Distributed flexible load scheduling control method

Abstract

The invention discloses a distributed flexible load dispatching control method in the field of power dispatching, which comprises the following steps of: step 1, load classification: dividing distributed flexible loads into various flexible loads according to user power utilization characteristics and user response behaviors, the various flexible loads comprise transferable flexible loads, translational flexible loads and reducible flexible loads; and step 2, load response characteristic modeling. The distributed flexible load is classified and responded to scheduling modeling, and the particle swarm optimization algorithm and the artificial bee colony algorithm are utilized to solve the optimal scheduling control strategy, so that efficient and flexible scheduling of the large-scale distributed flexible load can be realized, the dynamic change of a power system can be better adapted, and the scheduling efficiency of the large-scale distributed flexible load is improved. According to the method, the influence of new energy power generation fluctuation on the power grid is effectively relieved, the stability of the power grid is enhanced, the system frequency stability is obviously improved, the scheduling flexibility and response speed are improved, and the communication and control requirements of large-scale distributed flexible loads can be met.

Description

Technical Field

[0001] The present invention relates to the field of power dispatching, and in particular to a distributed flexible load dispatching control method. Background Art

[0002] As the power system transitions to a new energy structure, distributed flexible loads are crucial for improving the capacity to accommodate renewable energy, and their dispatching and control methods have become a research focus. However, current distributed flexible load dispatching and control faces numerous challenges. Renewable energy generation is volatile and uncertain. Large-scale integration of renewable energy leads to significant challenges such as insufficient grid regulation, reduced power quality, and increased distribution network operational risks. Existing grid dispatching and control methods struggle to adapt to the large-scale integration and flexible dispatching demands of distributed energy resources.

[0003] Traditional centralized control methods have limitations when handling large-scale distributed flexible loads. For example, Chinese patent CN109888598A discloses a "distributed power source and flexible load coordinated scheduling method." This method relies on a central controller for optimization calculations. As the number of distributed flexible loads increases, the burden on the central controller increases, and performance bottlenecks emerge. Furthermore, distributed flexible load information privacy and security are difficult to ensure, and the system lacks plug-and-play functionality, making it difficult to expand its application.

[0004] While traditional distributed control methods have alleviated these issues to a certain extent, they still struggle to meet the diverse communication and control requirements of large-scale distributed flexible loads. For example, Chinese patent CN112563777A, "A Coordinated Control Method for Flexible Loads in a Distribution Network Including Distributed Power Sources," has limitations in handling large-scale optimization problems within the context of intelligent scheduling, limiting the advancement of intelligent scheduling and control capabilities for distributed flexible loads.

[0005] Furthermore, with the increasing integration of new energy generation equipment, the need for grid balancing between distributed flexible loads and these new energy generation equipment places higher demands on the response speed of distributed controllers. In new power systems with source-load interaction, traditional distributed control methods need to be improved in terms of quickly and accurately responding to real-time grid changes. Existing distributed flexible load dispatching and control methods face prominent challenges such as communication bottlenecks, insufficient computing power, and delayed response speeds when faced with large-scale, multi-modal loads and the need for rapid response.

[0006] Therefore, those skilled in the art provide a distributed flexible load scheduling control method to solve the problems raised in the above background technology. Summary of the Invention

[0007] The purpose of the present invention is to provide a distributed flexible load dispatching control method to solve the problems raised in the above background technology.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] A distributed flexible load dispatching control method comprises the following steps:

[0010] Step 1: Load classification: Based on user power consumption characteristics and user response behaviors, the distributed flexible loads are divided into various types of flexible loads, including transferable flexible loads, shiftable flexible loads, and curtailable flexible loads.

[0011] Step 2: Load response characteristic modeling: Response characteristics of various flexible loads are modeled. Based on historical data, equipment technical parameters, and user behavior, the transferable power ratio coefficient α, the shiftable power change ratio β, and the curtailable power ratio γ are determined to derive the load regulation potential of various flexible loads at different times.

[0012] Furthermore, for the transferable flexible load, let its transferable power in time period t be The transferable time range is The transferable power can be calculated using the following formula:

[0013]

[0014] in: is the reference power of the load;

[0015] α is the transferable power ratio coefficient obtained based on historical data statistics, and α∈[0,1].

[0016] For a flexible load that can be translated, let its load translation time window be [T shift-start ,T shift-end ], the power before translation is

[0017] The power after translation is The power change before and after translation can be determined by analyzing its production process or equipment operation logic. For example, for some equipment, the power change may be related to the operating speed. Assuming that the speed change ratio is β, then β∈[0,1] or β≥1, depending on the device operating characteristics.

[0018] For the flexible load that can be reduced, let the power ratio that can be reduced in time period t be γt, and the original power is Power can be reduced γt∈[0,1], γt can be determined based on factors such as the energy-saving potential evaluation of the equipment and the degree of comfort reduction acceptable to users. Through the above method, the load regulation potential of various flexible loads in different time periods can be obtained.

[0019] Step 3: Establish a response dispatch model: Based on the load regulation potential and total load power data of various flexible loads in different periods, combined with the new energy power generation data, use the formula P flexible-need =P total -P new-energy -P fixed Establish a multi-type flexible load response scheduling model that meets the constraints of each type of flexible load;

[0020] Step 4: Day-ahead dispatch: Based on the power demand of various flexible loads at different times, professional power forecasting software, big data analysis, and machine learning algorithms are used. In combination with historical electricity consumption data, weather forecasts, business activity schedules, enterprise production plans, and residential electricity consumption behavior, a multi-type flexible load response dispatch model is used to dispatch various flexible loads at different times to obtain day-ahead dispatch data.

[0021] Step 5: Model Solution: Based on the day-ahead dispatch data, the particle swarm optimization algorithm and artificial bee colony algorithm are used to solve the pre-built distributed flexible load multi-objective dispatch optimization model. The model objectives include reducing grid operating costs, increasing the proportion of renewable energy consumption, ensuring grid stability, improving energy utilization efficiency, and enhancing power supply reliability. The optimal dispatch control strategy for distributed flexible loads is obtained.

[0022] Furthermore, based on the day-ahead dispatch data, a particle swarm optimization algorithm and an artificial bee colony algorithm were used to solve a pre-built multi-objective dispatch optimization model for distributed flexible loads, resulting in an optimal dispatch control strategy for distributed flexible loads. This multi-objective dispatch optimization model aims to reduce grid operating costs (C), increase the proportion of renewable energy consumption (R), and ensure grid stability indicators (S).

[0023] Assuming that the grid operation cost includes electricity purchase cost, equipment operation and maintenance cost, etc., the electricity purchase cost can be expressed as

[0024] C buy =t∑p t ×P buy,t , where pt is the electricity price at time period t, P buy,t is the amount of electricity purchased from the grid during period t; the equipment operation and maintenance cost can be modeled based on factors such as equipment type and operating time, such as Cmaintain = j∑k j ×h j , where k j is the operation and maintenance cost coefficient per unit time of the jth type of equipment, h j is the operating time of the jth device, then the grid operating cost C=C buy +Cmaintain. New energy consumption ratio in is the amount of renewable energy power generation effectively consumed in period t.

[0025] The grid stability index S can be determined by measuring factors such as grid voltage fluctuation and frequency deviation. For example, if the voltage fluctuation range is [V min ,V max ], the actual voltage is V, then the voltage stability index can be expressed as:

[0026]

[0027] (V nominal The frequency stability index can be similarly calculated using the frequency deviation. By using the synergistic effect of the particle swarm optimization algorithm and the artificial bee colony algorithm, an optimal scheduling solution is found that minimizes C, maximizes R, and satisfies a certain threshold.

[0028] Step 6: Dispatching and controlling: According to the optimal dispatching and controlling strategy, the distributed flexible loads are dispatched and controlled through the intelligent control system, energy management system, smart home system, etc., the grid parameters and equipment operating status are monitored in real time, and the operating status of the power system is adjusted.

[0029] As a further solution of the present invention: in the step 2, when modeling the response characteristics of various flexible loads, when modeling the response characteristics of various flexible loads, for transferable flexible loads, the transferable power proportion coefficient α is determined by analyzing historical electricity consumption data, user electricity consumption habits and the influence of electricity price policies, and then the power size and time range that can be transferred in different time periods are obtained; for shiftable flexible loads, the power change proportion β is determined based on factors such as equipment operation logic, production process, passenger flow, and the like, and the time window for load shifting and the power change before and after the shift are studied; for reducible flexible loads, the reducible power proportion γ is determined in combination with equipment technical parameters, energy-saving transformation plans, user comfort feedback, etc., and the power proportion and power size that can be reduced in different time periods are determined.

[0030] As a further solution of the present invention: when modeling the response characteristics of various flexible loads in step 2, the adjustment capabilities of various flexible loads, the total load demand of the system, the new energy power generation situation and the constraints of various flexible loads themselves are comprehensively considered to achieve coordinated scheduling of different types of flexible loads.

[0031] As a further solution of the present invention: in step four, the professional power forecasting software, big data analysis, machine learning algorithms, etc. used need to integrate multi-source data, including but not limited to historical power grid operation data, new energy power generation forecast data, user electricity consumption behavior data, meteorological data, production plan data, commercial activity data, etc., to accurately predict load demand and formulate reasonable scheduling plans.

[0032] As a further solution of the present invention: in the distributed flexible load multi-objective scheduling optimization model in step five, reducing the operating cost of the power grid requires considering factors such as electricity purchase cost and equipment operation and maintenance cost to construct a cost model; increasing the proportion of new energy consumption is measured by calculating the ratio of new energy power generation to total power generation; ensuring the stability of the power grid is determined by evaluating indicators such as power grid voltage fluctuation and frequency deviation; improving energy utilization efficiency can be measured by calculating the ratio of the total input energy of the system to the effective output energy; improving power supply reliability is evaluated by statistically evaluating indicators such as power outage time and number of power outages.

[0033] As a further solution of the present invention: in the step 5, when using the particle swarm optimization algorithm and the artificial bee colony algorithm to solve the distributed flexible load multi-objective scheduling optimization model, by setting reasonable algorithm parameters, such as the inertia weight and learning factor of the particle swarm algorithm, the number of leading bees and follower bees of the artificial bee colony algorithm, and determining the appropriate number of iterations, the synergy between the two is achieved and the optimal scheduling solution is searched.

[0034] As a further solution of the present invention: in step six, when dispatching and controlling the distributed flexible load, according to the optimal dispatching scheme obtained by solving, the control instructions are accurately transmitted to the controllers of various flexible load devices through intelligent control systems, energy management systems, smart home systems, etc., to achieve real-time adjustment of the equipment operating status and ensure stable and efficient operation of the power system.

[0035] As a further solution of the present invention: before establishing a multi-type flexible load response scheduling model, it also includes continuously collecting user electricity consumption data, including parameters such as power and electricity consumption time, by installing smart meters, electricity consumption behavior monitoring equipment, industrial-grade electricity consumption monitoring terminals, etc., to provide data support for load classification and characteristic modeling.

[0036] As a further solution of the present invention: after obtaining the optimal scheduling control strategy, it also includes using a risk assessment model to conduct a risk assessment of the strategy from the aspects of power grid security, equipment operation reliability, user satisfaction, etc. If there is a risk, the strategy is adjusted and optimized by adjusting algorithm parameters, optimizing model structure, re-collecting data, etc.

[0037] As a further solution of the present invention: it also includes real-time monitoring of the real-time operation status of the power grid, including parameters such as voltage, frequency, power, and the operating status of various flexible load devices during the dispatching and control process, and dynamically adjusting the optimal dispatching and control strategy using a dynamic adjustment algorithm according to changes in the monitoring data to adapt to real-time changes in the power system.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] The present invention classifies distributed flexible loads and models response scheduling, and uses particle swarm optimization algorithm and artificial bee colony algorithm to solve the optimal scheduling control strategy. It can achieve efficient and flexible scheduling of large-scale distributed flexible loads, better adapt to the dynamic changes of the power system, effectively alleviate the impact of renewable energy power generation volatility on the power grid, enhance power grid stability, significantly improve system frequency stability, and at the same time improve scheduling flexibility and response speed, and can adapt to the communication and control requirements of large-scale distributed flexible loads. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 FIG is a control method diagram of the present invention; DETAILED DESCRIPTION

[0041] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0042] See also Figure 1 In an embodiment of the present invention, a distributed flexible load dispatching control method includes the following steps:

[0043] Step 1: Load classification: Based on user power consumption characteristics and user response behaviors, the distributed flexible loads are divided into various types of flexible loads, including transferable flexible loads, shiftable flexible loads, and curtailable flexible loads.

[0044] Step 2: Load response characteristic modeling: Response characteristics of various flexible loads are modeled. Based on historical data, equipment technical parameters, and user behavior, the transferable power ratio coefficient α, the shiftable power change ratio β, and the curtailable power ratio γ are determined to derive the load regulation potential of various flexible loads at different times.

[0045] For the transferable flexible load, let its transferable power in time period t be The transferable time range is The transferable power can be calculated using the following formula:

[0046]

[0047] in: is the reference power of the load;

[0048] α is the transferable power ratio coefficient obtained based on historical data statistics, and α∈[0,1].

[0049] For a flexible load that can be translated, let its load translation time window be [T shift-start ,T shift-end ], the power before translation is

[0050] The power after translation is The power change before and after translation can be determined by analyzing its production process or equipment operation logic. For example, for some equipment, the power change may be related to the operating speed. Assuming that the speed change ratio is β, then β∈[0,1] or β≥1, depending on the device operating characteristics.

[0051] For the flexible load that can be reduced, let the power ratio that can be reduced in time period t be γt, and the original power is Power can be reduced γt∈[0,1], γt can be determined based on factors such as the energy-saving potential evaluation of the equipment and the degree of comfort reduction acceptable to users. Through the above method, the load regulation potential of various flexible loads in different time periods can be obtained.

[0052] Step 3: Establish a response dispatch model: Based on the load regulation potential and total load power data of various flexible loads in different periods, combined with the new energy power generation data, use the formula P flexible-need =P total -P new-energy -P fixed Establish a multi-type flexible load response scheduling model that meets the constraints of each type of flexible load;

[0053] Step 4: Day-ahead dispatch: Based on the power demand of various flexible loads at different times, professional power forecasting software, big data analysis, and machine learning algorithms are used. In combination with historical electricity consumption data, weather forecasts, business activity schedules, enterprise production plans, and residential electricity consumption behavior, a multi-type flexible load response dispatch model is used to dispatch various flexible loads at different times to obtain day-ahead dispatch data.

[0054] Step 5: Model Solution: Based on the day-ahead dispatch data, the particle swarm optimization algorithm and artificial bee colony algorithm are used to solve the pre-built distributed flexible load multi-objective dispatch optimization model. The model objectives include reducing grid operating costs, increasing the proportion of renewable energy consumption, ensuring grid stability, improving energy utilization efficiency, and enhancing power supply reliability. The optimal dispatch control strategy for distributed flexible loads is obtained.

[0055] Furthermore, based on the day-ahead dispatch data, a particle swarm optimization algorithm and an artificial bee colony algorithm were used to solve a pre-built multi-objective dispatch optimization model for distributed flexible loads, resulting in an optimal dispatch control strategy for distributed flexible loads. This multi-objective dispatch optimization model aims to reduce grid operating costs (C), increase the proportion of renewable energy consumption (R), and ensure grid stability indicators (S).

[0056] Assuming that the grid operation cost includes electricity purchase cost, equipment operation and maintenance cost, etc., the electricity purchase cost can be expressed as

[0057] C buy =t∑p t ×P buy,t , where pt is the electricity price at time period t, P buy,t is the amount of electricity purchased from the grid during period t; the equipment operation and maintenance cost can be modeled based on factors such as equipment type and operating time, such as Cmaintain = j∑k j ×h j , where k j is the operation and maintenance cost coefficient per unit time of the jth type of equipment, h j is the operating time of the jth device, then the grid operating cost C=C buy +Cmaintain. New energy consumption ratio in is the amount of renewable energy power generation effectively consumed in period t.

[0058] The grid stability index S can be determined by measuring factors such as grid voltage fluctuation and frequency deviation. For example, if the voltage fluctuation range is [V min ,V max ], the actual voltage is V, then the voltage stability index can be expressed as:

[0059]

[0060] (V nominal The frequency stability index can be similarly calculated using the frequency deviation. By using the synergistic effect of the particle swarm optimization algorithm and the artificial bee colony algorithm, an optimal scheduling solution is found that minimizes C, maximizes R, and satisfies a certain threshold.

[0061] Step 6: Dispatching and controlling: According to the optimal dispatching and controlling strategy, the distributed flexible loads are dispatched and controlled through the intelligent control system, energy management system, smart home system, etc., the grid parameters and equipment operating status are monitored in real time, and the operating status of the power system is adjusted.

[0062] In this embodiment, in the step 2, when modeling the response characteristics of various flexible loads, for transferable flexible loads, the transferable power ratio coefficient α is determined by analyzing historical electricity consumption data, user electricity consumption habits and the influence of electricity price policies, and then the power size and time range that can be transferred in different time periods are obtained; for shiftable flexible loads, the power change ratio β is determined based on factors such as equipment operation logic, production process, passenger flow, and the time window of its load shift and the power change before and after the shift are studied; for reducible flexible loads, the reducible power ratio γ is determined in combination with equipment technical parameters, energy-saving transformation plans, user comfort feedback, etc., and the power ratio and power size that can be reduced in different time periods are determined.

[0063] In this embodiment, when modeling the response characteristics of various flexible loads in step 2, the adjustment capabilities of various flexible loads, the total load demand of the system, the new energy power generation situation, and the constraints of various flexible loads themselves are comprehensively considered to achieve coordinated scheduling of different types of flexible loads.

[0064] In this embodiment, in step four, the professional power forecasting software, big data analysis, machine learning algorithms, etc. used need to integrate multi-source data, including but not limited to historical power grid operation data, new energy power generation forecast data, user electricity consumption behavior data, meteorological data, production plan data, commercial activity data, etc., to accurately predict load demand and formulate reasonable scheduling plans.

[0065] In this embodiment, in the distributed flexible load multi-objective scheduling optimization model in step five, reducing the operating cost of the power grid requires considering factors such as electricity purchase cost and equipment operation and maintenance cost to construct a cost model; increasing the proportion of new energy consumption is measured by calculating the ratio of new energy power generation to total power generation; ensuring the stability of the power grid is determined by evaluating indicators such as power grid voltage fluctuation and frequency deviation; improving energy utilization efficiency can be measured by calculating the ratio of the total input energy of the system to the effective output energy; improving power supply reliability is evaluated by counting indicators such as power outage time and number of power outages.

[0066] In this embodiment, in step five, when using the particle swarm optimization algorithm and the artificial bee colony algorithm to solve the distributed flexible load multi-objective scheduling optimization model, by setting reasonable algorithm parameters, such as the inertia weight and learning factor of the particle swarm algorithm, the number of leading bees and follower bees of the artificial bee colony algorithm, and determining the appropriate number of iterations, the synergy between the two is achieved and the optimal scheduling solution is searched.

[0067] In this embodiment, when the distributed flexible load is dispatched and controlled in step six, the control instructions are accurately transmitted to the controllers of various flexible load devices through the intelligent control system, energy management system, smart home system, etc. according to the optimal dispatching scheme obtained by solution, so as to realize real-time adjustment of the operating status of the equipment and ensure the stable and efficient operation of the power system.

[0068] In this embodiment, before establishing a multi-type flexible load response scheduling model, it is also included to continuously collect user electricity consumption data, including parameters such as power and electricity consumption time, by installing smart meters, electricity consumption behavior monitoring equipment, industrial-grade electricity consumption monitoring terminals, etc., to provide data support for load classification and characteristic modeling.

[0069] In this embodiment, after obtaining the optimal dispatching and control strategy, a risk assessment model is used to conduct a risk assessment of the strategy from aspects such as power grid security, equipment operation reliability, and user satisfaction. If there is a risk, the strategy is adjusted and optimized by adjusting algorithm parameters, optimizing model structure, and re-collecting data.

[0070] In this embodiment, the real-time operation status of the power grid, including parameters such as voltage, frequency, power, and the operating status of various flexible load devices, is also monitored in real time during the dispatching and control process. According to changes in the monitoring data, the optimal dispatching and control strategy is dynamically adjusted using a dynamic adjustment algorithm to adapt to real-time changes in the power system.

[0071] Example 1

[0072] Step 1. Load classification: In a certain industrial park, cooperate with enterprises to conduct in-depth research on production processes and the layout of electrical equipment. Install industrial-grade electricity monitoring terminals to collect the company's electricity consumption data for the past three months. Distributed flexible load types are divided according to factors such as the degree of impact of production process interruptions on product quality and the flexibility of equipment start-up and shutdown. The electrolytic aluminum equipment of some enterprises has little impact on product quality when adjusting production loads and can be quickly adjusted to a certain extent, so it is classified as a reducible flexible load; the production tasks of automated production lines can be flexibly arranged within a certain time range, which is a transferable flexible load; the energy storage equipment within the enterprise can be charged and discharged according to electricity prices and grid demand, and is determined to be a transferable flexible load.

[0073] Step 2: Load response characteristic modeling:

[0074] Electrolytic aluminum equipment (flexible load reduction possible):

[0075] By analyzing the production records and electricity consumption data of the electrolytic aluminum enterprises in the industrial park during the peak hours of 8:00-12:00 and 18:00-22:00 on weekdays over the past year, and combining the equipment energy-saving transformation plan with the enterprise production plan, we determined the power reduction ratio γ. After evaluation, the value of γ is in the range of 10%-20%. Assuming that the original power of the current electrolytic aluminum equipment P or =5000kW, when γ=0.15, according to the formula P cut =γ×P or , which can reduce power P cut =0.15×5000=750kW.

[0076] Automated production line (flexible load with translation): Study the production process logic of the automated production line and determine that its production time can be translated within the range of ±2 hours. By analyzing historical production data and order delivery time, the power change ratio after translation is set to β = 0.9. The power before translation P is known. pre =800kW, according to the formula P post =β×P pre , power after translation P post =0.9×800=720kW.

[0077] Internal energy storage equipment (transferable flexible load):

[0078] Collect the charging and discharging data of the energy storage device over the past six months to determine its capacity E = 1000kWh and charging power P charge =200kW, discharge power P discharge = 150kW. Based on analysis of electricity pricing policies and corporate electricity demand, charging is scheduled between 11:00 PM and 7:00 AM the following day, and discharging is scheduled between 5:00 PM and 8:00 PM.

[0079] Step 3: Establish a response scheduling model:

[0080] Build an integrated energy management platform for industrial parks, integrating the electricity consumption data of enterprises in the park, new energy generation data and grid access parameters. Assume that the total load power of the system is P total , the total load demand during the peak period is predicted to be P total =10000kW, fixed load power P fixed =4000kW, and the new energy power generation power P in this period is obtained through the new energy power generation prediction model new-energy =1500kW. According to the formula P flexible-need =P total -P new-energy -P fixe d. Calculate the required flexible load regulation power P flexible-need=10,000-1,500-4,000=4,500kW. The model develops a coordinated scheduling plan based on the characteristics of various flexible loads and the production constraints of the enterprise.

[0081] Step 4: Day-ahead scheduling:

[0082] Using big data analysis and machine learning algorithms, combined with factors such as adjustments to the company's production plans and the impact of weather changes on production, the total load demand for the next day was predicted to be Y = 10,500 kW. This includes the power demand of the electrolytic aluminum equipment (Y1 = 5,500 kW), the power demand of the automated production line (Y2 = 1,200 kW), and the power demand of the energy storage equipment (Y3 = 800 kW). Based on a multi-type flexible load response scheduling model, a day-ahead scheduling plan was developed: During peak hours, the load power of the electrolytic aluminum equipment was reduced by 15%, or 0.15 × 5,500 = 825 kW; the production schedule of some automated production lines was shifted to off-peak hours; and the energy storage equipment was scheduled to charge during off-peak hours and discharge during peak hours.

[0083] Step 5: Solve the type:

[0084] In a cluster computing environment, a particle swarm optimization algorithm and an artificial bee colony algorithm were used to solve a multi-objective scheduling optimization model for distributed flexible loads. With the goal of improving the industrial park's energy efficiency and reducing grid operating pressure, a comprehensive cost model was established, including electricity purchase costs and equipment operation and maintenance costs. After 300 iterative calculations, the optimal scheduling control strategy was obtained. The results showed that after optimization, the industrial park's energy efficiency increased by 12% and the grid's peak load decreased by 10%, from 8,000 kW to 7,200 kW.

[0085] Step 6: Scheduling control:

[0086] The next day, the optimized dispatch and control strategy was communicated to each enterprise's electricity management system through the industrial park's energy management system. Following the instructions, each enterprise adjusted the production load of its electrolytic aluminum equipment, the operating hours of its automated production lines, and the charge and discharge status of its energy storage devices. The system monitors grid operating parameters and enterprise electricity usage in real time, providing feedback every 15 minutes. Actual operational results indicate that the industrial park's power system operates stably, with no significant impact on enterprise production. Total electricity consumption in the park that day decreased by 5,000 kWh compared to the pre-optimization period, achieving the goals of energy conservation, emission reduction, and electricity cost reduction.

[0087] Analysis of the data from the above examples demonstrates that the present invention achieves efficient and flexible scheduling of large-scale distributed flexible loads by classifying and modeling distributed flexible load response, and utilizing particle swarm optimization and artificial bee colony algorithms to solve for optimal scheduling control strategies. In actual testing of a regional power grid, the application of the present method increased the scheduling efficiency of distributed flexible loads by 30%, enabling better adaptation to dynamic changes in the power system.

[0088] It also significantly improves the regulation capability and operating efficiency of the power system, effectively alleviates the impact of the volatility of renewable energy generation on the power grid, and enhances the stability of the power grid. In a power grid scenario with a high proportion of distributed renewable energy, after adopting the method of the present invention, the voltage fluctuation range of the power grid is reduced by 20%, the system frequency stability is significantly improved, and the flexibility and response speed of the dispatch are improved, which can adapt to the communication and control requirements of large-scale distributed flexible loads.

[0089] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A distributed flexible load dispatching control method, characterized by: The steps include: Step 1: Load classification: Based on user power consumption characteristics and user response behaviors, the distributed flexible loads are divided into various types of flexible loads, including transferable flexible loads, shiftable flexible loads, and curtailable flexible loads. Step 2: Load response characteristic modeling: Response characteristics of various flexible loads are modeled. Based on historical data, equipment technical parameters, and user behavior, the transferable power ratio coefficient α, the shiftable power change ratio β, and the curtailable power ratio γ are determined to derive the load regulation potential of various flexible loads at different time periods. Step 3: Establish a response dispatch model: Based on the load regulation potential and total load power data of various flexible loads in different periods, combined with the new energy power generation data, use the formula P flexible-need =P total -P new-energy -P fixed Establish a multi-type flexible load response scheduling model that meets the constraints of each type of flexible load; Step 4: Day-ahead dispatch: Based on the power demand of various flexible loads at different times, we utilize professional power forecasting software, big data analysis, and machine learning algorithms, combined with historical electricity consumption data, weather forecasts, business activity schedules, enterprise production plans, and residential electricity consumption behavior. We utilize a multi-type flexible load response dispatch model to perform day-ahead dispatch of various flexible loads at different times, generating day-ahead dispatch data. Step 5: Model Solution: Based on the day-ahead dispatch data, the particle swarm optimization algorithm and artificial bee colony algorithm are used to solve the pre-built multi-objective dispatch optimization model for distributed flexible loads. The model objectives include reducing grid operating costs, increasing the proportion of renewable energy consumption, ensuring grid stability, improving energy utilization efficiency, and enhancing power supply reliability. The optimal dispatch control strategy for distributed flexible loads is obtained. Step 6: Dispatching and controlling: According to the optimal dispatching and controlling strategy, the distributed flexible loads are dispatched and controlled through the intelligent control system, energy management system, and smart home system, and the grid parameters and equipment operating status are monitored in real time to adjust the operating status of the power system.

2. A distributed flexible load dispatching control method according to claim 1, characterized in that: In the step 2, when modeling the response characteristics of various flexible loads, for transferable flexible loads, the transferable power ratio coefficient α is determined by analyzing historical electricity consumption data, user electricity consumption habits and the influence of electricity price policies, and then the power size and time range that can be transferred in different time periods are obtained; for shiftable flexible loads, the power change ratio β is determined based on the equipment operation logic, production process, and passenger flow factors, and the time window for load shifting and the power change before and after the shift are studied; for reducible flexible loads, the reducible power ratio γ is determined in combination with equipment technical parameters, energy-saving transformation plans, and user comfort feedback, and the power ratio and power size that can be reduced in different time periods are determined.

3. A distributed flexible load dispatching control method according to claim 1, characterized in that: In the step 2, when modeling the response characteristics of various flexible loads, the adjustment capabilities of various flexible loads, the total load demand of the system, the renewable energy power generation situation, and the constraints of various flexible loads themselves are comprehensively considered to achieve coordinated scheduling of different types of flexible loads.

4. A distributed flexible load dispatching control method according to claim 1, characterized in that: In step 4, the professional power forecasting software, big data analysis, and machine learning algorithms used must integrate multi-source data, including but not limited to historical power grid operation data, new energy power generation forecast data, user electricity consumption behavior data, meteorological data, production plan data, and commercial activity data, in order to accurately predict load demand and formulate reasonable scheduling plans.

5. A distributed flexible load dispatching control method according to claim 1, characterized in that: In the distributed flexible load multi-objective scheduling optimization model in step 5, reducing the grid operation cost requires considering the power purchase cost and equipment operation and maintenance cost factors to construct a cost model; Improving the proportion of renewable energy consumption can be measured by calculating the ratio of renewable energy power generation to total power generation; ensuring grid stability can be determined by evaluating grid voltage fluctuations and frequency deviation indicators; improving energy efficiency can be measured by calculating the ratio of total system input energy to effective output energy; Improve power supply reliability by evaluating the power outage duration and number of power outages.

6. A distributed flexible load dispatching control method according to claim 1, characterized in that: In step five, when using the particle swarm optimization algorithm and the artificial bee colony algorithm to solve the distributed flexible load multi-objective scheduling optimization model, by setting reasonable algorithm parameters, such as the inertia weight and learning factor of the particle swarm algorithm, the number of leading bees and follower bees of the artificial bee colony algorithm, and determining the appropriate number of iterations, the synergy between the two is achieved and the optimal scheduling solution is searched.

7. A distributed flexible load dispatching control method according to claim 1, characterized in that: In step six, when dispatching and controlling the distributed flexible loads, according to the optimal dispatching scheme obtained, the control instructions are accurately transmitted to the controllers of various flexible load devices through the intelligent control system, energy management system, and smart home system, so as to realize real-time adjustment of the equipment operating status and ensure the stable and efficient operation of the power system.

8. The distributed flexible load dispatching control method according to claim 1, characterized in that: It also includes continuously collecting user electricity consumption data, including power and electricity consumption time parameters, by installing smart meters, electricity consumption behavior monitoring equipment, and industrial-grade electricity consumption monitoring terminals before establishing multi-type flexible load response scheduling models to provide data support for load classification and characteristic modeling.

9. A distributed flexible load dispatching control method according to claim 1, characterized in that: It also includes using a risk assessment model to conduct a risk assessment of the strategy from the aspects of grid security, equipment operation reliability, and user satisfaction after obtaining the optimal dispatching and control strategy. If there is a risk, the strategy will be adjusted and optimized by adjusting algorithm parameters, optimizing model structure, and re-collecting data.

10. A distributed flexible load dispatching control method according to claim 1, characterized in that: It also includes real-time monitoring of the real-time operation of the power grid during the dispatching and control process, including voltage, frequency, power parameters, and the operating status of various flexible load equipment. According to changes in monitoring data, the dynamic adjustment algorithm is used to dynamically adjust the optimal dispatching and control strategy to adapt to real-time changes in the power system.

Images