-discharge penalty factors, dynamically adjust the user's callable capacity and the discharge limit of schedulable equipment based on rebound risk penalty factors and over-discharge penalty factors, and update the demand response resource table in a timely manner.

Images