Multi-dimensional cost-benefit balanced switching control method, system and device for high-proportion new energy access power grid and storage medium

Abstract

The present application relates to a kind of high proportion new energy access grid's multi-dimensional cost benefit balance switching control method, system, equipment and storage medium, the method includes: obtaining the operation data of new energy station, energy storage device etc.;According to multiple time scales, new energy output and load curve are handled in layers, and deviation quantity is calculated etc.;Energy storage marginal cost, load regulation marginal cost, new energy marginal consumption benefit and grid operation marginal benefit model are constructed;Between energy storage regulation, load reduction or peak shifting and wind and light abandonment, dynamic switching is carried out;When there is risk of exceeding limit, according to the contribution of new energy station to limit exceeding and marginal consumption benefit, differentiated limit distribution is executed;The switching strategy and limit distribution result are corrected with security constraint, obtain the control action that satisfies power balance, energy storage boundary, load boundary and section safety constraint and execute, then according to the updated data, it is repeated rolling. The present application realizes the dynamic balance between new energy consumption benefit and system comprehensive regulation cost.

Description

Technical Field

[0001] This invention relates to the field of new power system dispatching and control technology, and in particular to a multi-dimensional cost-benefit balance switching control method, system, equipment and storage medium for high-proportion renewable energy grid access. Background Technology

[0002] With the continuous increase in installed capacity of new energy sources such as wind power and photovoltaics, the proportion of new energy output in regional power grids, local power grids, and distribution networks is constantly rising. Due to the randomness, volatility, and intermittency of new energy sources, when a high proportion is integrated into the grid, the system is prone to problems such as increased power balance deviations, frequent calls to regulation resources, increased risk of critical sections exceeding limits, and increased overall system operating costs. To address these challenges, existing technologies typically prioritize energy storage to absorb the fluctuations in new energy output. However, in actual operation, energy storage not only incurs costs related to power call-up and capacity occupancy, but also costs related to lifespan reduction, charging and discharging losses, and additional costs associated with rapid ramp-up. During certain high-load conditions, high-frequency switching, or rapid ramp-up periods, the marginal cost of continuing to utilize energy storage may exceed the marginal benefit of continuing to absorb new energy, indicating that relying solely on energy storage for regulation is not always economically optimal.

[0003] Meanwhile, adjustable load resources such as interruptible industrial loads, flexible commercial loads, and electric vehicle charging loads typically possess the ability to reduce, transfer, postpone, or advance loads, and can, to some extent, replace high-cost energy storage regulation. However, different load regulation resources differ significantly in terms of economics, duration, recovery characteristics, and user compensation mechanisms. Existing technologies generally lack a mechanism for unified quantitative comparison of different regulation resources and dynamic switching accordingly. Furthermore, in scenarios where transmission bottlenecks or critical sections are restricted, even if renewable energy can continue to be absorbed from the perspective of overall grid power balance, local lines or sections may still experience power flow exceeding limits due to concentrated renewable energy transmission. In such cases, relying solely on energy storage regulation or implementing renewable energy generation restrictions according to a uniform ratio may not only lead to excessively high regulation costs but also result in unreasonable selection of restriction targets, affecting the economics and fairness of system operation.

[0004] In summary, existing technologies have at least the following problems: insufficient utilization of the multi-timescale fluctuation characteristics of new energy sources and loads; incomplete assessment of the actual marginal cost of energy storage; lack of a dynamic switching mechanism between energy storage regulation, load regulation, and wind and solar curtailment; lack of a differentiated power limiting mechanism for new energy sites under cross-sectional constraints; and lack of a unified safety constraint correction mechanism before the execution of control actions. Summary of the Invention

[0005] To address the aforementioned shortcomings, this invention proposes a multi-dimensional cost-benefit equilibrium switching control method, system, equipment, and storage medium for high-proportion renewable energy grid integration. This invention constructs a multi-dimensional marginal cost and marginal benefit model, dynamically switching between energy storage regulation, load reduction or peak shifting, and wind / solar curtailment. It also incorporates the contribution of transmission bottlenecks or cross-section over-limits to implement differentiated power generation restriction allocation for renewable energy stations. Finally, through safety constraint correction, it obtains control actions that satisfy the grid operation boundary, thereby achieving a balanced control between renewable energy consumption benefits and overall system regulation costs.

[0006] The technical solution adopted in this invention is a multi-dimensional cost-benefit equilibrium switching control method for high-proportion renewable energy grid access, comprising the following steps: Step 1: Obtain operational data for new energy stations, energy storage devices, adjustable loads, and power transmission bottlenecks / sections; Step 2: Perform layered processing on the new energy output curve and load curve according to multiple time scales, calculate the deviation between new energy output and load demand, as well as the corresponding regulation power demand, regulation energy demand and regulation rate demand. Step 3: Construct models for the marginal cost of energy storage, the marginal cost of load regulation, the marginal benefit of renewable energy consumption, and the marginal benefit of grid operation; Step 4: Based on the comparison results of the marginal cost of energy storage, the marginal cost of load regulation, the marginal benefit of renewable energy consumption, and the marginal benefit of grid operation, determine the resource switching strategy among energy storage regulation, load reduction or peak shifting, and wind and solar curtailment. Step 5: When there is a risk of exceeding the limit at a power transmission bottleneck or section, differentiated power generation restriction allocation is implemented for multiple new energy stations based on the contribution of the new energy station to the limit exceedance and the corresponding marginal consumption revenue. Step 6: Perform safety constraint correction on the resource switching strategy and differentiated power generation limitation allocation results to obtain control actions that meet power balance constraints, energy storage operation boundary constraints, load regulation boundary constraints, and transmission bottleneck or section safety constraints. Step 7: Execute the control action for the current scheduling cycle, and repeat steps 1 to 7 on a rolling basis according to the updated running data.

[0007] Preferably, the multiple time scales in step 2 include intraday scales and cross-day scales, and the cross-day scale includes at least one of weekly scale, monthly scale, quarterly scale or annual scale.

[0008] Preferably, the marginal cost of energy storage in step 3 consists of at least two of the following: energy storage power call-up cost, energy storage capacity occupancy cost, energy storage cycle life depreciation cost, energy storage charge / discharge loss cost, and energy storage ramp-up capability cost; and / or The marginal cost of load adjustment in step 3 includes at least the cost of load reduction compensation and the cost of load peak shifting compensation; and / or The marginal benefit of renewable energy consumption in step 3 includes at least one of the following: renewable energy grid connection electricity revenue, green environmental attribute revenue, and carbon emission reduction revenue; and / or The marginal benefits of power grid operation in step 3 include at least one of the following: peak shaving and valley filling benefits, reserve capacity saving benefits, congestion mitigation benefits, and delayed capacity expansion benefits.

[0009] Preferably, the resource switching strategy in step 4 includes: When the marginal cost of load regulation is lower than the marginal cost of energy storage, and the adjustable load capacity meets the current regulation needs, load reduction or peak shifting should be prioritized; and / or When both the marginal cost of energy storage and the marginal cost of load regulation exceed the sum of the marginal revenue from renewable energy integration and the marginal revenue from grid operation, wind and solar curtailment control is triggered; and / or When the marginal cost of energy storage is not higher than the sum of the marginal revenue from renewable energy consumption and the marginal revenue from grid operation, energy storage regulation should be given priority.

[0010] Preferably, the contribution of the renewable energy station to the over-limit situation in step 5 is determined jointly by the power output of the renewable energy station and the corresponding power flow distribution factor; and / or The differentiated emission restriction allocation in step 5 is determined by jointly ranking the contribution of new energy stations to exceeding the limit, the current ramp intensity, and the marginal consumption benefit.

[0011] Preferably, the safety constraint correction in step 6 is achieved by constructing an optimization problem with the goal of minimizing the deviation between the corrected control action and the original action. The constraints of the optimization problem include at least power balance constraints, energy storage state boundary constraints, energy storage power boundary constraints, energy storage ramp rate constraints, load regulation boundary constraints, and power flow safety constraints at transmission bottlenecks / sections.

[0012] Preferably, at least one of steps 4 to 6 is executed by an intelligent decision-making model trained based on historical operational data; the intelligent decision-making model includes one or more of the following: constrained reinforcement learning model, hierarchical reinforcement learning model, multi-agent reinforcement learning model, rolling optimization model, or rule optimization model.

[0013] This invention also discloses a multi-dimensional cost-benefit balance switching control system for high-proportion renewable energy grid access, comprising: The data acquisition module is used to acquire operational data from new energy stations, energy storage devices, adjustable loads, and power transmission checkpoints or sections. The multi-timescale analysis module is used to perform layered processing of the new energy output curve and load curve according to multiple timescales, calculate the deviation between new energy output and load demand, as well as the corresponding regulation power demand, regulation energy demand and regulation rate demand. The cost-benefit modeling module is used to construct models for the marginal cost of energy storage, the marginal cost of load regulation, the marginal benefit of renewable energy consumption, and the marginal benefit of power grid operation. The resource switching decision module is used to determine the resource switching strategy among energy storage regulation, load reduction or peak shifting, and wind and solar curtailment based on the comparison results of the marginal cost of energy storage, the marginal cost of load regulation, the marginal benefit of renewable energy consumption, and the marginal benefit of grid operation. The differentiated power generation restriction allocation module is used to implement differentiated power generation restriction allocation for multiple renewable energy stations based on the contribution of renewable energy stations to the over-limit and the corresponding marginal consumption revenue when there is a risk of exceeding the limit at a power transmission bottleneck or section. The safety constraint correction module is used to perform safety constraint correction on the resource switching strategy and the differentiated power generation limit allocation results to obtain control actions that meet power balance constraints, energy storage operation boundary constraints, load regulation boundary constraints and transmission bottleneck or section safety constraints. The control execution and rolling update module is used to execute control actions for the current scheduling cycle and trigger the rolling repetition of each module based on the updated running data.

[0014] This invention also discloses a multi-dimensional cost-benefit balancing switching control device for high-proportion renewable energy grid access, comprising: Memory, used to store computer programs; A processor for executing the computer program to implement the methods described above.

[0015] The present invention also discloses a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention improves the accuracy of state characterization by simultaneously identifying the adjustment needs of new energy sources and loads under short-term fluctuations, medium-term cyclical changes, and long-term trend shifts through multi-timescale state analysis. 2. This invention achieves a unified quantitative comparison among different regulation resources by constructing models for the marginal cost of energy storage, the marginal cost of load regulation, the marginal benefit of renewable energy absorption, and the marginal benefit of power grid operation; 3. This invention can dynamically switch between energy storage regulation, load reduction or peak shifting, and wind and solar curtailment, avoiding the high operating costs caused by relying solely on energy storage; 4. Under the condition of power transmission bottlenecks or cross-section limitations, this invention implements differentiated power generation restriction allocation based on the contribution of new energy stations to exceeding the limit and the marginal consumption revenue, thereby improving the pertinence and economy of power generation restriction control. 5. This invention improves the feasibility of engineering implementation by using safety constraint correction to ensure that control actions meet the safety constraints of power balance, energy storage boundary, load boundary, and transmission section. 6. The intelligent decision-making model in this invention is only a preferred implementation method and is compatible with rule-based decision-making, rolling optimization and other methods. Therefore, it has strong adaptability and scalability. Attached Figure Description

[0017] The present invention will now be described in detail with reference to the embodiments and accompanying drawings, wherein: Figure 1 A flowchart of a multi-dimensional cost-benefit equilibrium switching control method for high-proportion renewable energy grid access; Figure 2 This is the overall functional architecture diagram of the high-proportion renewable energy grid connection equalization switching control. Figure 3 This is a framework diagram for multi-timescale state analysis; Figure 4 This is a flowchart of the decision-making process for resource switching and differentiated issuance restrictions; Figure 5 This is a flowchart of safety constraint correction; Figure 6 This is the online scrolling control flowchart; Figure 7 This is a graph showing the results of running the example. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar components or components having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0019] In one embodiment, a multi-dimensional cost-benefit equilibrium switching control method for high-proportion renewable energy grid connection is described in [reference]. Figure 1 This includes the following steps.

[0020] Step 1: Obtain operational data from new energy stations, energy storage devices, adjustable loads, and power transmission bottlenecks / sections.

[0021] Specifically, the new energy station includes wind farms and photovoltaic power stations. Wind farms are equipped with wind turbine generators, box-type transformers, and data acquisition and monitoring control systems. Photovoltaic power stations include photovoltaic modules, combiner boxes, inverters, and environmental monitoring instruments. Energy storage devices utilize lithium-ion battery energy storage systems, consisting of battery modules, a battery management system, an energy storage converter, and an energy management platform. Adjustable loads include industrial interruptible loads, commercial flexible loads, and electric vehicle charging loads, each equipped with intelligent control terminals and remote communication modules. Synchronous phasor measurement units, line protection and control devices, and data concentrators are installed at transmission checkpoints or sections. Operating data includes at least the predicted and measured output of new energy sources, predicted and measured load values, energy storage state of charge and health status, maximum and minimum charging and discharging power of energy storage, upper and lower limits of energy storage capacity, upper and lower limits of load shedding capacity, upper and lower limits of load peak shifting capacity, and active power transmission limits at transmission checkpoints or sections. All of the above data can be collected and sent to the dispatch center's data server via fiber optic Ethernet or a 4G wireless network.

[0022] Step 2: Perform layered processing on the renewable energy output curve and load curve according to multiple time scales, calculate the deviation between renewable energy output and load demand, as well as the corresponding regulation power demand, regulation energy demand and regulation rate demand.

[0023] Specifically, multiple time scales are used, including short-term scales such as 15 minutes to 4 hours, medium-term scales such as 1 to 7 days, and long-term scales such as 1 month to 1 year. Layered processing employs discrete wavelet transform or empirical mode decomposition to decompose the original time series into components of different frequency bands, each component corresponding to the fluctuation trend at a specific time scale. For each time scale, the difference between the total output of new energy sources and the total load at that scale is calculated to obtain the deviation. Based on this, the peak value of the deviation is identified as the regulating power demand, the integral of the deviation over time is identified as the regulating energy demand, and the maximum difference between adjacent time periods of the deviation is identified as the regulating rate demand. These calculations are performed by the online analysis server of the dispatch center, and the results are stored in numerical vector form for use in subsequent steps.

[0024] Step 3: Construct models for the marginal cost of energy storage, the marginal cost of load regulation, the marginal benefit of renewable energy absorption, and the marginal benefit of grid operation.

[0025] Specifically, the energy storage marginal cost model characterizes the increased cost of additionally calling up a unit of power or a unit of electricity for energy storage. This cost includes power call cost, capacity occupancy cost, cycle life loss cost, charging and discharging loss cost, and ramping capability cost. Each cost can be expressed as a derivative near the current operating point. The load regulation marginal cost model characterizes the increased compensation cost of additionally reducing or transferring a unit of load. For industrial interruptible loads, the compensation unit price is determined by the annual agreement; for commercial flexible loads and electric vehicle charging loads, the compensation unit price is obtained through time-of-use or real-time bidding. The renewable energy marginal consumption revenue model characterizes the comprehensive revenue obtained from additionally consuming a unit of renewable energy electricity. This revenue includes revenue from grid-connected electricity, revenue from green environmental attributes, and revenue from carbon emission reduction. The revenue from green environmental attributes is referenced to the green certificate market price, and the carbon emission reduction revenue is referenced to the carbon quota trading price. The grid operation marginal revenue model characterizes the operating cost savings of the system after additionally performing a unit of regulation action. This cost saving is reflected in four aspects: peak shaving and valley filling, release of reserve capacity, relief of line congestion, and delay of expansion of transmission and transformation equipment. The mathematical expressions for all four models can be taken as linear functions, with coefficients obtained through regression fitting of historical data or online identification. Model construction is performed by the cost-benefit modeling module, which runs on a high-performance computing server.

[0026] Step 4: Based on the comparison results of the marginal cost of energy storage, the marginal cost of load regulation, the marginal benefit of renewable energy absorption, and the marginal benefit of grid operation, determine the resource switching strategy among energy storage regulation, load reduction or peak shifting, and wind and solar curtailment. See also Figure 4 .

[0027] Specifically, the comparison process first determines whether the marginal cost of energy storage is not higher than the sum of the marginal benefits of renewable energy absorption and the marginal benefits of grid operation. If this condition is met, energy storage regulation is prioritized for the current scheduling cycle, i.e., the energy storage device is commanded to charge or discharge to balance the deviation between renewable energy and load. If this condition is not met, it further determines whether the marginal cost of load regulation is lower than the marginal cost of energy storage and whether the current adjustable load capacity meets the regulation requirements. If both conditions are met, load reduction or peak shifting is prioritized, i.e., reducing electricity consumption or shifting the electricity consumption period to another time by issuing control commands to the adjustable load. If neither of the above conditions is met, i.e., both the marginal cost of energy storage and the marginal cost of load regulation are higher than the sum of the marginal benefits of renewable energy absorption and the marginal benefits of grid operation, then wind and solar curtailment control is triggered, i.e., reducing the output of renewable energy stations. The resource switching strategy is generated by the resource switching decision module, which is implemented using a rule engine or decision tree. The output result is a triplet identifier indicating the type of regulation resource to be used in the current scheduling cycle.

[0028] Step 5: When there is a risk of exceeding the limit at a power transmission bottleneck or section, differentiated power generation restriction allocation is implemented for multiple new energy stations based on the contribution of the new energy station to the limit exceedance and the corresponding marginal consumption revenue.

[0029] Specifically, the criterion for judging the risk of exceeding the limit can be the ratio of the current power flow value of the transmission bottleneck or section to the transmission limit. When this ratio exceeds a preset threshold, such as 0.9, and the power flow is predicted to continue to rise in the next scheduling cycle, it is determined that there is a risk of exceeding the limit. The contribution of the renewable energy station to exceeding the limit is obtained by multiplying the station's current output by the power flow distribution factor of the section. The power flow distribution factor is obtained through offline power flow calculation or online state estimation, and its physical meaning is the change in power flow of the section when the renewable energy station injects a unit power. The larger the contribution value, the greater the impact of the station on exceeding the limit. Based on this, a comprehensive ranking is performed by combining the marginal absorption benefit of each renewable energy station. Stations with lower marginal absorption benefit and higher contribution are given priority for power restriction. The result of the differentiated power restriction allocation is that each renewable energy station is assigned a power restriction value, and the sum of all power restrictions equals the power reduction required to eliminate the limit. This allocation is performed by the differentiated power restriction allocation module, which has a built-in priority ranking algorithm.

[0030] Step 6: Perform safety constraint correction on the resource switching strategy and differentiated power generation limitation allocation results to obtain control actions that satisfy power balance constraints, energy storage operation boundary constraints, load regulation boundary constraints, and transmission bottleneck or section safety constraints. See also Figure 5 .

[0031] Safety constraint correction can be achieved using optimization methods, with the objective function being to minimize the sum of squared deviations between the corrected control actions and the original actions. Power balance constraints require that the sum of the output of all power generation units, including renewable energy stations and energy storage, plus the power received at the cross-section minus the power consumed by the load, equals the total system loss. Energy storage operation boundary constraints include a state of charge not exceeding the upper limit and not falling below the lower limit, charging and discharging power not exceeding the maximum allowable value, and the rate of change of charging and discharging power not exceeding the ramp rate limit. Load regulation boundary constraints include power reduction not exceeding the maximum allowable value of interruptible loads, peak-shifting power not exceeding the maximum allowable value of transferable loads, and the total electricity consumption before and after peak-shifting remaining constant. Transmission bottleneck or cross-section safety constraints require that the active power flow at the cross-section does not exceed its thermal stability limit and transient stability limit. The optimization problem can be solved using a quadratic programming algorithm, with the solver running on a real-time control server, outputting feasible control actions that satisfy all constraints within milliseconds.

[0032] Step 7: Execute the control action for the current scheduling cycle, and repeat steps 1 to 7 on a rolling basis according to the updated running data.

[0033] Specifically, control actions can be remotely commanded to the energy management platform of the renewable energy station, the energy storage converter controller of the energy storage device, and the load control terminal of the adjustable load. Upon receiving the command, the renewable energy station's energy management platform adjusts the active power setpoint of the inverter; the energy storage converter controller adjusts the charging and discharging power; and the load control terminal sends interruption or delay commands to industrial equipment, commercial air conditioners, or electric vehicle charging piles. After execution, the system waits for the start of the next scheduling cycle, collects the latest operating data again, and then automatically jumps back to step 1 to repeat the entire process. The length of the rolling cycle can be configured between 5 minutes and 1 hour depending on the grid regulation needs; a shorter cycle provides a stronger response to rapid fluctuations. Through continuous rolling execution, this method can maintain the most economically optimal switching between energy storage regulation, load regulation, and wind and solar curtailment during continuous changes in grid operating conditions, achieving a dynamic balance between renewable energy consumption benefits and overall system regulation costs, while ensuring the safe and stable operation of the grid.

[0034] The method in this embodiment is an online rolling control process, which in practical applications usually runs continuously until the system stops or is manually terminated.

[0035] In one specific embodiment, see Figure 3 The multiple time scales in step 2 include intraday scales and cross-day scales, and the cross-day scales include at least one of weekly scales, monthly scales, quarterly scales, or annual scales.

[0036] Specifically, the intraday scale is used to capture short-term fluctuations in renewable energy output and load demand, such as minute- to hourly output fluctuations in photovoltaic power plants caused by cloud cover, and intraday variations in morning and evening peak loads. The cross-day scale is used to analyze the periodic shifts in renewable energy generation and the medium- to long-term trends in load demand. The weekly scale corresponds to the differences in electricity consumption patterns between weekdays and weekends, the monthly scale corresponds to changes in sunlight and wind caused by seasonal transitions, the quarterly scale corresponds to the renewable energy output characteristics of different seasons, and the annual scale corresponds to interannual climate fluctuations and equipment aging trends.

[0037] In practical applications, the dispatch center selects an appropriate combination of time scales based on the grid scale and regulation needs. For example, in regional power grids with a high proportion of wind power, both intraday and weekly scales are used simultaneously; in systems with seasonal hydropower and photovoltaic complementarity, a quarterly scale is added. The historical data window length corresponding to each time scale is at least 3 to 5 times the scale to ensure the representativeness of statistical characteristics. The hierarchical processing of multiple time scales is completed by a data preprocessing server equipped with large-capacity memory and high-speed solid-state drives, which uses sliding window technology to update the feature parameters of each scale online.

[0038] In a specific embodiment, the marginal cost of energy storage in step 3 consists of at least two of the following: energy storage power call-up cost, energy storage capacity occupancy cost, energy storage cycle life loss cost, energy storage charging and discharging loss cost, and energy storage ramp-up capability cost.

[0039] Specifically, the energy storage power call cost reflects the additional losses of the energy storage converter and the energy consumption of auxiliary equipment per unit power call. This cost has a linear or quadratic relationship with the current charge / discharge power. The energy storage capacity occupancy cost reflects the impact on subsequent regulation capabilities when a unit of electricity occupies energy storage capacity. When the energy storage state of charge approaches its upper or lower limit, the capacity occupancy cost increases significantly. The energy storage cycle life reduction cost is calculated based on the battery's cycle life curve. The cycle life of lithium iron phosphate batteries and ternary lithium batteries has an exponential relationship with the depth of discharge. Each additional charge / discharge cycle or deeper discharge reduces the remaining battery life, which is converted into an additional cost per charge / discharge. The energy storage charge / discharge loss cost includes battery internal resistance loss, converter switching loss, and line loss. The charge / discharge efficiency of lithium-ion batteries at a 0.5C rate is approximately 92% to 95%, and the lost electricity is converted into a cost based on the real-time electricity price. The energy storage ramp-up capability cost characterizes the additional stress on the battery's internal structure and converter switching devices caused by rapidly changing the charge / discharge power. This cost is triggered when the ramp rate exceeds the rated value of 0.2C per minute. Among the aforementioned costs, the cycle life reduction cost and the ramp-up capability cost can be calculated in real time using the cumulative charge and discharge amount and power change rate recorded by the battery management system. In practice, a lookup table method or a neural network fitting function is used.

[0040] In a specific embodiment, the marginal cost of load adjustment in step 3 includes at least the cost of load reduction compensation and the cost of load peak shifting compensation. For adjustable loads participating in peak shifting, the total electricity consumption before and after peak shifting is kept in line with the constraint.

[0041] Specifically, load shedding compensation cost refers to the fee paid by the power system to users in exchange for their temporary power outages. The unit price for load shedding compensation for industrial interruptible loads is stipulated in the annual interruptible load contract signed by both parties, and is usually priced in segments based on the interruption power and time, for example, 0.5 to 1.2 yuan per kilowatt. For commercial flexible loads such as shopping mall air conditioning and office building lighting, the load shedding compensation is determined through demand response bidding, and the unit price fluctuates with the number of responses and the time period. Peak shaving compensation cost refers to the fee paid by the power system to users in exchange for them shifting their electricity consumption from peak to off-peak hours. Peak shaving compensation for electric vehicle charging loads can utilize time-of-use pricing differences; users actively shift their charging time to off-peak hours at night, and the system compensates for the difference between the charging cost and the usual charging cost.

[0042] For adjustable loads participating in peak shaving, the total electricity consumption before and after peak shaving must be equal in value within a complete cycle, such as 24 hours, to ensure that the electricity moved out equals the electricity moved in, thus ensuring that the user's overall electricity demand is not affected. The metering of peak shaving is done by smart meters. Meter data is uploaded to the demand response platform via power line carrier or wireless public network. The platform calculates the actual transferred electricity based on the load curves before and after peak shaving and checks the conservation constraints. If the deviation exceeds the allowable range, such as 5%, the compensation fee is adjusted or the peak shaving request is rejected.

[0043] In one specific embodiment, the marginal consumption benefit of new energy in step 3 includes at least one of the following: new energy grid connection electricity revenue, green environmental attribute revenue, and carbon emission reduction revenue.

[0044] Specifically, the revenue from renewable energy grid connection refers to the electricity revenue earned by a renewable energy power plant for each additional kilowatt-hour generated. For onshore wind power and centralized photovoltaic power plants, the grid connection price is based on the nationally approved benchmark price or the contract price formed through market-based transactions. The revenue from green environmental attributes refers to the revenue obtained from selling green electricity certificates generated by renewable energy power generation on the green certificate trading market. The price of green certificates fluctuates with market supply and demand. The revenue from carbon emission reduction refers to the revenue from the reduction in carbon dioxide emissions resulting from renewable energy replacing fossil fuel power generation on the carbon quota trading market. The emission reduction factor is calculated based on the average carbon emission intensity of the local power grid, and the carbon quota price is based on the national carbon emission trading market price. These three revenue streams can be used individually, or a combination of two or three, depending on the policies and market environment of the region where the power grid is located. The renewable energy marginal consumption revenue model is updated in real time by the revenue assessment module. This module connects to the power trading center and the carbon emission trading exchange through an application programming interface, obtaining the latest price data at regular intervals for subsequent resource switching comparisons.

[0045] In one specific embodiment, the marginal benefits of power grid operation in step 3 include at least one of peak shaving and valley filling benefits, reserve capacity saving benefits, congestion mitigation benefits, and delayed capacity expansion benefits.

[0046] Specifically, peak shaving and valley filling benefits refer to the cost savings in power generation achieved by adjusting energy storage charging and discharging or load shifting to reduce peak loads and increase off-peak loads. During peak periods, high-cost gas turbines or pumped-storage hydroelectric power plants are used, while during off-peak periods, low-cost baseload units are fully utilized. After peak shaving and valley filling, the average power generation cost of the system decreases, and the amount of this decrease is the marginal benefit of peak shaving and valley filling. Reserve capacity savings refer to the reduction in demand for spinning reserves due to adjustments in resource input. Generators that would otherwise need to be reserved can be released to participate in the market or reduce no-load losses. The savings in reserve capacity purchase or operating costs are the reserve capacity savings. Congestion mitigation benefits refer to the reduction in power flow exceeding limits on transmission lines or sections by adjusting the output of new energy sources or load distribution, thereby avoiding or delaying losses caused by high-priced unit substitution or load shedding due to congestion. Congestion mitigation benefits equal the cost of avoiding congestion. Delayed expansion revenue refers to the revenue generated by postponing the upgrading and transformation of power transmission and transformation equipment through demand-side regulation and energy storage support. This postpones the expansion investment that would normally be made in the current year to future years and converts it into current revenue based on the time value of money. Delayed expansion revenue equals the annualized amortization of the expansion investment multiplied by the number of years of delay and then multiplied by the discount factor.

[0047] The above four benefits can be selected to be included in the model based on the actual operation of the power grid. For example, for urban distribution networks, peak shaving and valley filling benefits and delayed capacity expansion benefits are more significant; for inter-regional interconnection sections, congestion mitigation benefits are more important. The marginal benefit model of power grid operation is calculated by the economic operation assessment module, which calls power system production simulation software and power flow calculation tools, and updates the values ​​of each marginal benefit once per scheduling cycle.

[0048] In one specific embodiment, the resource switching strategy in step 4 includes one or more of the following three rules.

[0049] The first rule is: when the marginal cost of load adjustment is lower than the marginal cost of energy storage, and the adjustable load capacity meets the current adjustment demand, load reduction or peak shifting should be prioritized. Adjustable load capacity is statistically analyzed in real time by the load aggregation platform. This platform collects current interruptible and transferable capacity through intelligent control terminals installed on industrial equipment, commercial air conditioning units, and electric vehicle charging piles, and uploads this data to the dispatch center. If the marginal cost of load adjustment is lower than the marginal cost of energy storage, it indicates that calling user-side resources is more economical than calling energy storage. Simultaneously, if the adjustable load capacity is greater than or equal to the adjustment power demand, the resource switching decision module outputs a load adjustment priority instruction.

[0050] The second rule is as follows: when both the marginal cost of energy storage and the marginal cost of load regulation exceed the sum of the marginal benefits of renewable energy absorption and the marginal benefits of grid operation, wind and solar curtailment control is triggered. In this case, regardless of whether energy storage or load regulation continues to be used, the unit regulation cost exceeds the comprehensive benefits that can be obtained by continuing to absorb renewable energy. Therefore, from an economic perspective, it is more cost-effective to give up some renewable energy output than to call up additional regulation resources, and the resource switching decision module outputs a wind and solar curtailment command.

[0051] The third rule is: when the marginal cost of energy storage is not higher than the sum of the marginal revenue from renewable energy absorption and the marginal revenue from grid operation, energy storage regulation should be prioritized. In this case, the cost of energy storage deployment can be covered by the revenue from renewable energy absorption and the revenue from system operation, making energy storage regulation economically feasible. Therefore, the resource switching decision module outputs a priority instruction for energy storage regulation.

[0052] The above rules can be used individually or in combination to form a complete decision-making logic. Specifically, within each scheduling cycle, the system first checks if the energy storage priority condition is met. If it is, energy storage regulation is employed; if not, the system further checks the load regulation alternative condition. If the load regulation is met, load regulation is employed; if neither of the above conditions is met, wind and solar power curtailment is triggered. The rule engine is deployed on a real-time control server and implemented using a finite state machine, with output results generated within milliseconds.

[0053] In a specific embodiment, the contribution of the renewable energy station to the limit exceedance in step 5 is determined jointly by the renewable energy station's output and the corresponding power flow distribution factor. The power flow distribution factor represents the change in active power flow at the target transmission bottleneck or section when the renewable energy station injects a unit of active power. Its value can be obtained through offline power flow calculation or updated in real time using online state estimation methods. Specifically, for a regional power grid containing multiple renewable energy stations, firstly, based on the ground-state power flow, a small disturbance is added to the output of each renewable energy station, and the increment of power flow at the section is calculated separately. The ratio of the increments is used as the distribution factor of that station. The contribution calculation formula is: the station's output multiplied by the power flow distribution factor of that station at the section. The larger the product, the greater the impact of the station on the limit exceedance at the section.

[0054] The differentiated power restriction allocation in step 5 is determined by jointly ranking the renewable energy stations based on their over-limit contribution, current ramp intensity, and marginal absorption benefit. The current ramp intensity is taken from the actual output change rate of the renewable energy station, i.e., the output difference between the last two scheduling cycles divided by the time interval. Stations with high ramp intensity experience drastic output fluctuations, and power restriction on them has a smaller impact on system stability. During joint ranking, a comprehensive ranking index is calculated for each renewable energy station. For example, the over-limit contribution is multiplied by the ramp intensity and then divided by the marginal absorption benefit. Stations are ranked from largest to smallest index, with priority given to stations with higher indexes. The ranking algorithm is executed by the differentiated power restriction allocation module, which uses a fast ranking algorithm and can handle the ranking needs of dozens of renewable energy stations simultaneously. After determining the power restriction order, the power restriction amount is allocated sequentially from the station ranked first, according to the total power restriction amount, until the cumulative power restriction amount reaches the required reduction value. The power restriction amount for each station does not exceed the difference between its current output and minimum technical output to ensure the safe operation of the renewable energy station.

[0055] In a specific embodiment, the safety constraint correction in step 6 is achieved by constructing an optimization problem with the objective of minimizing the deviation between the corrected control action and the original action. The original action refers to the uncorrected control commands generated by resource switching strategies and differentiated power rationing, including energy storage charging and discharging power, load reduction or peak shifting, and power rationing at renewable energy stations. Since the original action may violate the grid operation boundary, a feasible corrected action with the minimum deviation needs to be found in the correction process. The objective function for minimizing deviation is usually in quadratic form, that is, the sum of squares of the differences between the corrected action and the original action.

[0056] The constraints of the optimization problem include at least the following six categories: power balance constraints, requiring the sum of total grid power generation, received power, load power consumption, and grid losses to be equal; energy storage state boundary constraints, requiring the energy storage state of charge to be between the allowable minimum and maximum values; energy storage power boundary constraints, requiring the energy storage charging and discharging power to not exceed the rated power and the power limit corresponding to the current state of charge; energy storage ramp rate constraints, requiring the change in charging and discharging power between adjacent scheduling cycles to not exceed the maximum ramp rate capability of the energy storage converter; load regulation boundary constraints, requiring the load reduction to not exceed the contracted capacity of interruptible loads, the load peak shifting to not exceed the maximum shifting capacity of transferable loads, and the total electricity consumption before and after peak shifting to be conserved; and transmission bottleneck or cross-section power flow safety constraints, requiring the active power flow at the cross-section to not exceed the thermal stability limit and the dynamic stability limit. This optimization problem is a convex quadratic programming problem, which can be solved using an open-source solver running on a safety constraint correction server equipped with a multi-core CPU and large-capacity memory to ensure completion within seconds. The solution result is the feasible control action that satisfies all constraints, serving as the final issued command.

[0057] In one specific embodiment, at least one of steps 4 to 6 is executed by an intelligent decision-making model trained based on historical operational data; the intelligent decision-making model includes one or more of the following: constrained reinforcement learning model, hierarchical reinforcement learning model, multi-agent reinforcement learning model, rolling optimization model, or rule optimization model. Steps 4 to 6 are also the resource switching strategy, differentiated limited-release allocation, and security constraint correction.

[0058] Historical operational data includes at least the past year's renewable energy output sequences, load sequences, energy storage operation records, cross-sectional power flow data, and actual control actions and corresponding cost-benefit results. During the training phase, when using a constrained reinforcement learning model, the agent learns the optimal strategy through interaction with the environment. Constraints act as an action shield, prohibiting the agent from outputting actions that violate safety boundaries. The hierarchical reinforcement learning model decomposes resource switching and differentiated power curtailment allocation into upper-layer strategy selection subtasks and lower-layer parameter optimization subtasks. The upper layer decides whether to use energy storage, load, or wind / solar curtailment, while the lower layer determines the specific power values. The multi-agent reinforcement learning model models each renewable energy station, energy storage device, and load aggregator as an independent agent, with each agent achieving global optimum through a collaborative mechanism. The rolling optimization model solves a finite-time-domain optimization problem in each scheduling cycle, executing only the actions for the first time period, and re-optimizing actions for subsequent time periods in the next cycle. The rule-based optimization model automatically mines decision rules from historical data, forming a rule base, and makes decisions based on rules in the rule base during online runtime.

[0059] The training of the aforementioned intelligent decision-making model is completed on an offline server, and the trained model parameters are then deployed to a real-time control server. The model type can be flexibly selected based on the power grid scale and computing resources; for example, a hierarchical reinforcement learning model can be used for large regional power grids, while a rule-based optimization model can be used for distribution networks. By replacing the traditional rule engine with an intelligent decision-making model, better decision-making strategies can be learned from historical operational experience, further improving the benefits of renewable energy consumption and the economic efficiency of regulation.

[0060] In one embodiment, the overall control architecture is described in [reference]. Figure 2 A multi-dimensional cost-benefit balance switching control system for high-proportion renewable energy grid access is provided for executing the methods described in the above embodiments. The system includes a data acquisition module, a multi-timescale analysis module, a cost-benefit modeling module, a resource switching decision module, a differentiated limited generation allocation module, a safety constraint correction module, and a control execution and rolling update module.

[0061] The data acquisition module is used to collect data on the operating status of new energy sources, energy storage, load, and the power grid; the multi-timescale analysis module is used to perform hierarchical processing of new energy output and load curves and extract key state features; the cost-benefit modeling module is used to establish models for the marginal cost of energy storage, the marginal cost of load regulation, the marginal benefit of new energy absorption, and the marginal benefit of power grid operation; the resource switching decision module is used to determine the switching strategy among energy storage regulation, load reduction or peak shifting, and wind and solar curtailment; the differentiated power generation restriction allocation module is used to determine the power generation restriction targets and amounts of different new energy stations when cross-sectional constraints exist; the safety constraint correction module is used to perform feasibility corrections before the action is executed; and the control execution and rolling update module is used to execute the control actions of the current scheduling cycle and roll into the next cycle.

[0062] Specifically, the data acquisition module is used to acquire operational data from renewable energy stations, energy storage devices, adjustable loads, and transmission bottlenecks or sections. The multi-timescale analysis module is used to perform layered processing of renewable energy output curves and load curves according to multiple time scales, calculating the deviation between renewable energy output and load demand, as well as the corresponding regulation power demand, regulation energy demand, and regulation rate demand. The cost-benefit modeling module is used to construct models for the marginal cost of energy storage, the marginal cost of load regulation, the marginal benefit of renewable energy absorption, and the marginal benefit of grid operation. The resource switching decision module is used to determine resource switching strategies among energy storage regulation, load reduction or peak shifting, and wind and solar curtailment based on the comparison results of the aforementioned marginal costs of energy storage, load regulation, renewable energy absorption, and grid operation. The differentiated generation restriction allocation module is used to implement differentiated generation restriction allocation for multiple renewable energy stations based on their contribution to the over-limit and the corresponding marginal absorption benefit when there is a risk of exceeding the limit at a transmission bottleneck or section. The safety constraint correction module is used to perform safety constraint correction on the resource switching strategy and the differentiated generation restriction allocation results to obtain control actions that meet power balance constraints, energy storage operation boundary constraints, load regulation boundary constraints, and transmission bottleneck or section safety constraints. The control execution and rolling update module is used to execute the control actions of the current scheduling cycle and trigger the rolling repetition of each module based on the updated operating data.

[0063] In one embodiment, a multi-dimensional cost-benefit balancing switching control device for high-proportion renewable energy grid access includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described in the above embodiment.

[0064] In one embodiment, a computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the methods described above.

[0065] In a more specific embodiment, the multi-timescale state analysis is as follows.

[0066] Time scale definition, let the set of time scales be: In practical applications, all or part of the time scales can be selected based on the scheduling level and control objectives.

[0067] Feature extraction. Let the predicted output of renewable energy station i at time t be... The baseline value of the load at time t is L. t The output and load of new energy sources are aggregated, analyzed using sliding statistics, or decomposed into trend fluctuations at different time scales to obtain a standardized sequence. and .

[0068] At any time scale s, the following features are extracted.

[0069] The maximum predicted output of the new energy station r over a time scale s: .

[0070] Minimum predicted output of the new energy station r over time scale s: .

[0071] Output fluctuation at this scale: The rate of increase at this scale: Correspondingly, the load-side characteristics include the following.

[0072] Maximum and minimum load values ​​over time scale s: Load fluctuation: Load ramp rate: Furthermore, adjustable load capacity, transfer duration, cross-sectional remaining margin, and over-limit risk indicators can also be extracted.

[0073] Demand adjustment calculation. On a time scale *s*, the deviation between new energy sources and load is defined as: in, Indicates the amount of basic regulatory resources available. This represents the equivalent adjustment margin after considering network constraints.

[0074] Therefore, the following calculation is performed: in, To adjust power demand, To regulate energy demand, To adjust the rate requirement.

[0075] In a more specific embodiment, the multidimensional cost-benefit model is as follows.

[0076] Energy storage marginal cost model. The marginal cost of energy storage is used to characterize the comprehensive cost brought about by a unit of newly added energy storage regulation capacity, and can be expressed by the following formula: in: Power is allocated for energy storage; This refers to the amount of energy storage capacity occupied. This is a lifespan reduction function; To consume electricity; The rate of change of energy storage power; a 1- a5 represents the corresponding cost coefficient.

[0077] Load adjustment marginal cost model. The marginal cost of load adjustment is used to characterize the compensation cost of adjusting a unit of additional load, and can be expressed as: in, The marginal amount of reduction per unit load, This is the marginal amount of peak shifting per unit load. and These are the unit price for reduced compensation and the unit price for peak shifting compensation, respectively.

[0078] For peak-shifting loads, the following conditions must be met: .

[0079] The marginal benefit model for renewable energy consumption. The marginal benefit of renewable energy consumption characterizes the benefit generated by consuming a unit of newly added renewable energy, and can be expressed as: in, For electricity revenue, Benefiting from green environmental attributes, For carbon emission reduction benefits.

[0080] The marginal revenue model for power grid operation. The marginal revenue of power grid operation is used to characterize the system operating revenue brought about by a unit of additional adjustment action, and can be expressed as: in , , and These represent the marginal amounts corresponding to peak shaving and valley filling benefits, reserve saving benefits, congestion mitigation benefits, and delayed capacity expansion benefits, respectively.

[0081] In a more specific embodiment, the core of this invention lies in dynamically switching between energy storage regulation, load reduction or peak shifting, and wind and solar curtailment based on a comparison of multidimensional marginal costs and marginal benefits. The resource switching strategy is as follows.

[0082] Energy storage priority conditions. When the following conditions are met: When the cost of new energy storage regulation is no higher than the sum of the new energy consumption and the system operation benefits, energy storage regulation should be given priority.

[0083] Load adjustment alternative conditions. When the following conditions are met: When the adjustable load capacity meets the current regulation needs, load reduction or peak shifting should be given priority to replace high-cost energy storage regulation.

[0084] Conditions for triggering wind and solar power curtailment. When the following conditions are met: ,and: Whether energy storage or load regulation continues to be used, the marginal cost is higher than the marginal benefit of continuing to absorb new energy sources, thus triggering wind and solar curtailment control.

[0085] Through the above switching logic, this invention achieves a balanced control mechanism where the cost is absorbed when revenue covers the cost, the load is switched to the load when the load is cheaper than energy storage, and power generation is limited when all are uneconomical.

[0086] In a more specific embodiment, when there is a risk of exceeding the limit at a transmission bottleneck or section, this invention does not adopt a uniform proportional limit allocation, but rather allocates the limit based on the contribution of new energy sites to the limit exceedance and the corresponding marginal revenue. The specific differentiated limit allocation mechanism is as follows.

[0087] Calculation of Exceeding Limits Contribution. For line or section l, the exceeding limit contribution of renewable energy station i can be expressed as: .in, The power flow distribution factor of the new energy station i for the line or section l.

[0088] Site-level priority indicators for restricted power generation. Further defining the priority indicators for restricted power generation at new energy stations as: .in, Indicates the current slope intensity of the station. This indicates the marginal absorption revenue of the site. It is a very small positive number. Prioritize... Larger new energy power stations will be subject to emission restrictions.

[0089] This mechanism allows the selection of entities subject to issuance restrictions to consider network impact, dynamic fluctuation characteristics, and economic efficiency simultaneously, rather than simply restricting issuance based on installed capacity or a uniform ratio.

[0090] In a more specific embodiment, the safety constraint correction mechanism is as follows.

[0091] To ensure that control actions meet the power grid operation boundaries, this invention sets a unified safety constraint correction step before the action is executed. Let the original actions obtained from the resource switching strategy and differentiated generation limiting allocation be... Then the safety constraint correction problem can be expressed as: The constraints must include at least the following:

[0092] First, power balance constraints: , Wherein, Nr represents the set of new energy stations, Ne represents the set of energy storage devices, and Nl represents the set of load nodes. To contribute to actual corrections; This refers to the discharge power. This refers to the charging power. Supports external power transmission; Baseline value for basic electrical load; Reduce power to the load; This refers to the amount of load reduction. This represents the increase in load.

[0093] Secondly, energy storage operation boundary constraints: ;in, These represent the energy state (or charge capacity) of the energy storage device at the current time and the next time, respectively. and The charging efficiency coefficient and discharging efficiency coefficient of energy storage devices; ,in , These represent the minimum and maximum energy boundaries that allow the energy storage device to operate. ,in , These represent the maximum allowable charging power limit and discharging power limit for the energy storage device, respectively. ,in, This indicates the maximum permissible ramp rate for the energy storage device.

[0094] Third, load regulation boundary constraints: , This represents the maximum load reduction capacity. , , These are the maximum power limits for outgoing and incoming power, respectively. This is the constraint on the conservation of electricity consumption within the cycle.

[0095] Fourth, cross-sectional power flow safety constraints: Actual power flow calculation values: ,in, Basic power flow benchmark value For power transfer distribution factor, This represents the net increase in injected power. , This is the upper limit for transmission.

[0096] Corrected actions As the final control action to be executed in the current scheduling cycle, This represents the final output control motion vector, which is the corrected control vector after optimization and projection correction through all the above safety constraints.

[0097] In a more specific embodiment, the core of this invention lies in the resource switching and differentiated allocation mechanism, rather than being limited to a specific algorithm. To improve the efficiency and adaptability of strategy generation, at least one of the resource switching strategy, differentiated limited allocation, or security constraint correction can be generated by an intelligent decision-making model trained based on historical operational data. Preferred implementations of the intelligent decision-making model may include, but are not limited to: constrained reinforcement learning models; hierarchical reinforcement learning models; multi-agent reinforcement learning models; rolling optimization models; rule-based optimization models; and combinations of the above methods.

[0098] In a preferred embodiment, an upper-level coordination model can output resource weights or quotas for energy storage, load, and wind and solar curtailment, while a lower-level execution model outputs the curtailment of renewable energy stations, the charging and discharging power of energy storage, and the load reduction or peak shifting. Furthermore, safety constraint correction can be modeled as a feasible domain projection or an online optimization correction process.

[0099] In a more specific embodiment, see Figure 6 The online rolling control process of the present invention includes: Step S1: Collect operational data of new energy sources, energy storage, loads, and transmission bottlenecks or sections for the current scheduling cycle; Step S2: Perform multi-timescale state analysis to extract the characteristics of new energy sources and load fluctuations; Step S3: Calculate the deviation, regulation power requirement, regulation energy requirement, and regulation rate requirement; Step S4: Establish and update models for the marginal cost of energy storage, the marginal cost of load regulation, the marginal benefit of renewable energy consumption, and the marginal benefit of grid operation; Step S5: Determine the resource switching strategy based on the marginal comparison results; Step S6: When there is a risk of exceeding the limit at a cross section, implement differentiated issuance restriction allocation; Step S7: Perform safety constraint correction on the original control action; Step S8: Execute the control action for the current scheduling cycle; Step S9: Update the time status variables and roll into the next scheduling cycle.

[0100] In a more specific embodiment, a regional power grid connects to one wind-solar combined renewable energy station, one electrochemical energy storage system, and three types of adjustable loads, and is connected to the main grid through a critical section. The system operates on a 15-minute dispatch cycle, with 96 time periods per day. The installed wind power capacity is 120 MW, the installed photovoltaic capacity is 80 MW, the rated energy storage power is 50 MW, the rated capacity is 100 MWh, the initial state of charge is 45%, the operating range is 15% to 95%, and the maximum ramp rate is 20 MW / 15min. The maximum reduction limit for industrial interruptible loads is 10 MW, the maximum peak shift limit for commercial flexible loads is 8 MW, and the maximum peak shift limit for electric vehicle charging loads is 12 MW. The transmission limit at the critical section is 150 MW.

[0101] The comprehensive unit revenue for new energy is 0.43 yuan / kWh; the compensation for industrial load reduction is 0.78 yuan / kWh; the compensation for commercial load peak shifting is 0.30 yuan / kWh; and the compensation for electric vehicle load peak shifting is 0.22 yuan / kWh. The marginal cost of energy storage changes dynamically according to the conditions, ranging from 0.24 to 0.34 yuan / kWh under normal operating conditions, and rising to 0.50 to 0.68 yuan / kWh under high SOC or rapid ramp-up conditions.

[0102] During the peak solar PV period at midday, as the SOC of energy storage increases, the marginal cost of energy storage rises. After comparing the marginal cost of energy storage, the marginal cost of load regulation, and the marginal benefit of renewable energy absorption, the system prioritizes switching some regulation tasks to the forward shift of electric vehicle charging loads and the peak shifting of commercial loads. When continued absorption will cause the cross-sectional power flow to exceed the limit and neither energy storage nor load regulation is sufficient to eliminate the risk, the target and amount of power generation to be restricted are determined based on the contribution of renewable energy stations to the cross-sectional limit exceedance and their marginal benefits.

[0103] During the rapid net load ramp-up period in the evening, if the marginal cost of energy storage is lower than the marginal cost of load regulation, the system will prioritize the discharge of energy storage, and only trigger a small amount of industrial load reduction when the energy storage is close to the power boundary.

[0104] Compared with traditional fixed-priority rule control schemes, the present invention can reduce high-cost energy storage dispatch, decrease the number of cross-section overruns, reduce the risk of energy storage overruns, and improve the overall net benefit of the system. See also Figure 7 .

[0105] This invention discloses a multi-dimensional cost-benefit equilibrium switching control method, system, equipment, and storage medium for high-proportion renewable energy grid integration. The method acquires operational data from renewable energy stations, energy storage devices, adjustable loads, and transmission bottlenecks or sections. It performs layered processing of renewable energy output curves and load curves according to multiple time scales, calculating the deviation between renewable energy output and load demand, as well as the corresponding regulation power demand, regulation energy demand, and regulation rate demand. It constructs models for energy storage marginal cost, load regulation marginal cost, renewable energy marginal absorption benefit, and grid operation marginal benefit. Based on the comparison results of the marginal costs and marginal benefits, it determines resource switching strategies among energy storage regulation, load reduction or peak shifting, and wind and solar curtailment. When there is a risk of exceeding limits at transmission bottlenecks or sections, it performs differentiated generation restriction allocation for multiple renewable energy stations based on the renewable energy station's contribution to exceeding limits and the corresponding marginal absorption benefit. It performs safety constraint correction on the resource switching strategies and differentiated generation restriction allocation results to obtain control actions that satisfy power balance constraints, energy storage operation boundary constraints, load regulation boundary constraints, and transmission bottleneck or section safety constraints. This invention can achieve a dynamic balance between the benefits of renewable energy consumption and the overall system regulation cost while ensuring the safe operation of the power grid.

[0106] In the description of this specification, the use of terms such as "Embodiment 1," "this embodiment," or "in one embodiment" indicates that the specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example; moreover, the specific features, structures, materials, or characteristics described may be combined in any appropriate manner in one or more embodiments or examples.

[0107] In the description of this specification, the terms "connection," "installation," "fixing," "setting," and "having" are interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0108] In the description of this specification, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0109] The above description of the embodiments is intended to enable those skilled in the art to understand and apply the technology of this invention. Those skilled in the art can easily make various modifications to these examples and apply the general principles described herein to other embodiments without creative effort. Therefore, this invention is not limited to the above embodiments. Modifications in the following situations should be within the scope of protection of this invention: ① New technical solutions implemented based on the technical solution of this invention and combined with existing common knowledge, where the technical effects of the new technical solution do not exceed the technical effects of this invention; ② Equivalent substitutions of some features of the technical solution of this invention using known technology, resulting in the same technical effects as those of this invention; ③ Extendable technical solutions based on the technical solution of this invention, where the substantive content of the extended technical solution does not exceed the technical solution of this invention; ④ Equivalent transformations made using the content of this specification and drawings, directly or indirectly applied to other related technical fields.

Claims

1. A multi-dimensional cost-benefit equilibrium switching control method for high-proportion renewable energy grid connection, characterized in that, Includes the following steps: Step 1: Obtain operational data for new energy stations, energy storage devices, adjustable loads, and power transmission bottlenecks / sections; Step 2: Perform layered processing on the new energy output curve and load curve according to multiple time scales, calculate the deviation between new energy output and load demand, as well as the corresponding regulation power demand, regulation energy demand and regulation rate demand. Step 3: Construct models for the marginal cost of energy storage, the marginal cost of load regulation, the marginal benefit of renewable energy consumption, and the marginal benefit of grid operation; Step 4: Based on the comparison results of the marginal cost of energy storage, the marginal cost of load regulation, the marginal benefit of renewable energy consumption, and the marginal benefit of grid operation, determine the resource switching strategy among energy storage regulation, load reduction or peak shifting, and wind and solar curtailment. Step 5: When there is a risk of exceeding the limit at a power transmission bottleneck or section, differentiated power generation restriction allocation is implemented for multiple new energy stations based on the contribution of the new energy station to the limit exceedance and the corresponding marginal consumption revenue. Step 6: Perform safety constraint correction on the resource switching strategy and differentiated power generation limitation allocation results to obtain control actions that meet power balance constraints, energy storage operation boundary constraints, load regulation boundary constraints, and transmission bottleneck or section safety constraints. Step 7: Execute the control action for the current scheduling cycle, and repeat steps 1 to 7 on a rolling basis according to the updated running data.

2. The method according to claim 1, characterized in that, The multiple time scales in step 2 include intraday scales and cross-day scales, and the cross-day scales include at least one of weekly scales, monthly scales, quarterly scales, or annual scales.

3. The method according to claim 1, characterized in that, The marginal cost of energy storage in step 3 consists of at least two of the following: energy storage power call-up cost, energy storage capacity occupancy cost, energy storage cycle life depreciation cost, energy storage charge and discharge loss cost, and energy storage ramp-up capability cost; and / or The marginal cost of load adjustment in step 3 includes at least the cost of load reduction compensation and the cost of load peak shifting compensation; and / or The marginal benefit of renewable energy consumption in step 3 includes at least one of the following: renewable energy grid connection electricity revenue, green environmental attribute revenue, and carbon emission reduction revenue; and / or The marginal benefits of power grid operation in step 3 include at least one of the following: peak shaving and valley filling benefits, reserve capacity saving benefits, congestion mitigation benefits, and delayed capacity expansion benefits.

4. The method according to claim 1, characterized in that, The resource switching strategy in step 4 includes: When the marginal cost of load regulation is lower than the marginal cost of energy storage, and the adjustable load capacity meets the current regulation needs, load reduction or peak shifting should be prioritized; and / or When both the marginal cost of energy storage and the marginal cost of load regulation exceed the sum of the marginal revenue from renewable energy integration and the marginal revenue from grid operation, wind and solar curtailment control is triggered; and / or When the marginal cost of energy storage is not higher than the sum of the marginal revenue from renewable energy consumption and the marginal revenue from grid operation, energy storage regulation should be given priority.

5. The method according to claim 1, characterized in that, The contribution of the renewable energy station to the overload in step 5 is determined by the power output of the renewable energy station and the corresponding power flow distribution factor; and / or The differentiated emission restriction allocation in step 5 is determined by jointly ranking the contribution of new energy stations to exceeding the limit, the current ramp intensity, and the marginal consumption benefit.

6. The method according to claim 1, characterized in that, The safety constraint correction in step 6 is achieved by constructing an optimization problem with the goal of minimizing the deviation between the corrected control action and the original action. The constraints of the optimization problem include at least power balance constraints, energy storage state boundary constraints, energy storage power boundary constraints, energy storage ramp rate constraints, load regulation boundary constraints, and power flow safety constraints at transmission bottlenecks / sections.

7. The method according to any one of claims 1-6, characterized in that, At least one of steps 4 to 6 is executed by an intelligent decision-making model trained based on historical operational data; the intelligent decision-making model includes one or more of the following: constrained reinforcement learning model, hierarchical reinforcement learning model, multi-agent reinforcement learning model, rolling optimization model, or rule optimization model.

8. A multi-dimensional cost-benefit equilibrium switching control system for high-proportion renewable energy grid integration, characterized in that, include: The data acquisition module is used to acquire operational data from new energy stations, energy storage devices, adjustable loads, and power transmission checkpoints or sections. The multi-timescale analysis module is used to perform layered processing of the new energy output curve and load curve according to multiple timescales, calculate the deviation between new energy output and load demand, as well as the corresponding regulation power demand, regulation energy demand and regulation rate demand. The cost-benefit modeling module is used to construct models for the marginal cost of energy storage, the marginal cost of load regulation, the marginal benefit of renewable energy consumption, and the marginal benefit of power grid operation. The resource switching decision module is used to determine the resource switching strategy among energy storage regulation, load reduction or peak shifting, and wind and solar curtailment based on the comparison results of the marginal cost of energy storage, the marginal cost of load regulation, the marginal benefit of renewable energy consumption, and the marginal benefit of grid operation. The differentiated power generation restriction allocation module is used to implement differentiated power generation restriction allocation for multiple renewable energy stations based on the contribution of renewable energy stations to the over-limit and the corresponding marginal consumption revenue when there is a risk of exceeding the limit at a power transmission bottleneck or section. The safety constraint correction module is used to perform safety constraint correction on the resource switching strategy and the differentiated power generation limit allocation results to obtain control actions that meet power balance constraints, energy storage operation boundary constraints, load regulation boundary constraints and transmission bottleneck or section safety constraints. The control execution and rolling update module is used to execute control actions for the current scheduling cycle and trigger the rolling repetition of each module based on the updated running data.

9. A multi-dimensional cost-benefit balancing switching control device for high-proportion renewable energy grid connection, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores a computer program thereon, which, when executed by a processor, implements the method of any one of claims 1 to 7.

Images