A strategy optimization method and system for data centers to participate in the electricity and carbon markets

Abstract

This invention provides a strategy optimization method and system for data centers participating in the electricity and carbon markets. The method calculates the data center's electricity demand adjustment range based on energy storage charging and discharging constraints and diesel generator output constraints, combined with the data center's power demand model. It calculates the data center's actual carbon emissions based on the data center's actual total power and green electricity power. Based on the data center's carbon emission quota and actual carbon emissions, it uses a tiered reward-penalty carbon trading model to calculate the data center's carbon trading costs. Based on the carbon trading costs and the clearing price in the spot electricity market, and using the electricity demand adjustment range as the boundary, it optimizes and solves a two-layer optimization model for data center participation in the electricity and carbon markets, obtaining a coordinated electricity and carbon strategy for the data center. In this invention, data center participation in the electricity and carbon markets enables collaborative scheduling of multiple markets for joint optimization, achieving better collaborative optimization results.

Description

Technical Field

[0001] This invention relates to the field of collaborative optimization technology for electricity markets and carbon trading, specifically to a strategy optimization method and system for data centers participating in electricity and carbon markets. Background Technology

[0002] Compared with traditional fragmented resources on the user side, data centers consume a lot of electricity, resulting in a large amount of indirect carbon emissions. On the other hand, the workload of data centers has the characteristics of spatiotemporal transferability, and the power consumption behavior of data centers can be adjusted by delaying offline load processing or transferring online load between data centers. Therefore, in the context of power market reform, the potential and demand for data centers to participate in the electricity market and carbon market are becoming increasingly prominent.

[0003] Existing technologies lack multi-market collaborative optimization. The optimization strategies for data centers participating in the electricity market or carbon market are mostly single-dimensional. For example, some studies reduce electricity costs through load shifting or reduce carbon emissions through carbon trading accounting. The potential for low-carbon collaboration is limited, resulting in poor optimization effects. Summary of the Invention

[0004] To overcome the shortcomings of traditional technologies, such as the lack of multi-market collaborative optimization and the fact that optimization strategies for data centers participating in the electricity or carbon markets are mostly single-dimensional, resulting in poor optimization effects, this invention provides a strategy optimization method for data centers participating in the electricity and carbon markets, comprising: Based on the constraints of energy storage charging and discharging and diesel generator output, combined with the power demand model of the data center, the power demand adjustment range of the data center is calculated. Based on the actual total power and green electricity power of the data center, the actual carbon emissions of the data center are calculated; based on the carbon emission quota of the data center and the actual carbon emissions, a tiered reward and punishment carbon trading model is used to calculate the carbon trading cost of the data center. Based on the carbon trading costs and the clearing price of the spot electricity market, and taking the electricity demand adjustment range as the boundary, the two-layer optimization model for data centers participating in the electricity carbon market is optimized and solved to obtain the data center's electricity carbon coordination strategy. The data center's electricity carbon coordination strategy includes load spatiotemporal transfer schemes, energy storage scheduling strategies, and market pricing strategies.

[0005] Optionally, the power demand model includes the relationship between the total power of the data center, the processing load power of the data center, the energy storage charging and discharging power, and the diesel generator power; the load includes online load and offline load. The power demand model satisfies the following formula:

[0006] in, fort Time zone k Data Center i Total power, for t Time zone k Data Center i The processing load power, for t Time zone k Data Center i Energy storage charging power, for t Time zone k Data Center i The energy storage discharge power, for t Time zone k Data Center i The power of the diesel generator; These represent the unit energy consumption of data center i when handling online and offline loads, respectively. They are respectively t Time zone k Data Center i Handling online and offline workloads; This refers to the data center's operational efficiency factor.

[0007] Optionally, the power demand model is established based on the spatiotemporal regulation characteristics of the load; the load includes online loads and offline loads; The spatiotemporal regulation characteristics of the load include spatial transfer constraints of online loads and time delay constraints of offline loads; The spatial transfer constraint of the online load satisfies the following formula:

[0008] in, for t Time zone k Data Center i The total load processing Let t represent the online load arriving at data center i from region k during time period k. The time delay constraint of the offline load satisfies the following formula:

[0009] in, for t Time zone k Data Center i Offline workload of storage, for t-1 Time zone k Data Center i Offline workload of storage, for t Time zone k data center i The online load being processed, for t Time zone k data center i The offline load being processed satisfy , for t Time zone k data center i The arriving offline load, The deadline for scheduling, For the region k Data Center i The maximum offline load of storage. For the region k Data Center i The offline load limit stored at the scheduling deadline; They are respectively regions k Data Center i The number of servers, service rate, and memory; For the delay limits of the Service Level Agreement; c, d These represent the data volume for online load and offline load, respectively.

[0010] Optionally, the energy storage charge and discharge constraints satisfy the following formula:

[0011] in, , They are respectively t Time zone k data center i The energy storage charging and discharging states; , They are respectively t Time zone k data center i The charging and discharging power of energy storage for t Time zone k data center i The upper limit of the discharge power of energy storage; , For respectively t Time period, t-1 Time zone k data center i The energy storage capacity; , They are respectively regions k data centeri Energy storage charging and discharging efficiency; and for t Time zone k data center i The minimum and maximum values ​​of the energy storage charge capacity; The output constraint of the diesel generator satisfies the following formula:

[0012] in, for t Time zone k data center i The power of the diesel generator, , They are respectively regions k data center i The upper and lower limits of the diesel generator power.

[0013] Optionally, the calculation of the carbon trading cost of the data center, based on the data center's carbon emission quota and actual carbon emissions, using a tiered reward and penalty carbon trading model, includes: If the carbon emission allowance of the data center is less than the actual carbon emission, then based on the carbon emission allowance, the actual carbon emission, and the basic carbon trading price of the data center, the carbon emission excess model in the tiered reward and punishment carbon trading model is used to calculate the carbon emission excess cost of the data center as the carbon trading cost. If the carbon emission allowance is not less than the actual carbon emission, then based on the carbon emission allowance, the actual carbon emission, and the carbon trading base price of the data center, the carbon emission surplus model in the tiered reward and punishment carbon trading model is used to calculate the carbon emission surplus revenue of the data center as the carbon trading cost.

[0014] Optionally, the carbon emission quota is determined based on the sum of the carbon emission quotas of the computing unit, storage unit, and communication unit of the data center; The carbon emission quotas satisfy the following formula:

[0015] in, For the region k data center i Carbon emission allowances; For the region k data center i The carbon emission allowance of the calculation unit; For the region k data center i Carbon emission quotas for storage units; For the region k data center i Carbon emission quotas for communication units; For data center operating efficiency factors; The actual carbon emissions are determined based on the difference between the actual total power of the data center and the green electricity power. The actual carbon emissions satisfy the following formula:

[0016] in, For the region k data center i The actual carbon emissions; As an indirect carbon emission factor for data centers, for t Time zone k data center i The actual total power, for t Time zone k data center i green electricity power, The time period for calculating carbon emissions.

[0017] Optionally, the carbon emission excess model satisfies the following formula:

[0018] in, For the carbon trading costs of data centers, The base price for carbon trading in data centers, The interval length is... For price increases, This represents the actual carbon emissions of the data center. Carbon emission allowances for data centers; The carbon emission surplus model satisfies the following formula:

[0019] in, The carbon trading cost for data centers (in this scenario, the revenue from carbon emission surplus). The base price for carbon trading in data centers, The interval length is... As the reward coefficient, This represents the actual carbon emissions of the data center. Carbon emission allowances for data centers.

[0020] Optionally, the two-layer optimization model includes an upper-layer data center energy consumption decision model and a lower-layer electricity market clearing model; The data center energy consumption decision model includes an optimization objective of minimizing energy costs; the optimization objective satisfies the following formula:

[0021] in, for t Time zone k data center i The cost of purchasing electricity in the electricity market for t Time zone k data center i The cost of generating electricity from a diesel generator. For the region k data center i The cost of carbon trading; The electricity market purchase cost satisfies the following formula:

[0022] in, for t Time zone k data center i The cost of purchasing electricity in the electricity market for t Time zone k data center i Medium- to long-term market-clearing electricity prices for t Time zone k data center i The spot electricity market clearing price, for t Time zone k data center i Electricity volume under signed medium- and long-term contracts for t Time zone k data center i Total demand; The power generation cost of the diesel generator satisfies the following formula:

[0023] in, for t Time zone k data center i The cost of generating electricity from a diesel generator. For the region k data center i The unit power generation cost of diesel generator sets, for tTime zone k data center i The power of the diesel generator; The electricity market clearing model includes a medium- to long-term market model and a spot electricity market model. The medium- to long-term market model minimizes the market clearing cost and deviation assessment cost for data centers in the medium to long term; the medium- to long-term market model satisfies the following formula:

[0024] in, for t Time zone k data center i The medium- to long-term market clearing costs, for t Time zone k data center i The spot electricity market clearing price, for t Time zone k data center i Electricity volume under signed medium- and long-term contracts Costs for deviation assessment; for t Time-of-use contract The amount of electricity, , They are respectively t Time-of-use contract The lower and upper limits of battery capacity; for t Time zone k data center i Total demand The deviation penalty coefficient is T, where T is the number of time periods. The spot electricity market model minimizes the difference between the application cost of generating units and the application cost of data centers in the spot market; The spot electricity market model satisfies the following formula:

[0025] in, This indicates the declared cost of the generating unit in the spot market. For the units in the current market j The first application s Price range In order to be in the market recently t Time-of-use units j No. s Duan made great efforts, This indicates the declared cost of data centers in the spot market. For the first data center application in the market recently Price range In order to be in the market recently t Time period data center Electricity purchased in segments, where T represents the number of time segments. Number of generating units The number of power sections of the unit. This refers to the number of power bands in the data center. In order to be in the market recently t Other users during the same period k Electricity purchase segment For the remaining users' power segment numbers; This is the upper limit of the generating capacity of the unit; , The lower and upper limits for electricity purchases for data centers.

[0026] Optionally, based on the carbon trading cost and the clearing price of the spot electricity market, and using the electricity demand adjustment range as the boundary, the two-level optimization model for data center participation in the electricity carbon market is optimized and solved to obtain the data center's electricity carbon coordination strategy, including: Transform the two-layer optimization model for data center participation in the electricity carbon market into a single-layer optimization model; Based on the carbon trading costs and the clearing price of the spot electricity market, and using the electricity demand adjustment range as the boundary, the single-layer optimization model is solved to obtain the data center's electricity-carbon coordination strategy.

[0027] Optionally, the single-layer optimization model satisfies the following formula:

[0028] Where F() represents the function of a single-layer optimization model, x For the decision vector of the data center; y, z These are variables related to the clearing issues in the medium- and long-term electricity market and the spot electricity market, respectively. G ( ) H ( ) represents a constraint on collaborative decision-making in data centers across multiple market types; 、 It serves as the dual variable for the constraints in the medium- and long-term electricity market clearing problem; 、 The dual variable for the constraints in the joint electricity market clearing problem; , The constraint set corresponding to the KKT system for medium- and long-term electricity market and joint market clearing issues.

[0029] On the other hand, the present invention also provides a strategy optimization system for data centers participating in the electricity and carbon markets, comprising: The adjustment range determination module is used to calculate the power demand adjustment range of the data center based on the energy storage charging and discharging constraints and the diesel generator output constraints, combined with the power demand model of the data center. The carbon trading cost determination module is used to calculate the actual carbon emissions of the data center based on the actual total power and green electricity power of the data center; and to calculate the carbon trading cost of the data center based on the carbon emission quota of the data center and the actual carbon emissions using a tiered reward and punishment carbon trading model. The two-layer model collaborative optimization module is used to optimize and solve the two-layer optimization model for data centers participating in the electricity carbon market based on the carbon trading cost and the clearing price of the spot electricity energy market, with the electricity demand adjustment range as the boundary, so as to obtain the data center's electricity carbon collaborative strategy. The data center's electricity carbon collaborative strategy includes load spatiotemporal transfer scheme, energy storage scheduling strategy and market bidding strategy.

[0030] Optionally, the power demand model includes the relationship between the total power of the data center, the processing load power of the data center, the energy storage charging and discharging power, and the diesel generator power; the load includes online load and offline load. The power demand model satisfies the following formula:

[0031] in, for t Time zone k Data Center i Total power, for t Time zone k Data Center i The processing load power, for t Time zone k Data Center i Energy storage charging power, for t Time zone k Data Center i The energy storage discharge power, for t Time zone k Data Center i The power of the diesel generator; These represent the unit energy consumption of data center i when handling online and offline loads, respectively. They are respectively t Time zone k Data Center i Handling online and offline workloads; This refers to the data center's operational efficiency factor.

[0032] Optionally, the power demand model is established based on the spatiotemporal regulation characteristics of the load; the load includes online loads and offline loads; The spatiotemporal regulation characteristics of the load include spatial transfer constraints of online loads and time delay constraints of offline loads; The spatial transfer constraint of the online load satisfies the following formula:

[0033] in, for t Time zone k Data Center i The total load processing Let t represent the online load arriving at data center i from region k during time period k. The time delay constraint of the offline load satisfies the following formula:

[0034] in, for t Time zone k Data Center i Offline workload of storage, for t-1 Time zone k Data Center i Offline workload of storage, for t Time zone k data center i The online load being processed, for t Time zone k data center i The offline load being processed satisfy , for t Time zone k data center i The arriving offline load, The deadline for scheduling, For the region k Data Center i The maximum offline load of storage. For the region k Data Center i The offline load limit stored at the scheduling deadline; They are respectively regions k Data Center i The number of servers, service rate, and memory; For the delay limits of the Service Level Agreement; c, d These represent the data volume for online load and offline load, respectively.

[0035] Optionally, the energy storage charge and discharge constraints satisfy the following formula:

[0036] in, , They are respectively t Time zone k data center i The energy storage charging and discharging states; , They are respectively t Time zone k data center i The charging and discharging power of energy storage for t Time zone k data center i The upper limit of the discharge power of energy storage; , For respectively t Time period, t-1 Time zone k data center i The energy storage capacity; , They are respectively regions k data center i Energy storage charging and discharging efficiency; and for t Time zone k data center i The minimum and maximum values ​​of the energy storage charge capacity; The output constraint of the diesel generator satisfies the following formula:

[0037] in, for t Time zone k data center i The power of the diesel generator, , They are respectively regions k data center i The upper and lower limits of the diesel generator power.

[0038] Optionally, the carbon trading cost determination module is specifically used for: If the carbon emission allowance of the data center is less than the actual carbon emission, then based on the carbon emission allowance, the actual carbon emission, and the basic carbon trading price of the data center, the carbon emission excess model in the tiered reward and punishment carbon trading model is used to calculate the carbon emission excess cost of the data center as the carbon trading cost. If the carbon emission allowance is not less than the actual carbon emission, then based on the carbon emission allowance, the actual carbon emission, and the carbon trading base price of the data center, the carbon emission surplus model in the tiered reward and punishment carbon trading model is used to calculate the carbon emission surplus revenue of the data center as the carbon trading cost.

[0039] Optionally, the carbon emission quota is determined based on the sum of the carbon emission quotas of the computing unit, storage unit, and communication unit of the data center; The carbon emission quotas satisfy the following formula:

[0040] in, For the region k data center i Carbon emission allowances; For the region k data center i The carbon emission allowance of the calculation unit; For the region k data center i Carbon emission quotas for storage units; For the region k data center i Carbon emission quotas for communication units; For data center operating efficiency factors; The actual carbon emissions are determined based on the difference between the actual total power of the data center and the green electricity power. The actual carbon emissions satisfy the following formula:

[0041] in, For the region k data center i The actual carbon emissions; As an indirect carbon emission factor for data centers, for t Time zone k data center i The actual total power, for t Time zone k data center i green electricity power, The time period for calculating carbon emissions.

[0042] Optionally, the carbon emission excess model satisfies the following formula:

[0043] in, For the carbon trading costs of data centers, The base price for carbon trading in data centers, The interval length is... For price increases, This represents the actual carbon emissions of the data center. Carbon emission allowances for data centers; The carbon emission surplus model satisfies the following formula:

[0044] in, The carbon trading cost for data centers (in this scenario, the revenue from carbon emission surplus). The base price for carbon trading in data centers, The interval length is... As the reward coefficient, This represents the actual carbon emissions of the data center. Carbon emission allowances for data centers.

[0045] Optionally, the two-layer optimization model includes an upper-layer data center energy consumption decision model and a lower-layer electricity market clearing model; The data center energy consumption decision model includes an optimization objective of minimizing energy costs; the optimization objective satisfies the following formula:

[0046] in, for t Time zone k data center i The cost of purchasing electricity in the electricity market for t Time zone k data center i The cost of generating electricity from a diesel generator. For the region k data center i The cost of carbon trading; The electricity market purchase cost satisfies the following formula:

[0047] in, for t Time zone k data center iThe cost of purchasing electricity in the electricity market for t Time zone k data center i Medium- to long-term market-clearing electricity prices for t Time zone k data center i The spot electricity market clearing price, for t Time zone k data center i Electricity volume under signed medium- and long-term contracts for t Time zone k data center i Total demand; The power generation cost of the diesel generator satisfies the following formula:

[0048] in, for t Time zone k data center i The cost of generating electricity from a diesel generator. For the region k data center i The unit power generation cost of diesel generator sets, for t Time zone k data center i The power of the diesel generator; The electricity market clearing model includes a medium- to long-term market model and a spot electricity market model. The medium- to long-term market model minimizes the market clearing cost and deviation assessment cost for data centers in the medium to long term; the medium- to long-term market model satisfies the following formula:

[0049] in, for t Time zone k data center i The medium- to long-term market clearing costs, for t Time zone k data center i The spot electricity market clearing price, for t Time zone k data center i Electricity volume under signed medium- and long-term contracts Costs for deviation assessment; for t Time-of-use contract The amount of electricity, , They are respectively t Time-of-use contract The lower and upper limits of battery capacity; for t Time zone k data center i Total demand The deviation penalty coefficient is T, where T is the number of time periods. The spot electricity market model minimizes the difference between the application cost of generating units and the application cost of data centers in the spot market; The spot electricity market model satisfies the following formula:

[0050] in, This indicates the declared cost of the generating unit in the spot market. For the units in the current market j The first application s Price range In order to be in the market recently t Time-of-use units j No. s Duan made great efforts, This indicates the declared cost of data centers in the spot market. For the first data center application in the market recently Price range In order to be in the market recently t Time period data center Electricity purchased in segments, where T represents the number of time segments. Number of generating units The number of power sections of the unit. This refers to the number of power bands in the data center. In order to be in the market recently t Other users during the same period k Electricity purchase segment For the remaining users' power segment numbers; This is the upper limit of the generating capacity of the unit; , The lower and upper limits for electricity purchases for data centers.

[0051] Optionally, the two-layer model collaborative optimization module includes: The model conversion unit is used to convert the two-layer optimization model for data centers participating in the electricity carbon market into a single-layer optimization model. An optimization solution unit is used to solve the single-layer optimization model based on the carbon trading cost and the spot electricity market clearing price, with the electricity demand adjustment range as the boundary, to obtain the data center's electricity-carbon coordination strategy.

[0052] Optionally, the single-layer optimization model satisfies the following formula:

[0053] Where F() represents the function of a single-layer optimization model, x For the decision vector of the data center; y, z These are variables related to the clearing issues in the medium- and long-term electricity market and the spot electricity market, respectively. G ( ) H ( ) represents a constraint on collaborative decision-making in data centers across multiple market types; 、 It serves as the dual variable for the constraints in the medium- and long-term electricity market clearing problem; 、 The dual variable for the constraints in the joint electricity market clearing problem; , The constraint set corresponding to the KKT system for medium- and long-term electricity market and joint market clearing issues.

[0054] On the other hand, the present invention also provides a computer device, comprising: one or more processors; The processor is used to store one or more programs; When the one or more programs are executed by the one or more processors, the strategy optimization method for data centers to participate in the electricity and carbon markets described in any one of the above-mentioned methods is implemented.

[0055] On the other hand, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed, it implements the strategy optimization method for data centers participating in the electricity and carbon markets as described in any one of the above.

[0056] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention provides a strategy optimization method and system for data centers participating in the electricity and carbon markets. The method calculates the data center's electricity demand adjustment range based on energy storage charging and discharging constraints and diesel generator output constraints, combined with the data center's power demand model. It calculates the data center's actual carbon emissions based on the data center's actual total power and green electricity power. Based on the data center's carbon emission quotas and actual carbon emissions, it uses a tiered reward-penalty carbon trading model to calculate the data center's carbon trading costs. Based on the carbon trading costs and the clearing price in the spot electricity market, and using the electricity demand adjustment range as the boundary, it optimizes a two-layer optimization model for data center participation in the electricity and carbon markets, obtaining a data center's electricity-carbon coordination strategy. This strategy includes load spatiotemporal transfer schemes, energy storage scheduling strategies, and market pricing strategies. This invention, by considering the characteristics of power sources such as energy storage and diesel generators, the data center's electricity demand characteristics, and combining a tiered reward-penalty carbon trading model, enables collaborative optimization of multiple markets, stimulating the synergistic potential of flexible resources and low-carbon technologies. This can improve the carbon emission incentive effect, reduce carbon emissions, and achieve better collaborative optimization results. Attached Figure Description

[0057] Figure 1 This is a flowchart illustrating the strategy optimization method for data centers participating in the electricity and carbon markets according to the present invention. Figure 2 This is a schematic diagram illustrating the decision-making process of a data center participating in the electricity spot market according to the present invention; Figure 3 This is a schematic diagram of the online load scheduling of the regional center 1 according to the present invention; Figure 4 This is a schematic diagram of the online load scheduling of the regional center 2 according to the present invention; Figure 5 This is a schematic diagram of the online load scheduling of the regional center 3 according to the present invention; Figure 6 This is a schematic diagram of the system architecture for optimizing data center participation in the electricity and carbon markets according to the present invention. Figure 7 This is a schematic diagram of the electronic device of the present invention. Detailed Implementation

[0058] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0059] Example 1: This invention provides a strategy optimization method for data centers participating in the electricity and carbon markets, such as... Figure 1 As shown, it includes: Step 101: Based on the energy storage charging and discharging constraints and the diesel generator output constraints, combined with the power demand model of the data center, calculate the power demand adjustment range of the data center. Step 102: Based on the actual total power and green electricity power of the data center, calculate the actual carbon emissions of the data center; based on the carbon emission quota and actual carbon emissions of the data center, use a tiered reward and punishment carbon trading model to calculate the carbon trading cost of the data center. Step 103: Based on carbon trading costs and spot electricity market clearing prices, and taking the electricity demand adjustment range as the boundary, optimize the two-layer optimization model for data centers participating in the electricity carbon market to obtain the data center's electricity carbon coordination strategy. The data center's electricity carbon coordination strategy includes load spatiotemporal transfer schemes, energy storage scheduling strategies, and market pricing strategies.

[0060] This invention proposes a process for determining a joint trading strategy for data centers participating in the electricity-carbon trading market. By involving data centers in the electricity-carbon market, considering the characteristics of power sources such as energy storage and diesel generators, the electricity demand characteristics of data centers, and combining a tiered reward-based carbon trading model, it is possible to coordinate and optimize the joint operation of multiple markets, stimulate the synergistic potential of flexible resources and low-carbon technologies, improve the carbon emission incentive effect, reduce carbon emissions, and achieve better synergistic optimization results.

[0061] The power demand model in step 101 above includes the relationship between the total power of the data center, the processing load power of the data center, the energy storage charging and discharging power, and the diesel generator power. The load includes online loads and offline loads; online loads are generally real-time tasks, while offline loads are generally deferred tasks. For example, the power demand model satisfies the following formula:

[0062] in, for t Time zone k Data Center i Total power, for t Time zone k Data Center i The processing load power, for t Time zone k Data Center i Energy storage charging power, for t Time zone k Data Center i The energy storage discharge power, for t Time zone k Data Center iThe power of the diesel generator; These represent the unit energy consumption of data center i when handling online and offline loads, respectively. They are respectively t Time zone k Data Center i Handling online and offline workloads; This refers to the data center's operational efficiency factor. It can be the energy consumption ratio of the data center to the IT (Information Technology) equipment in the data center. The closer the energy consumption ratio is to 1, the higher the operating efficiency of the data center.

[0063] In one possible implementation, the power demand model described above can be established based on the spatiotemporal adjustability of the load (including online and offline loads). The spatiotemporal adjustability of the load includes spatial transfer constraints for online loads and time delay constraints for offline loads; the spatial transfer constraint for online loads indicates that the total load processing capacity of the data center equals the total load arrivals. For example, the spatial transfer constraint for online loads satisfies the following formula:

[0064] in, for t Time zone k Data Center i The total load processing Let t represent the online load arriving at data center i from region k during time period k. The time latency constraint for offline workloads means that the cumulative latency of the offline workloads must not exceed the data center's storage limit and must be completed before the scheduling deadline. For example, the time latency constraint for offline workloads satisfies the following formula:

[0065] in, for t Time zone k Data Center i Offline workload of storage, for t-1 Time zone k Data Center i Offline workload of storage, for t Time zone k data center i The online load being processed, for t Time zone k data center i The offline load being processed satisfy , for t Time zone k data center i The arriving offline load, The deadline for scheduling, For the region k Data Center i The maximum offline load of storage. For the region k Data Center i The offline load limit stored at the scheduling deadline; They are respectively regions k Data Center i The number of servers, service rate, and memory; For the delay limits of the Service Level Agreement; c, d These represent the data volume for online load and offline load, respectively.

[0066] The power demand model is used to quantify the adjustment potential of data centers. In step 101 above, based on the spatial transfer constraints of online loads and the time delay constraints of offline loads, combined with the energy storage charging and discharging constraints and diesel generator output limits, the upper and lower limits of the data center's power demand are quantified to determine its adjustment range, thus obtaining the data center's power demand adjustment range. .in, , They are respectively t Time zone k data center i The lower and upper limits for adjusting electricity demand.

[0067] The example energy storage charge and discharge constraints include energy storage charge and discharge power constraints and state of charge boundaries, such as satisfying the following formula:

[0068] in, , They are respectively t Time zone k data center i The energy storage charging and discharging states; , They are respectively t Time zone k data center i The charging and discharging power of energy storage for t Time zone k data center i The upper limit of the discharge power of energy storage; , For respectively t Time period,t-1 Time zone k data center i The energy storage capacity; , They are respectively regions k data center i Energy storage charging and discharging efficiency; and for t Time zone k data center i The minimum and maximum values ​​of the energy storage charge capacity; For example, the output constraint of a diesel generator satisfies the following formula:

[0069] in, for t Time zone k data center i The power of the diesel generator, , They are respectively regions k data center i The upper and lower limits of the diesel generator power.

[0070] In step 102 above, free carbon emission allowances are allocated based on the service volume indicators (such as, but not limited to, number of instructions, data interaction volume, and throughput) of the data center's computing units, storage units, and communication units. A tiered reward and punishment carbon trading model is adopted to dynamically calculate the cost / benefit of carbon emission excess or surplus in order to incentivize proactive emission reduction.

[0071] When calculating carbon emission allowances, data throughput can be used to measure the activity level of communication units, setting baselines for the service volume of different units, refining allowance allocation, and defining carbon emission factors and energy efficiency values ​​to determine free carbon emission allowances. For example, carbon emission allowances are determined based on the sum of carbon emission allowances for the computing units, storage units, and communication units of a data center, for instance, by formulating the following:

[0072] in, For the region k data center i Carbon emission allowances; For the region k data center i The carbon emission allowance of the calculation unit; For the region k data center i Carbon emission quotas for storage units; For the region kdata center i Carbon emission quotas for communication units; This refers to the data center's operational efficiency factor.

[0073] Actual carbon emissions can be calculated based on the difference between the actual total power of the data center and the green electricity output. The actual total power of the data center can be collected data, while the green electricity output can be collected data or provided by the power dispatching department. For example, the actual carbon emissions calculation can satisfy the following formula:

[0074] in, For the region k data center i The actual carbon emissions; As an indirect carbon emission factor for data centers, for t Time zone k data center i The actual total power, for t Time zone k data center i green electricity power, The time period for calculating carbon emissions.

[0075] The tiered reward and penalty carbon trading model includes a carbon emission excess model and a carbon emission surplus model. In one implementation, in step 102 above, based on the data center's carbon emission quotas and actual carbon emissions, when calculating the data center's carbon trading costs using the tiered reward and penalty carbon trading model, if the data center's carbon emission quota is less than its actual carbon emissions, then based on the carbon emission quota, actual carbon emissions, and the data center's base carbon trading price, the carbon emission excess model within the tiered reward and penalty carbon trading model is used to calculate the data center's carbon emission excess cost, which is then used as the carbon trading cost. If the carbon emission quota is not less than the actual carbon emissions, then based on the carbon emission quota, actual carbon emissions, and the data center's base carbon trading price, the carbon emission surplus model within the tiered reward and penalty carbon trading model is used to calculate the data center's carbon emission surplus revenue, which is then used as the carbon trading cost. In this implementation, if the data center's total carbon emission quota is less than its actual carbon emissions, it falls under the carbon emission excess scenario; otherwise, it falls under the carbon emission surplus scenario.

[0076] In the carbon emission excess model, when carbon emissions exceed the limit, the cost increases in stages with the amount of excess, for example, satisfying the following formula:

[0077] in, The carbon trading cost for data centers (in this scenario, the cost of exceeding carbon emission limits). The base price for carbon trading in data centers, The interval length is... For price increases, This represents the actual carbon emissions of the data center. Carbon emission allowances for data centers; In the carbon emission surplus model, when there is a carbon emission surplus, the revenue decreases in stages as the surplus increases, for example, satisfying the following formula:

[0078] in, The carbon trading cost for data centers (in this scenario, the revenue from carbon emission surplus). The base price for carbon trading in data centers, The interval length is... As the reward coefficient, This represents the actual carbon emissions of the data center. Carbon emission allowances for data centers.

[0079] The two-layer optimization model for data centers participating in the electricity carbon market in step 103 above includes an upper-layer data center energy consumption decision model and a lower-layer electricity market clearing model.

[0080] In one implementation, the upper-level data center energy consumption decision model includes an optimization objective of minimizing energy costs. The objective function is a joint optimization of electricity market purchase costs (e.g., electricity purchase costs in the power energy market), diesel generation costs (specifically, diesel generator generation costs), carbon trading costs, and ancillary service market revenues. Constraints include load shifting, energy storage charging and discharging, reserve capacity, and market bidding boundaries. For example, this optimization objective satisfies the following formula:

[0081] in, for t Time zone k data center i The cost of purchasing electricity in the electricity market for t Time zone k data center i The cost of generating electricity from a diesel generator. For the region k data center i The cost of carbon trading; The cost of purchasing electricity in the electricity market is related to the clearing price in the spot electricity market, the clearing price in the medium- and long-term (electricity) market, the volume of medium- and long-term contracts, and total demand. The clearing price in the spot electricity market can be the day-ahead clearing price or the real-time clearing price obtained from spot electricity market transactions. The clearing price in the medium- and long-term market can be the price during the contractually stipulated performance period. For example, the cost of purchasing electricity in the electricity market satisfies the following formula:

[0082] in, for t Time zone k data center i The cost of purchasing electricity in the electricity market for t Time zone k data center i The spot electricity market clearing price, for t Time zone k data center i Medium- to long-term market-clearing electricity prices for t Time zone k data center i Electricity volume under signed medium- and long-term contracts for t Time zone k data center i Total demand; The cost of generating electricity from a diesel generator is related to both the unit cost of generating electricity and the power output. The unit cost of generating electricity is the cost of generating electricity per unit time, that is, the cost incurred by the diesel generator per unit time when it operates at a given power level. For example, the cost of generating electricity from a diesel generator satisfies the following formula:

[0083] in, for t Time zone k data center i The cost of generating electricity from a diesel generator. For the region k data center i The unit power generation cost of diesel generator sets, for t Time zone k data center i The power of the diesel generator.

[0084] The constraints on load shifting, energy storage charging and discharging, and reserve capacity in the upper-level data center energy consumption decision model can be found in the constraints of the power demand model mentioned above; the constraint formula for the market price boundary satisfies the following formula:

[0085] in, For the region k data center i Medium and long-term market quotations, For the region k data center i The lower limit of medium- and long-term market quotations. For the region k data center i The upper limit of medium- and long-term market quotations, For the region k data center i Spot electricity market quotes, For the region k data center i The lower limit of the spot electricity market price. For the region k data center i The upper limit of spot electricity market price.

[0086] The lower-level electricity market clearing model includes a medium-to-long-term market model and a spot electricity market model. This lower-level model is an electricity market clearing model, targeting both the spot electricity market and the medium-to-long-term electricity market, with the goal of minimizing total regional costs. This goal can achieve the objective of maximizing social welfare. Constraints in the spot electricity market include power balance, line transmission capacity, and bidding boundaries for market participants.

[0087] In the medium- to long-term market model, data centers participate in the market by signing medium- to long-term electricity price contracts. This includes minimizing the market clearing costs and deviation assessment costs for data centers in the medium- to long-term market. Constraints include the contracted electricity volume range and deviation penalty mechanisms, such as satisfying the following formula:

[0088] in, for t Time zone k data center i The medium- to long-term market clearing costs, for t Time zone k data center i The spot electricity market clearing price, for t Time zone k data centeri Electricity volume under signed medium- and long-term contracts Costs for deviation assessment; for t Time-of-use contract The amount of electricity, , They are respectively t Time-of-use contract The lower and upper limits of battery capacity; for t Time zone k data center i Total demand Here, T represents the deviation penalty coefficient, and T represents the number of time periods.

[0089] In the spot energy market, it is assumed that generating units and data centers participate by submitting bids, while other users act as price takers. The spot energy market model aims to minimize the difference between the bidding costs of generating units and data centers. Constraints include power balance, line transmission capacity, and market participant bidding boundaries, for example, satisfying the following formula:

[0090] in, This indicates the declared cost of the generating unit in the spot market. For the units in the current market j The first application s Price range In order to be in the market recently t Time-of-use units j No. s Duan made great efforts, This indicates the declared cost of data centers in the spot market. For the first data center application in the market recently Price range In order to be in the market recently t Time period data center Electricity purchased in segments, where T represents the number of time segments. Number of generating units The number of power sections of the unit. This refers to the number of power bands in the data center. In order to be in the market recently t Other users during the same period k Electricity purchase segment For the remaining users' power segment numbers; This is the upper limit of the generating capacity of the unit; , The lower and upper limits for electricity purchases for data centers.

[0091] In one implementation, step 103 above can transform the two-layer optimization model for data centers participating in the electricity carbon market into a single-layer optimization model. Based on carbon trading costs and the clearing price of the spot electricity market, and using the electricity demand adjustment range as the boundary, the single-layer optimization model is solved to obtain the data center's electricity carbon coordination strategy. For example, but not limited to, KKT conditions (Kuhn-Tucker conditions) can be used to transform the objective function in the two-layer optimization model into corresponding constraints, thereby transforming the two-layer optimization model into a single-layer optimization model. The single-layer optimization model can satisfy the following formula:

[0092] Where F() represents the function of a single-layer optimization model, x For the decision vector of the data center; y, z These are variables related to the clearing issues in the medium- and long-term electricity market and the spot electricity market, respectively. G ( ) H ( ) represents a constraint on collaborative decision-making in data centers across multiple market types; 、 It serves as the dual variable for the constraints in the medium- and long-term electricity market clearing problem; 、 The dual variable for the constraints in the joint electricity market clearing problem; , The constraint set corresponding to the KKT system for medium- and long-term electricity market and joint market clearing issues.

[0093] In step 103 above, the optimized solution yields a load spatiotemporal transfer scheme, an energy storage scheduling strategy, and a market pricing strategy. Figure 2 Taking the decision-making process of a data center participating in the electricity spot market as an example, it includes the input stage, the optimization decision stage, the market interaction stage, and the feedback response stage.

[0094] Input phase: Input task queues, energy consumption data, medium and long-term power purchase agreements, etc. into the decision center.

[0095] Optimization Decision-Making Phase: The decision-making center combines the power demand model, energy storage system characteristics, green electricity access, carbon emission quotas and trading rules to construct and solve a two-layer optimization model, outputting load scheduling plans, energy storage actions, and market pricing strategies.

[0096] Market interaction phase: Submit day-ahead market quotes to the day-ahead market; dynamically adjust prices based on actual operating conditions during the real-time phase and quote prices in the real-time market; simultaneously send load scheduling decisions to the data center for execution.

[0097] Feedback and Response Phase: After the data center executes the scheduling, it feeds back the actual operating results (such as actual power and green electricity ratio) to the decision center; this is used to update carbon emission calculations, correct deviations, and support the next round of optimization (closed-loop control).

[0098] The method provided in this invention embodiment can verify the effectiveness of the strategy through multi-scenario simulation: Assuming we consider data centers located in three different regions, Data Center 1 is located in a region with tight power supply and high electricity prices; Data Center 2 is located in a region with a higher proportion of renewable energy and generally lower electricity prices, with a clear "floor price" and price peaks; Data Center 3 is located in a region with relatively balanced power supply and demand and moderate electricity prices. Multi-scenario simulations include: Scenario 1: Three data centers participate only in the medium to long term market, and their workloads are not transferable; Scenario 2: Data Center 1 only participates in the medium- to long-term market, while Data Centers 2 and 3 only participate in the day-to-day spot market, and their workloads are non-transferable; Scenario 3: Data Center 1 only participates in the medium- to long-term market, while Data Centers 2 and 3 only participate in the day-to-day spot market, and their workloads are transferable; Scenario 4: Data centers 1, 2, and 3 simultaneously participate in the medium- to long-term and day-ahead spot markets, and their workloads are transferable.

[0099] The method provided in this embodiment of the invention calculates the adjustable load range of the data center, uses this adjustable load range as the boundary of the data center's load in the two-layer model, analyzes the changes in the data center's revenue expenditure and energy costs under different scenarios, and thus verifies the superiority of this method.

[0100] Table 1 below shows the revenue distribution of data centers in various scenarios. Comparing scenarios 1 and 2, it can be seen that after participating in the spot market, the energy cost of data centers decreases by approximately 24%. Comparing scenarios 2 and 3, it can be seen that when considering the ability to transfer workloads, the workload can be transferred to nodes and time periods with lower electricity prices, thereby reducing the cost of purchasing electricity. Comparing scenarios 3 and 4, it can be seen that data centers can participate in both medium- and long-term and spot electricity markets simultaneously, generating revenue through electricity sales or demand response. Combined with the ability to transfer workloads, the selection of electricity purchase time periods is optimized, avoiding electricity consumption during high-price periods, further reducing energy costs.

[0101] Table 1

[0102] Table 2 below shows the carbon emissions and green electricity consumption of data centers in various scenarios. Comparing scenarios 1 and 2, it can be seen that the green electricity consumption ratio increases, but carbon emissions actually rise. This is because long-term contracts are tied to green electricity, and the non-transferable load leads to the passive matching of high-carbon power sources during the electricity consumption period in the spot market. Comparing scenarios 2 and 3, it can be seen that with the participation of the spot market and the transferability of the load, the data center can flexibly schedule to avoid high-carbon periods and prioritize matching green electricity output windows, resulting in reduced carbon emissions and an increased green electricity consumption ratio. Comparing scenarios 3 and 4, carbon emissions are further reduced by 7.8%. Data center 1 reduces its dependence on regional high-carbon electricity by participating in the spot market. Overall, scenario 4 performs best in terms of economic and social benefits, but it also poses challenges to data center load transfer technology and cross-regional collaborative optimization capabilities.

[0103] Table 2

[0104] Figure 3 , Figure 4 and Figure 5 The online load scheduling of Regional Center 1, Regional Center 2, and Regional Center 3 are shown respectively. Changes in data center power consumption behavior depend on workload transfer. Analyzing the workload scheduling of the three data centers in Scenario 4, it can be concluded that when there is a significant difference in electricity prices across regions, the scheduling of online load among data centers is mainly determined by the relative size of the day-ahead market clearing price in their respective regions. For example, during periods 1 to 18, the electricity price in the region where interconnected data center 1 is located is much higher than that of the other two data centers. Therefore, the front-end server will transfer part or even all of the online load destined for data center 1 to data centers 2 and 3 for processing. Thus, it can be seen that in this embodiment of the invention, the adjustable space of data center power consumption is accurately quantified and assessed based on the power demand of the data center, and a power demand model is constructed; a model for data center participation in the electricity-carbon market is constructed to calculate carbon emission costs; and a two-layer optimization model for data center participation in the electricity-carbon market is constructed to coordinate the scheduling of data center loads, energy storage, and other power sources, achieving multi-market joint optimization, reducing energy costs, reducing carbon emissions, and improving the renewable energy absorption rate. Through simulation examples, it is verified that this embodiment of the invention can reduce data center energy costs and improve environmental benefits.

[0105] Traditional technologies lack multi-market collaborative optimization, and optimization strategies for data centers participating in the electricity or carbon markets are mostly single-dimensional. For example, some studies reduce electricity costs through load shifting or reduce carbon emissions through carbon trading accounting. The quantification of adjustment potential is insufficient, failing to accurately assess the spatiotemporal adjustability of data center loads and the flexibility of power sources such as energy storage, resulting in limited optimization space. The incentive effect of carbon emissions is limited; traditional fixed carbon price models are insufficient to dynamically incentivize data centers to proactively reduce emissions. There is an urgent need to research new carbon trading mechanisms to unleash the synergistic potential of flexible resources and low-carbon technologies. This paper combines a reward-penalty tiered carbon trading model and utilizes the spatiotemporal transferability of loads to construct a two-layer optimized operation model for data centers in a multi-market coupling environment of electricity and carbon trading. Through collaborative optimization of data center load handling arrangements and energy storage charging and discharging, the overall energy cost of data centers is reduced by approximately 56%, carbon emissions by approximately 15%, and photovoltaic consumption by approximately 20%, achieving the dual goals of improving economic and environmental benefits.

[0106] Example 2: Based on the same inventive concept, this invention also provides a strategy optimization system for data centers participating in the electricity and carbon markets, as shown in the schematic diagram below. Figure 6 As shown, it includes: The adjustment range determination module is used to calculate the power demand adjustment range of the data center based on the energy storage charging and discharging constraints and the diesel generator output constraints, combined with the power demand model of the data center. The carbon trading cost determination module is used to calculate the actual carbon emissions of data centers based on the actual total power and green electricity power obtained from the data centers; and to calculate the carbon trading costs of data centers based on the carbon emission quotas and actual carbon emissions of the data centers using a tiered reward and punishment carbon trading model. The two-layer model collaborative optimization module is used to optimize and solve the two-layer optimization model for data centers participating in the electricity carbon market based on carbon trading costs and spot electricity market clearing prices, with the electricity demand adjustment range as the boundary. The result is the data center's electricity carbon collaborative strategy, which includes load spatiotemporal transfer schemes, energy storage scheduling strategies, and market pricing strategies.

[0107] In one specific implementation, the power demand model includes the relationship between the total power of the data center, the processing load power of the data center, the energy storage charging and discharging power, and the diesel generator power; the load includes online load and offline load. The power demand model satisfies the following formula:

[0108] in, for t Time zone k Data Center iTotal power, for t Time zone k Data Center i The processing load power, for t Time zone k Data Center i Energy storage charging power, for t Time zone k Data Center i The energy storage discharge power, for t Time zone k Data Center i The power of the diesel generator; These represent the unit energy consumption of data center i when handling online and offline loads, respectively. They are respectively t Time zone k Data Center i Handling online and offline workloads; This refers to the data center's operational efficiency factor.

[0109] In one specific implementation, the power demand model is established based on the spatiotemporal regulation characteristics of the load; the load includes online loads and offline loads; The spatiotemporal regulation characteristics of loads include spatial transfer constraints for online loads and time delay constraints for offline loads; The spatial transfer constraints of online loads satisfy the following formula:

[0110] in, for t Time zone k Data Center i The total load processing Let t represent the online load arriving at data center i from region k during time period k. The time delay constraint for offline loads satisfies the following formula:

[0111] in, for t Time zone k Data Center i Offline workload of storage, for t-1 Time zone k Data Center i Offline workload of storage, for t Time zone kdata center i The online load being processed, for t Time zone k data center i The offline load being processed satisfy , for t Time zone k data center i The arriving offline load, The deadline for scheduling, For the region k Data Center i The maximum offline load of storage. For the region k Data Center i The offline load limit stored at the scheduling deadline; They are respectively regions k Data Center i The number of servers, service rate, and memory; For the delay limits of the Service Level Agreement; c, d These represent the data volume for online load and offline load, respectively.

[0112] In one specific implementation, the energy storage charging and discharging constraints satisfy the following formula:

[0113] in, , They are respectively t Time zone k data center i The energy storage charging and discharging states; , They are respectively t Time zone k data center i The charging and discharging power of energy storage for t Time zone k data center i The upper limit of the discharge power of energy storage; , For respectively t Time period, t-1 Time zone k data center i The energy storage capacity; , They are respectively regions k data center i Energy storage charging and discharging efficiency; and for t Time zone k data center i The minimum and maximum values ​​of the energy storage charge capacity; The output constraint of the diesel generator satisfies the following formula:

[0114] in, for t Time zone k data center i The power of the diesel generator, , They are respectively regions k data center i The upper and lower limits of the diesel generator power.

[0115] In one specific implementation, the carbon trading cost determination module is specifically used for: If the carbon emission allowance of the data center is less than the actual carbon emission, the carbon emission excess cost of the data center is calculated based on the carbon emission allowance, the actual carbon emission, and the basic carbon trading price of the data center, using the carbon emission excess model in the tiered reward and punishment carbon trading model, and is used as the carbon trading cost. If the carbon emission allowance is not less than the actual carbon emission, then based on the carbon emission allowance, the actual carbon emission, and the basic carbon trading price of the data center, the carbon emission surplus model in the tiered reward and punishment carbon trading model is used to calculate the carbon emission surplus revenue of the data center as the carbon trading cost.

[0116] In one specific implementation, carbon emission quotas are determined based on the sum of carbon emission quotas for the computing units, storage units, and communication units of the data center; Carbon emission quotas satisfy the following formula:

[0117] in, For the region k data center i Carbon emission allowances; For the region k data center i The carbon emission allowance of the calculation unit; For the region k data center i Carbon emission quotas for storage units; For the region k data center i Carbon emission quotas for communication units; For data center operating efficiency factors; Actual carbon emissions are determined based on the difference between the actual total power of the data center and the green electricity power. Actual carbon emissions satisfy the following formula:

[0118] in, For the region k data center i The actual carbon emissions; As an indirect carbon emission factor for data centers, for t Time zone k data center i The actual total power, for t Time zone k data center i green electricity power, The time period for calculating carbon emissions.

[0119] In one specific implementation, the carbon emission excess model satisfies the following formula:

[0120] in, For the carbon trading costs of data centers, The base price for carbon trading in data centers, The interval length is... For price increases, This represents the actual carbon emissions of the data center. Carbon emission allowances for data centers; The carbon emission surplus model satisfies the following formula:

[0121] in, The carbon trading cost for data centers (in this scenario, the revenue from carbon emission surplus). The base price for carbon trading in data centers, The interval length is... As the reward coefficient, This represents the actual carbon emissions of the data center. Carbon emission allowances for data centers.

[0122] In one specific implementation, the two-layer optimization model includes an upper-layer data center energy consumption decision model and a lower-layer electricity market clearing model. The data center energy consumption decision model includes an optimization objective of minimizing energy costs; the optimization objective satisfies the following formula:

[0123] in, for t Time zone k data center i The cost of purchasing electricity in the electricity market for t Time zone k data center i The cost of generating electricity from a diesel generator. For the region k data center i The cost of carbon trading; The cost of purchasing electricity in the electricity market satisfies the following formula:

[0124] in, for t Time zone k data center i The cost of purchasing electricity in the electricity market for t Time zone k data center i Medium- to long-term market-clearing electricity prices for t Time zone k data center i The spot electricity market clearing price, for t Time zone k data center i Electricity volume under signed medium- and long-term contracts for t Time zone k data center i Total demand; The cost of generating electricity using a diesel generator satisfies the following formula:

[0125] in, for t Time zone k data center i The cost of generating electricity from a diesel generator. For the region k data center i The unit power generation cost of diesel generator sets, for t Time zone k data center i The power of the diesel generator; Electricity market clearing models include medium- to long-term market models and spot electricity market models; The medium- to long-term market model minimizes the market clearing costs and deviation assessment costs for data centers in the medium to long term; the medium- to long-term market model satisfies the following formula:

[0126] in, for t Time zone k data center i The medium- to long-term market clearing costs, for t Time zone k data center i The spot electricity market clearing price, for t Time zone k data center i Electricity volume under signed medium- and long-term contracts Costs for deviation assessment; for t Time-of-use contract The amount of electricity, , They are respectively t Time-of-use contract The lower and upper limits of battery capacity; for t Time zone k data center i Total demand Here, T is the deviation penalty coefficient, and T is the number of time periods. The spot electricity market model minimizes the difference between the cost of generating units and the cost of data centers in the spot market. The spot electricity market model satisfies the following formula:

[0127] in, This indicates the declared cost of the generating unit in the spot market. For the units in the current market j The first application s Price range In order to be in the market recently t Time-of-use units j No. s Duan made great efforts, This indicates the declared cost of data centers in the spot market. For the first data center application in the market recently Price range In order to be in the market recently t Time period data center Electricity purchased in segments, where T represents the number of time segments. Number of generating units The number of power sections of the unit. This refers to the number of power bands in the data center. In order to be in the market recently t Other users during the same period k Electricity purchase segment For the remaining users' power segment numbers; This is the upper limit of the generating capacity of the unit; , The lower and upper limits for electricity purchases for data centers.

[0128] In one specific implementation, the two-layer model collaborative optimization module includes: The model conversion unit is used to convert the two-layer optimization model for data centers participating in the electricity carbon market into a single-layer optimization model. An optimization unit is used to solve a single-layer optimization model based on carbon trading costs and spot electricity market clearing prices, with the electricity demand adjustment range as the boundary, to obtain the data center's electricity-carbon synergy strategy.

[0129] In one specific implementation, the single-layer optimization model satisfies the following formula:

[0130] Where F() represents the function of a single-layer optimization model, x For the decision vector of the data center; y, z These are variables related to the clearing issues in the medium- and long-term electricity market and the spot electricity market, respectively. G ( ) H ( ) represents a constraint on collaborative decision-making in data centers across multiple market types; 、 It serves as the dual variable for the constraints in the medium- and long-term electricity market clearing problem; 、 The dual variable for the constraints in the joint electricity market clearing problem; , The constraint set corresponding to the KKT system for medium- and long-term electricity market and joint market clearing issues.

[0131] Example 3: like Figure 7As shown, the present invention also provides an electronic device, which may be a computer device, a microcontroller device, a smart mobile device, etc. The electronic device in this embodiment may include a processor, a memory, a transceiver component, etc. The memory, processor, and transceiver component are connected via a bus; the memory can be used to store executable programs, and an exemplary executable program may include instructions; the processor is used to execute the instructions stored in the memory. The memory can also be used to store data, which can be accessed and / or modified when instructions are executed.

[0132] The processor may be a Central Processing Unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, and it is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the storage medium to realize the corresponding method flow or corresponding function, so as to realize the steps of the strategy optimization method for data centers to participate in the electricity and carbon market in the above embodiments.

[0133] Example 4: Based on the same inventive concept, this invention also provides a readable storage medium, specifically an electronic device readable storage medium (Memory). This readable storage medium is a memory device within an electronic device used to store programs and data. It is understood that the storage medium here can include both built-in storage media within the electronic device and extended storage media supported by the electronic device. The storage medium provides storage space, which stores the terminal's operating system. Furthermore, this storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more executable programs (including program code). It should be noted that the storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device. Loading and executing one or more instructions stored in the storage medium by the processor can implement the steps of the strategy optimization method for data centers participating in the electricity and carbon markets described in the above embodiments.

[0134] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0135] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0136] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0137] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0138] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit its scope of protection. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that after reading the present invention, they can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the application, but these changes, modifications or equivalent substitutions are all within the scope of protection of the claims pending approval.

Claims

1. A strategy optimization method for data centers participating in the electricity and carbon markets, characterized in that, include: Based on the constraints of energy storage charging and discharging and diesel generator output, combined with the power demand model of the data center, the power demand adjustment range of the data center is calculated. Based on the actual total power and green electricity power of the data center, the actual carbon emissions of the data center are calculated; based on the carbon emission quota of the data center and the actual carbon emissions, a tiered reward and punishment carbon trading model is used to calculate the carbon trading cost of the data center. Based on the carbon trading costs and the clearing price of the spot electricity market, and taking the electricity demand adjustment range as the boundary, the two-layer optimization model for data centers participating in the electricity carbon market is optimized and solved to obtain the data center's electricity carbon coordination strategy. The data center's electricity carbon coordination strategy includes load spatiotemporal transfer schemes, energy storage scheduling strategies, and market pricing strategies.

2. The method as described in claim 1, characterized in that, The power demand model includes the relationship between the total power of the data center, the processing load power of the data center, the energy storage charging and discharging power, and the diesel generator power. The power demand model satisfies the following formula: in, for t Time zone k Data Center i Total power, for t Time zone k Data Center i The processing load power, for t Time zone k Data Center i Energy storage charging power, for t Time zone k Data Center i The energy storage discharge power, for t Time zone k Data Center i The power of the diesel generator; These represent the unit energy consumption of data center i when handling online and offline loads, respectively. They are respectively t Time zone k Data Center i Handling online and offline workloads; This refers to the data center's operational efficiency factor.

3. The method as described in claim 1 or 2, characterized in that, The power demand model is established based on the spatiotemporal regulation characteristics of the load; the load includes online load and offline load; The spatiotemporal regulation characteristics of the load include spatial transfer constraints of online loads and time delay constraints of offline loads; The spatial transfer constraint of the online load satisfies the following formula: in, for t Time zone k Data Center i The total load processing Let t represent the online load arriving at data center i from region k during time period k. The time delay constraint of the offline load satisfies the following formula: in, for t Time zone k Data Center i Offline workload of storage, for t-1 Time zone k Data Center i Offline workload of storage, for t Time zone k data center i The online load being processed, for t Time zone k data center i The offline load being processed satisfy , for t Time zone k data center i The arriving offline load, The deadline for scheduling, For the region k Data Center i The maximum offline load of storage. For the region k Data Center i The offline load limit stored at the scheduling deadline; They are respectively regions k Data Center i The number of servers, service rate, and memory; For the delay limits of the Service Level Agreement; c, d These represent the data volume for online load and offline load, respectively.

4. The method as described in claim 1 or 2, characterized in that, The energy storage charging and discharging constraints satisfy the following formula: in, , They are respectively t Time zone k data center i The energy storage charging and discharging states; , They are respectively t Time zone k data center i The charging and discharging power of energy storage for t Time zone k data center i The upper limit of the discharge power of energy storage; , For respectively t Time period, t-1 Time zone k data center i The energy storage capacity; , They are respectively regions k data center i Energy storage charging and discharging efficiency; and for t Time zone k data center i The minimum and maximum values ​​of the energy storage capacity; The output constraint of the diesel generator satisfies the following formula: in, for t Time zone k data center i The power of the diesel generator, , They are respectively regions k data center i The upper and lower limits of the diesel generator power.

5. The method as described in claim 1, characterized in that, The carbon trading cost of the data center is calculated using a tiered reward and penalty carbon trading model, based on the data center's carbon emission quota and actual carbon emissions. This includes: If the carbon emission allowance of the data center is less than the actual carbon emission, then based on the carbon emission allowance, the actual carbon emission, and the basic carbon trading price of the data center, the carbon emission excess model in the tiered reward and punishment carbon trading model is used to calculate the carbon emission excess cost of the data center as the carbon trading cost. If the carbon emission allowance is not less than the actual carbon emission, then based on the carbon emission allowance, the actual carbon emission, and the carbon trading base price of the data center, the carbon emission surplus model in the tiered reward and punishment carbon trading model is used to calculate the carbon emission surplus revenue of the data center as the carbon trading cost.

6. The method as described in claim 5, characterized in that, The carbon emission quota is determined based on the sum of the carbon emission quotas of the computing unit, storage unit, and communication unit of the data center; The carbon emission quotas satisfy the following formula: in, For the region k data center i Carbon emission allowances; For the region k data center i The carbon emission allowance of the calculation unit; For the region k data center i Carbon emission quotas for storage units; For the region k data center i Carbon emission quotas for communication units; For data center operating efficiency factors; The actual carbon emissions are determined based on the difference between the actual total power of the data center and the green electricity power. The actual carbon emissions satisfy the following formula: in, For the region k data center i The actual carbon emissions; As an indirect carbon emission factor for data centers, for t Time zone k data center i The actual total power, for t Time zone k data center i green electricity power, The time period for calculating carbon emissions.

7. The method as described in claim 5 or 6, characterized in that, The carbon emission excess model satisfies the following formula: in, For the carbon trading costs of data centers, The base price for carbon trading in data centers. The interval length is... For price increases, This represents the actual carbon emissions of the data center. Carbon emission allowances for data centers; The carbon emission surplus model satisfies the following formula: in, The carbon trading cost for data centers (in this scenario, the revenue from carbon emission surplus). The base price for carbon trading in data centers. The interval length is... As the reward coefficient, This represents the actual carbon emissions of the data center. Carbon emission allowances for data centers.

8. The method as described in claim 1, characterized in that, The two-layer optimization model includes an upper-layer data center energy consumption decision model and a lower-layer electricity market clearing model. The data center energy consumption decision model includes an optimization objective of minimizing energy costs; the optimization objective satisfies the following formula: in, for t Time zone k data center i The cost of purchasing electricity in the electricity market for t Time zone k data center i The cost of generating electricity from a diesel generator. For the region k data center i Carbon trading costs; The electricity market purchase cost satisfies the following formula: in, for t Time zone k data center i The cost of purchasing electricity in the electricity market for t Time zone k data center i Medium- to long-term market-clearing electricity prices for t Time zone k data center i The spot electricity market clearing price, for t Time zone k data center i Electricity volume under signed medium- and long-term contracts for t Time zone k data center i Total demand; The power generation cost of the diesel generator satisfies the following formula: in, for t Time zone k data center i The cost of generating electricity from a diesel generator. For the region k data center i The unit power generation cost of diesel generator sets, for t Time zone k data center i The power of the diesel generator; The electricity market clearing model includes a medium- to long-term market model and a spot electricity market model. The medium- to long-term market model minimizes the market clearing cost and deviation assessment cost for data centers in the medium to long term; the medium- to long-term market model satisfies the following formula: in, for t Time zone k data center i The medium- to long-term market clearing costs, for t Time zone k data center i The spot electricity market clearing price, for t Time zone k data center i Electricity volume under signed medium- and long-term contracts Costs for deviation assessment; for t Time-of-use contract The amount of electricity, , They are respectively t Time-of-use contract The lower and upper limits of battery capacity; for t Time zone k data center i Total demand The deviation penalty coefficient is T, where T is the number of time periods. The spot electricity market model minimizes the difference between the application cost of generating units and the application cost of data centers in the spot market; The spot electricity market model satisfies the following formula: in, This indicates the declared cost of the generating unit in the spot market. For the units in the current market j The first application s Price range In order to be in the market recently t Time-of-use units j No. s Duan made great efforts, This indicates the declared cost of data centers in the spot market. For the first data center application in the market recently Price range In order to be in the market recently t Time period data center Electricity purchased in segments, where T represents the number of time segments. Number of generating units The number of power sections of the unit. This refers to the number of power bands in the data center. In order to be in the market recently t Other users during the same period k Purchase electricity in stages, For the remaining users' power segment numbers; This is the upper limit of the generating capacity of the unit; , The lower and upper limits for electricity purchases for data centers.

9. The method as described in claim 1 or 8, characterized in that, Based on the carbon trading costs and the clearing price of the spot electricity market, and using the electricity demand adjustment range as the boundary, the two-level optimization model for data centers participating in the electricity carbon market is optimized and solved to obtain the data center's electricity carbon coordination strategy, including: Transform the two-layer optimization model for data center participation in the electricity carbon market into a single-layer optimization model; Based on the carbon trading costs and the clearing price of the spot electricity market, and using the electricity demand adjustment range as the boundary, the single-layer optimization model is solved to obtain the data center's electricity-carbon coordination strategy.

10. The method as described in claim 9, characterized in that, The single-layer optimization model satisfies the following formula: Where F() represents the function of a single-layer optimization model, x For the decision vector of the data center; y, z These are variables related to the clearing issues in the medium- and long-term electricity market and the spot electricity market, respectively. G ( ) H ( ) represents a constraint on collaborative decision-making in data centers across multiple market types; 、 It serves as the dual variable for the constraints in the medium- and long-term electricity market clearing problem; 、 The dual variable for the constraints in the joint electricity market clearing problem; , The constraint set corresponding to the KKT system for medium- and long-term electricity market and joint market clearing issues.

11. A strategy optimization system for data centers participating in the electricity and carbon markets, characterized in that, include: The adjustment range determination module is used to calculate the power demand adjustment range of the data center based on the energy storage charging and discharging constraints and the diesel generator output constraints, combined with the power demand model of the data center. The carbon trading cost determination module is used to calculate the actual carbon emissions of the data center based on the actual total power and green electricity power of the data center; and to calculate the carbon trading cost of the data center based on the carbon emission quota of the data center and the actual carbon emissions using a tiered reward and punishment carbon trading model. The two-layer model collaborative optimization module is used to optimize and solve the two-layer optimization model for data centers participating in the electricity carbon market based on the carbon trading cost and the clearing price of the spot electricity energy market, with the electricity demand adjustment range as the boundary, so as to obtain the data center's electricity carbon collaborative strategy. The data center's electricity carbon collaborative strategy includes load spatiotemporal transfer scheme, energy storage scheduling strategy and market bidding strategy.

12. The system as claimed in claim 11, characterized in that, The power demand model includes the relationship between the total power of the data center, the processing load power of the data center, the energy storage charging and discharging power, and the diesel generator power; the load includes online load and offline load. The power demand model satisfies the following formula: in, for t Time zone k Data Center i Total power, for t Time zone k Data Center i The processing load power, for t Time zone k Data Center i Energy storage charging power, for t Time zone k Data Center i The energy storage discharge power, for t Time zone k Data Center i The power of the diesel generator; These represent the unit energy consumption of data center i when handling online and offline loads, respectively. They are respectively t Time zone k Data Center i Handling online and offline workloads; This refers to the data center's operational efficiency factor.

13. The system as described in claim 11 or 12, characterized in that, The power demand model is established based on the spatiotemporal regulation characteristics of the load; the load includes online load and offline load; The spatiotemporal regulation characteristics of the load include spatial transfer constraints of online loads and time delay constraints of offline loads; The spatial transfer constraint of the online load satisfies the following formula: in, for t Time zone k Data Center i The total load processing Let t represent the online load arriving at data center i from region k during time period k. The time delay constraint of the offline load satisfies the following formula: in, for t Time zone k Data Center i Offline workload of storage, for t-1 Time zone k Data Center i Offline workload of storage, for t Time zone k data center i The online load being processed, for t Time zone k data center i The offline load being processed satisfy , for t Time zone k data center i The arriving offline load, The deadline for scheduling, For the region k Data Center i The maximum offline load of storage. For the region k Data Center i The offline load limit stored at the scheduling deadline; They are respectively regions k Data Center i The number of servers, service rate, and memory; For the delay limits of the Service Level Agreement; c, d These represent the data volume for online load and offline load, respectively.

14. The system as described in claim 11 or 12, characterized in that, The energy storage charging and discharging constraints satisfy the following formula: in, , They are respectively t Time zone k data center i The energy storage charging and discharging states; , They are respectively t Time zone k data center i The charging and discharging power of energy storage for t Time zone k data center i The upper limit of the discharge power of energy storage; , For respectively t Time period, t-1 Time zone k data center i The energy storage capacity; , They are respectively regions k data center i Energy storage charging and discharging efficiency; and for t Time zone k data center i The minimum and maximum values ​​of the energy storage capacity; The output constraint of the diesel generator satisfies the following formula: in, for t Time zone k data center i The power of the diesel generator, , They are respectively regions k data center i The upper and lower limits of the diesel generator power.

15. The system as claimed in claim 11, characterized in that, The carbon trading cost determination module is specifically used for: If the carbon emission allowance of the data center is less than the actual carbon emission, then based on the carbon emission allowance, the actual carbon emission, and the basic carbon trading price of the data center, the carbon emission excess model in the tiered reward and punishment carbon trading model is used to calculate the carbon emission excess cost of the data center as the carbon trading cost. If the carbon emission allowance is not less than the actual carbon emission, then based on the carbon emission allowance, the actual carbon emission, and the carbon trading base price of the data center, the carbon emission surplus model in the tiered reward and punishment carbon trading model is used to calculate the carbon emission surplus revenue of the data center as the carbon trading cost.

16. The system as described in claim 15, characterized in that, The carbon emission quota is determined based on the sum of the carbon emission quotas of the computing unit, storage unit, and communication unit of the data center; The carbon emission quotas satisfy the following formula: in, For the region k data center i Carbon emission allowances; For the region k data center i The carbon emission allowance of the calculation unit; For the region k data center i Carbon emission quotas for storage units; For the region k data center i Carbon emission quotas for communication units; For data center operating efficiency factors; The actual carbon emissions are determined based on the difference between the actual total power of the data center and the green electricity power. The actual carbon emissions satisfy the following formula: in, For the region k data center i The actual carbon emissions; As an indirect carbon emission factor for data centers, for t Time zone k data center i The actual total power, for t Time zone k data center i green electricity power, The time period for calculating carbon emissions.

17. The system as described in claim 15 or 16, characterized in that, The carbon emission excess model satisfies the following formula: in, For the carbon trading costs of data centers, The base price for carbon trading in data centers. The interval length is... For price increases, This represents the actual carbon emissions of the data center. Carbon emission allowances for data centers; The carbon emission surplus model satisfies the following formula: in, The carbon trading cost for data centers (in this scenario, the revenue from carbon emission surplus). The base price for carbon trading in data centers. The interval length is... As the reward coefficient, This represents the actual carbon emissions of the data center. Carbon emission allowances for data centers.

18. The system as claimed in claim 11, characterized in that, The two-layer optimization model includes an upper-layer data center energy consumption decision model and a lower-layer electricity market clearing model. The data center energy consumption decision model includes an optimization objective of minimizing energy costs; the optimization objective satisfies the following formula: in, for t Time zone k data center i The cost of purchasing electricity in the electricity market for t Time zone k data center i The cost of generating electricity from a diesel generator. For the region k data center i Carbon trading costs; The electricity market purchase cost satisfies the following formula: in, for t Time zone k data center i The cost of purchasing electricity in the electricity market for t Time zone k data center i Medium- to long-term market-clearing electricity prices for t Time zone k data center i The spot electricity market clearing price, for t Time zone k data center i Electricity volume under signed medium- and long-term contracts for t Time zone k data center i Total demand; The power generation cost of the diesel generator satisfies the following formula: in, for t Time zone k data center i The cost of generating electricity from a diesel generator. For the region k data center i The unit power generation cost of diesel generator sets, for t Time zone k data center i The power of the diesel generator; The electricity market clearing model includes a medium- to long-term market model and a spot electricity market model. The medium- to long-term market model minimizes the market clearing cost and deviation assessment cost for data centers in the medium to long term; the medium- to long-term market model satisfies the following formula: in, for t Time zone k data center i The medium- to long-term market clearing costs, for t Time zone k data center i The spot electricity market clearing price, for t Time zone k data center i Electricity volume under signed medium- and long-term contracts Costs for deviation assessment; for t Time-of-use contract The amount of electricity, , They are respectively t Time-of-use contract The lower and upper limits of battery capacity; for t Time zone k data center i Total demand The deviation penalty coefficient is T, where T is the number of time periods. The spot electricity market model minimizes the difference between the application cost of generating units and the application cost of data centers in the spot market; The spot electricity market model satisfies the following formula: in, This indicates the declared cost of the generating unit in the spot market. For the units in the current market j The first application s Price range In order to be in the market recently t Time-of-use units j No. s Duan made great efforts, This indicates the declared cost of data centers in the spot market. For the first data center application in the market recently Price range In order to be in the market recently t Time period data center Electricity purchased in segments, where T represents the number of time segments. Number of generating units The number of power sections of the unit. This refers to the number of power bands in the data center. In order to be in the market recently t Other users during the same period k Purchase electricity in stages, For the remaining users' power segment numbers; This is the upper limit of the generating capacity of the unit; , The lower and upper limits for electricity purchases for data centers.

19. The system as described in claim 11 or 18, characterized in that, The two-layer model collaborative optimization module includes: The model conversion unit is used to convert the two-layer optimization model for data centers participating in the electricity carbon market into a single-layer optimization model. An optimization solution unit is used to solve the single-layer optimization model based on the carbon trading cost and the spot electricity market clearing price, with the electricity demand adjustment range as the boundary, to obtain the data center's electricity-carbon coordination strategy.

20. The system as described in claim 19, characterized in that, The single-layer optimization model satisfies the following formula: Where F() represents the function of a single-layer optimization model, x For the decision vector of the data center; y, z These are variables related to the clearing issues in the medium- and long-term electricity market and the spot electricity market, respectively. G ( ) H ( ) represents a constraint on collaborative decision-making in data centers across multiple market types; 、 It serves as the dual variable for the constraints in the medium- and long-term electricity market clearing problem; 、 The dual variable for the constraints in the joint electricity market clearing problem; , The constraint set corresponding to the KKT system for medium- and long-term electricity market and joint market clearing issues.

21. An electronic device, characterized in that, include: At least one processor and memory; The memory and processor are connected via a bus; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, the strategy optimization method for data center participation in the electricity and carbon markets as described in any one of claims 1 to 10 is implemented.

22. A readable storage medium, characterized in that, It contains an execution program, which, when executed, implements the strategy optimization method for data centers participating in the electricity and carbon markets as described in any one of claims 1 to 10.

Images