CN114841510B - Data center cluster shared energy storage method, system, device and medium - Google Patents
Data center cluster shared energy storage method, system, device and medium Download PDFInfo
- Publication number
- CN114841510B CN114841510B CN202210322060.XA CN202210322060A CN114841510B CN 114841510 B CN114841510 B CN 114841510B CN 202210322060 A CN202210322060 A CN 202210322060A CN 114841510 B CN114841510 B CN 114841510B
- Authority
- CN
- China
- Prior art keywords
- data center
- energy
- representing
- scheduling period
- power
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000004146 energy storage Methods 0.000 title claims abstract description 101
- 238000000034 method Methods 0.000 title claims abstract description 27
- 239000007789 gas Substances 0.000 claims description 23
- 230000004044 response Effects 0.000 claims description 22
- 238000001816 cooling Methods 0.000 claims description 21
- 230000008901 benefit Effects 0.000 claims description 16
- 230000005611 electricity Effects 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 13
- 230000015654 memory Effects 0.000 claims description 11
- 238000003860 storage Methods 0.000 claims description 11
- 239000002699 waste material Substances 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 5
- 230000009467 reduction Effects 0.000 claims description 5
- 239000002737 fuel gas Substances 0.000 claims description 3
- 238000005265 energy consumption Methods 0.000 abstract description 9
- 238000010248 power generation Methods 0.000 description 10
- 238000010586 diagram Methods 0.000 description 7
- 230000006872 improvement Effects 0.000 description 7
- 230000017525 heat dissipation Effects 0.000 description 6
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 6
- 238000010521 absorption reaction Methods 0.000 description 4
- 238000007599 discharging Methods 0.000 description 4
- 238000005338 heat storage Methods 0.000 description 4
- 238000005457 optimization Methods 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 239000003345 natural gas Substances 0.000 description 3
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 3
- 230000001186 cumulative effect Effects 0.000 description 2
- 238000005315 distribution function Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000020169 heat generation Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000004931 aggregating effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000029087 digestion Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000011065 in-situ storage Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06315—Needs-based resource requirements planning or analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06312—Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/067—Enterprise or organisation modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/007—Arrangements for selectively connecting the load or loads to one or several among a plurality of power lines or power sources
- H02J3/0075—Arrangements for selectively connecting the load or loads to one or several among a plurality of power lines or power sources for providing alternative feeding paths between load and source according to economic or energy efficiency considerations, e.g. economic dispatch
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/12—Circuit arrangements for ac mains or ac distribution networks for adjusting voltage in ac networks by changing a characteristic of the network load
- H02J3/14—Circuit arrangements for ac mains or ac distribution networks for adjusting voltage in ac networks by changing a characteristic of the network load by switching loads on to, or off from, network, e.g. progressively balanced loading
- H02J3/144—Demand-response operation of the power transmission or distribution network
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/28—Arrangements for balancing of the load in a network by storage of energy
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2300/00—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation
- H02J2300/20—The dispersed energy generation being of renewable origin
- H02J2300/28—The renewable source being wind energy
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Tourism & Hospitality (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- Power Engineering (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- General Health & Medical Sciences (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a data center cluster sharing energy storage method, which promotes the consumption of renewable energy sources by utilizing the complementation of energy consumption of different data centers through the energy balance in the data center clusters; the shared energy storage model with the minimum daily energy cost is used for improving the utilization rate of renewable energy and simultaneously benefiting through selling energy, so that the daily energy cost is reduced, and the utilization rate of the total energy is improved.
Description
Technical Field
The present disclosure relates to data center electrical energy storage scheduling, and more particularly, to a data center cluster shared energy storage method, system, apparatus, and medium that take into account uncertainty of renewable energy sources.
Background
In order to meet the increasing demand for online computing, data centers (DATA CENTER, DC for short) providing information technology services have evolved faster and faster in recent years. Modern DC is typically very high occupancy, resulting in extremely large costs for DC. DC has a strong DR capability as a Demand Response (DR) resource.
Modern DC is typically very high in occupancy and power consumption, resulting in extremely large costs for DC. In order to reduce energy costs, more and more DC are installed with renewable energy power generation devices. However, DC is a demand response resource, has a strong demand response capability, and can result in different energy demands of different DCs after participating in DR. For DC with small energy demand and large wind power generation, there is excess energy. In contrast, for DC with a large energy demand and a small wind power generation, there is an energy shortage.
Therefore, there is a problem in that renewable energy sources inside the data center Cluster (DATA CENTER Cluster, DCC for short) are underutilized. Because of the uncertainty of renewable energy sources, DCC needs to store energy, but the energy storage device is high, and DCC has the problem that energy cannot be stored and taken.
Disclosure of Invention
Aiming at the prior art, the technical problem to be solved by the invention is to provide the data center cluster shared energy storage method, which not only can enable renewable energy sources in the data center cluster to be consumed, but also can realize optimal reserved discharge and charging power planning on the basis of ensuring safe operation of the data center cluster by cooperating with the comprehensive shared energy storage system, thereby reducing energy storage cost.
In order to solve the technical problems:
In a first aspect, the data center cluster sharing energy storage method of the present invention includes the following steps:
Establishing a server model, a hybrid cooling system model, a renewable energy model, a back pressure turbine model and a gas boiler model, thereby obtaining the energy demand of each data center in each period;
Establishing a demand response model, and carrying out energy scheduling in a data center cluster according to the predicted value of the renewable energy generating capacity device of each data center in each period and the energy demand of each data center;
According to the scheduling result, carrying out energy storage planning based on a shared energy storage model of the data center cluster;
The shared energy storage model is a model which aims at minimizing daily energy cost of the data center cluster; the daily cost comprises energy purchasing cost, renewable energy source cost reduction, energy storage service cost and energy selling benefit; the energy purchasing cost comprises electricity purchasing cost, heat purchasing cost and gas purchasing cost.
In the technical scheme, the energy requirements of all the data centers in the data center cluster are acquired, the energy consumption complementation of different data centers is utilized to balance internal energy, and the renewable energy consumption is promoted; the shared energy storage model with the minimum daily energy cost improves the utilization rate of renewable energy, benefits from selling energy benefit, reduces the daily energy cost, and improves the overall energy utilization rate.
As a further improvement of the above technical solution, the daily cost further includes a standby cost; the standby cost comprises standby discharging cost and standby charging cost, so that the shared energy storage model can process the power generation quantity prediction error of the renewable energy source device, and the safety and reliability of the cluster operation of the data center are improved.
As a further improvement of the above-described technical solution, the daily cost further includes a risk cost; the risk costs include the loss of load costs, relinquishing renewable energy costs, meeting confidence level requirements; the load loss cost is equal to the product of the unit load loss cost and the absence between the actual standby charging power and the standby charging power which just meets the confidence level requirement; the abandoned renewable energy cost is equal to the product of the unit abandoned renewable energy cost and the actual standby discharge power and the standby discharge power which just meets the confidence level requirement, so that the shared energy storage model can be flexibly adjusted based on the standby discharge and charge power which just meets the confidence level requirement, and the shared energy storage model becomes a scheduling model based on the opportunity constraint target planning, and the standby storage power is allowed to deviate from the standby discharge and charge power value which just meets the opportunity constraint to a certain extent.
As a further improvement of the above technical solution, the energy scheduling is implemented by computing task redistribution; the influence factors of the computing task reassignment include: calculating the delay characteristic and the interruptible characteristic of the task, whether the renewable energy device is generating in a peak period, whether the comprehensive shared energy storage system is in a low price energy price and/or an energy service price. The data center is a system with high coupling of data and energy, more calculation tasks are distributed in a period with larger renewable energy generating capacity through calculation task redistribution, so that the discarding amount of the renewable energy generating capacity is reduced, and the utilization rate of the renewable energy generating capacity is improved. The energy costs are advantageously reduced when the calculation tasks are performed during periods of low price energy prices and/or energy service prices.
As a further improvement of the technical scheme, the renewable energy generating device comprises a wind turbine generator set, the wind turbine generator set is arranged in a data center, and natural wind power is utilized for generating electricity, so that the device is clean and sanitary.
As a further improvement of the technical scheme, the wind turbine generator meets the following model when not participating in energy scheduling:
wherein:
Representing the wind power prediction error of the data center i in the t period; n (0, σ 2) represents a normal distribution with a mean of 0 and a variance of σ 2; representing the predicted wind power output power value of the ith data center in the t scheduling period; Representing the actual wind power utilization of the ith data center in the t scheduling period; representing the waste wind power of the ith data center in the t scheduling period; i represents a data center set; t represents a set of scheduling periods;
When internal energy is scheduled, the output model of the wind turbine generator meets the following conditions:
As a further improvement of the invention, the energy storage planning comprises: according to the price, the energy use condition and the capacity condition of the renewable energy device, whether to acquire direct use energy, whether to store energy from the comprehensive shared energy storage system and whether to release energy to the comprehensive shared energy storage system is determined, so that renewable energy consumption is promoted, the data center cluster obtains economic benefit through exothermic discharge, the purchasing and storing energy cost is further reduced, the requirements of the data center cluster on energy storage and taking are met, and the safe operation of the data center cluster is further ensured.
In a second aspect, the invention provides a data center cluster shared energy storage system, the system comprises a data center cluster and a comprehensive shared energy storage system, the data center cluster comprises a data center, and a renewable energy device is arranged in data. Establishing a server model, a hybrid cooling system model, a renewable energy model, a back pressure turbine model and a gas boiler model, thereby obtaining the energy demand of each data center in each period; establishing a demand response model, and carrying out energy scheduling in a data center cluster according to the predicted value of the renewable energy generating capacity device of each data center in each period and the energy demand of each data center; according to the scheduling result, carrying out energy storage planning based on a shared energy storage model of the data center cluster; the shared energy storage model is a model which aims at minimizing daily energy cost of the data center cluster; the daily cost comprises energy purchasing cost, renewable energy source cost reduction, energy storage service cost and energy selling benefit; the energy purchasing cost comprises electricity purchasing cost, heat purchasing cost and gas purchasing cost; the energy storage planning comprises the steps of determining whether to purchase directly used energy, store energy from the comprehensive shared energy storage system and release energy to the comprehensive shared energy storage system according to the price, the energy use condition and the capacity condition of the renewable energy device.
In the technical scheme, the system is used for fully utilizing renewable energy sources in the data center cluster, and by being matched with the comprehensive shared energy storage system, the data center cluster obtains economic benefits through exothermic discharge, so that the cost of purchasing and storing the energy sources is reduced, the requirements of the data center cluster on energy storage and taking are met, and the safe operation of the data center cluster is further ensured.
In a third aspect, the present invention provides a data center cluster shared storage device comprising a memory and a processor, the memory having stored thereon a computer program capable of being loaded by the processor and performing any of the methods described above.
In a fourth aspect, the present invention provides a computer readable storage medium storing a computer program capable of being loaded by a processor and performing any one of the methods described above.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort to a person skilled in the art.
FIG. 1 is a schematic diagram of an application mode of a shared energy storage method for a data center cluster and a comprehensive shared energy storage system according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of energy prices of the integrated energy system according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of an energy service price of the integrated energy system according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a wind power output prediction curve of a wind turbine generator system according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of power supply and power demand in a 24-hour data center cluster in accordance with one embodiment of the invention;
FIG. 6 is a schematic diagram of a data center cluster without internal energy balancing in accordance with one embodiment of the present invention;
FIG. 7 is a schematic diagram of internal energy balancing of a cluster of data centers in accordance with one embodiment of the invention.
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments.
In embodiment 1, the data center cluster and the comprehensive shared energy storage system adopt an application mode of the shared energy storage method, as shown in fig. 1. In fig. 1, the renewable energy power generation device is a wind turbine generator set and is installed in a data center. The renewable energy source may also be water energy, solar energy, light energy, etc. in other embodiments. Because the power consumption of the data center is huge, the renewable energy power generation device is arranged in the data center, so that the energy cost is reduced. In example 1, it is assumed that one wind turbine is installed in each data center to provide renewable energy.
Model 1: the output model of the wind turbine generator is expressed as:
wherein:
Representing the wind power prediction error of the data center i in the t period; n (0, σ 2) represents a normal distribution with a mean of 0 and a variance of σ 2; representing the predicted wind power output power value of the ith data center in the t scheduling period; Representing the actual wind power utilization of the ith data center in the t scheduling period; Representing the waste wind power of the ith data center in the t scheduling period; i represents a data center set; t denotes a set of scheduling periods.
The electric energy generated by the wind turbine generator system in model 1 is used by the data center where the wind turbine generator system is located.
There are four energy flows in the data center, including current, heat flow, natural gas flow, and cooling flow. The power consuming components of the data center include servers and electric chillers. The heat consumption components are a double-effect absorption refrigerator and a back pressure steam turbine. The natural gas consuming element is a gas fired boiler. The double-effect absorption refrigerator and the electric refrigerator provide cooling energy for heat dissipation of the server.
Model 2: the server model is as follows:
wherein:
C pro denotes a processor set, C mem denotes a memory set, C oth denotes a set of other electronic components in the server, and C1 is an element index in three sets; s ser represents a server set; a is a set of computing tasks; p i,s,t denotes a power value of an s-th server in the i-th data center at the t-th scheduling period; p i,c1,t denotes the power value of the c1 st element in the i-th data center at the t-th scheduling period; Representing a power value at a t-th scheduling period in an i-th data center; u i,s,t represents the utilization of the ith server in the ith data center during the ith scheduling period; representing the maximum power of the c1 st server component in the i-th data center; A state variable indicating whether the computing task a is executed in the s-th server of the i-th data center in the t-th scheduling period; representing power in an idle state of a c1 st server component in an i-th data center; f a is the CPU frequency occupied by processing computing task a.
Air cooling and water cooling are two typical cooling modes in modern hybrid cooling systems. The electric refrigerator uses electrically driven air cooling, while the double effect absorption refrigerator uses thermally driven liquid cooling. In modern data centers, with respect to server heat dissipation, water cooling is typically used for heat dissipation of processors and memory, while air cooling is typically used for heat dissipation of other electronic components.
Model 3: the hybrid cooling system model is expressed as follows:
wherein:
Representing a processor generated thermal power in an ith data center for an ith scheduling period; representing a processor electrical to thermal energy conversion coefficient in an ith data center; Representing an ith server processor electrical power in an ith data center for an ith schedule period; Representing a memory heat generation power in an ith data center of the ith scheduling period; representing a memory electric-to-thermal energy conversion coefficient in an ith data center; Representing an ith server memory electric power in an ith data center for an ith schedule period; Representing a t scheduling period i other electronic component heat generation power in the server in the data center; Representing the electrical to thermal energy conversion coefficients of other electronic components in the memory server in the ith data center; Representing the electric power of other electronic components in an s-th server in an i-th data center of an t-th schedule period; c EC represents an electric refrigerator set, C2 is an element in C EC; Representing the cold power output by the c2 nd element in the ith data center of the t scheduling period; Representing the coefficient of performance of the c2 th element input in the i-th data center; Representing the electric power input by the c2 nd element in the ith data center of the t scheduling period; Representing the heat dissipation power required by the c2 nd element in the ith data center of the t scheduling period; Representing the maximum electric power input by the c2 th element in the ith data center; c DAC represents a double-effect absorption refrigerator set, C3 is an element therein; representing the cold power output by the c3 rd element in the ith data center of the t scheduling period; representing the coefficient of performance of the c3 rd element input in the i-th data center; representing the thermal power input by the c3 rd element in the ith data center of the t scheduling period; Representing the coefficient of performance of the c3 rd element output in the i-th data center; Representing the heat dissipation power required by the c3 rd element in the ith data center of the t scheduling period; Representing the electric power required by the ith data center cooling system in the t-th scheduling period; representing the thermal power required by the ith data center cooling system in the t scheduling period; Representing the maximum thermal power input by the c3 rd element in the ith data center.
Model 4: the back pressure turbine model is as follows:
wherein:
C BPST represents a back pressure turbine set, and C4 is an element in C BPST; Representing the thermal power output by the c4 th element in the ith data center of the t scheduling period; Representing the coefficient of performance of the c4 th element input heat transfer in the i-th data center; representing the coefficient of performance of the c4 th element input heat to electricity in the ith data center; Representing the thermal power input by the c4 th element in the ith data center of the t scheduling period; representing the electric power output by the c4 th element in the ith data center of the t scheduling period; Representing the maximum thermal power input by the c4 th element in the ith data center of the t scheduling period.
Model 5: the gas boiler model is expressed as follows:
wherein:
C GB represents a collection of gas boilers, C5 is an element in C GB; epsilon is the natural gas heating value; representing the thermal power output by the c5 th element in the ith data center of the t scheduling period; representing the coefficient of performance of the c5 th element input in the i-th data center; representing the fuel gas power input by a c5 th element in an ith data center of a t scheduling period; Showing the thermal power input by the c4 th element in the ith data center of the t scheduling period; representing the maximum gas power input by the c5 th element in the ith data center; x c5 represents the energy loss rate of the c5 th element.
The computing tasks of the data center can be divided into deferrable tasks and undelayable tasks according to real-time requirements, and can be divided into continuous tasks and interruptable tasks according to continuity requirements. Because the computing task and the energy of the data center are highly coupled, the load of the data center can be flexibly transferred in time and space by fully utilizing the delay characteristic and the interruptible characteristic of the computing task, so that the energy consumption of different data centers in the data center cluster can be flexibly adjusted. In order to fully utilize the space-time transfer characteristics of the data center load, a demand response model of the data center is established as follows.
Model 6: the demand response model is expressed as follows:
If a is a epsilon A del,
If a is a epsilon A non,
If a is a epsilon A int,
If a is a epsilon A con,
a∈A，
Wherein:
U is a data center cluster set, and U is an element in U; Expressing the ith data center of the ith data center cluster in the ith scheduling period to transfer out the calculation task number; i -i represents the set of data centers other than the ith data center, I' being an element in I -i; the ith' data center of the ith data center cluster in the ith scheduling period is transferred to the computing task number; a is a computing task set, a is an element therein; representing the completion time of task a; Arrival time of task a; representing the deferrable time of task a; Representing the processing time length of the task a; representing the maximum deferrable time of task a; a del represents a deferrable task set, a non represents a non-deferrable task set, a int represents an interruptible task set, and a con represents a continuity task set; Representing the time required for task a to calculate; a start instruction amount for the task a representing the t scheduling period; Indicating the completion calculation indicating quantity of the task a in the t scheduling period; a state variable representing execution of task a for the t-th scheduling period; Representing the state variable of task a execution for the t-1 th scheduling period.
After the demand response, the computing tasks of the data center cluster are redistributed, and at this time, different energy demands of different data centers are caused: for data centers with small energy demands and large wind power generation, there is excess energy. In contrast, there is a shortage of energy for data centers with large energy demands and small wind power generation. The data centers in the same data center cluster exchange energy first to perform internal energy balance, so that complementary energy consumption of different data centers can be fully utilized. Therefore, the generated energy of the wind turbine is first used for internal energy scheduling balance, so the wind turbine output model in the model 1 is improved to the following model.
Model 7: and when internal energy is scheduled, the output model of the wind turbine generator is as follows:
Specifically, as shown in fig. 1, data centers in a cluster of data centers communicate data via internal data communication links. The transmission data comprise the electric load power and wind power predicted value of the data center in each period. And the data center cluster operator collects the data of each data center for centralized optimization and returns the dispatching optimization result to each data center. And according to the dispatching result of the data center cluster operators, the surplus wind power is transmitted to the electricity-shortage data center by the surplus wind power data center, so that the electricity supply interaction of the surplus wind power data center and the electricity-shortage data center is realized. Through an internal energy balance mechanism, the generated energy of the wind turbine generator in the data center cluster can be utilized by any data center in the data center cluster.
Further, the surplus energy in the data center cluster is stored through the comprehensive shared energy storage system, so that energy consumption in the data center cluster is promoted, the stored energy can be used by other users sharing the energy storage system, and therefore, the surplus energy can increase economic benefits for the data center cluster. The comprehensive shared energy storage system is provided with an electricity storage system and a heat storage system, and energy storage service which is collected by lease capacity and lease power is provided.
Model 8: the rental capacity-based charging method is expressed as follows:
wherein:
Cca u represents the cost of leased storage capacity for the nth data center cluster; SE cap represents the price of the electrical energy service per unit capacity; SH cap represents the thermal energy service price per unit capacity; representing the maximum leasing power of the leasing power storage system of the u-th data center cluster; representing the minimum leasing power of the leasing power storage system of the u-th data center cluster; representing the maximum leasing heat of the leasing heat storage system of the u-th data center cluster; Representing the minimum lease heat of the heat storage system leased by the ith data center cluster; e u,1 represents the leasing electricity quantity of the ith data center cluster of the 1 st scheduling period; e u,t represents the leasing electricity quantity of the ith data center cluster of the t scheduling period; e u,t+1 represents the leasing electricity quantity of the (t+1) th scheduling period (t+1) th data center cluster; the charging power of the u-th data center cluster representing the 1 st scheduling period; discharge power of the u-th data center cluster representing the 1 st scheduling period; The charging power of the nth data center cluster representing the nth scheduling period; The discharge power of the nth data center cluster representing the nth scheduling period; tsum represents the total number of scheduling time periods, t is a scheduling time period variable; the endothermic power of the u-th data center cluster representing the 1 st scheduling period; the endothermic power of the ith data center cluster representing the ith scheduling period; Exothermic power of the u-th data center cluster representing the 1 st scheduling period; a heat release power of a kth data center cluster representing a kth scheduling period; Δt represents the time interval of two adjacent scheduling periods; qu, 1 represents the u-th data center cluster leasing heat of the 1 st schedule period, Q u,t represents the u-th data center cluster leasing heat of the t-th schedule period, and Q u,t+1 represents the u-th data center cluster heat storage system leasing heat of the t+1 th schedule period.
Model 9: the rental power based charging method is expressed as follows:
wherein:
Cpo u represents the leased energy storage power cost of the u-th data center cluster; SE pow represents the price of electrical energy service per unit power; a nth data center cluster charging power representing a nth scheduling period; The discharge power of the nth data center cluster of the nth scheduling period is represented; SH pow represents the thermal energy service price per unit power; representing the maximum charging power of the u-th data center cluster; representing the maximum discharge power of the nth data center cluster; a state of charge variable representing a nth data center cluster of the nth scheduling period; a discharge state variable representing a nth data center cluster of the nth scheduling period; a nth data center cluster endothermic power representing a nth scheduling period; a nth data center cluster exothermic power representing a nth scheduling period; representing the maximum endothermic power of the u-th data center cluster; Representing a maximum heat release power of the first data center cluster; an endothermic state variable representing a nth data center cluster of the nth scheduling period; And (3) a heat release state variable of the ith data center cluster representing the t-th scheduling period.
According to model 8 and model 9, the economic benefit obtained by the data center cluster from the comprehensive shared energy storage system is calculated by the following economic benefit model:
Model 10:
wherein:
BSS u represents the economic benefit of the u-th data center cluster; CE dis represents a price of a discharge acquisition contract signed by the data center cluster and the comprehensive shared energy storage system; CH rel represents the exothermic acquisition contract price for the data center cluster with the integrated shared energy storage system.
Further, the data center cluster aims at minimizing daily cost, performs calculation task reassignment, and arranges an energy consumption plan, a purchase energy plan and an energy storage service purchase plan. The cost of the data center cluster comes from the cost of energy purchase from the integrated energy system, the cost of wind curtailment, and the cost of energy storage service. Data center clusters are economically efficient from both discharge and heat release.
Model 11: under the condition that the operation constraint of the data center cluster is met, the shared energy storage model based on the data center cluster is expressed as follows: min DC u＝CPIu+CWCu+CESu-BSSu
In the case where models 2-10 and their constraints are satisfied, the following constraints are added:
s.t.CPIu＝CPEIu+CPHIu+CPGIu
CESu＝Ccau+Cpou
wherein:
DC u represents the daily cost of the nth data center cluster; CPI u represents the purchase cost of the nth data center cluster; CWC u represents the wind curtailment cost of the u-th data center cluster; CES u represents the energy storage service cost of the u-th data center cluster; CPEI u represents the purchase cost of the ith data center cluster; CPHI u represents the purchase cost of the u-th data center cluster; CPGI u represents the gas purchase cost of the u-th data center cluster; The power purchasing power of the ith data center cluster of the t scheduling period is represented; EPI t represents the electricity purchase price at the t scheduling period; The heat purchasing power of the ith data center cluster of the t scheduling period is represented; HPI t represents the t-th scheduling period purchase price; The gas purchasing power of a ith data center cluster in the t scheduling period is represented; GPI t gas purchase price in the t scheduling period; l wc represents the unit cost of the waste wind; Representing the load electric power of the ith data center of the t-th scheduling period; representing the power supply system electric power of the ith data center of the t scheduling period; Representing the lighting system electric power of the ith data center of the t-th scheduling period; and represents the load thermal power of the ith data center of the t-th scheduling period.
In embodiment 2, the shared energy storage model based on the data center cluster in embodiment 1 is improved, and the discharging power and the charging power are reserved for the safe operation of the data center cluster so as to process the uncertainty variable wind power prediction error, so that a data center cluster optimization model based on opportunistic constraint planning is established.
Model 12: the data center cluster optimization model based on opportunistic constraint planning is as follows:
Among the constraints of the model 11, the constraints are:
The improvement is as follows:
Under the condition that the remaining constraint conditions in the model 11 need to be satisfied, adding constraints:
wherein:
Representing a daily cost of the u-th data center cluster based on the opportunity constraint planning; CRS u represents the u-th data center cluster spare cost; Representing the standby charging power of a ith data center cluster in a ith scheduling period; Representing the standby discharge power of a ith scheduling period and a ith data center cluster; beta is the confidence level; Representing the maximum charging standby electric power of the u-th data center cluster; Represents the maximum discharge reserve electric power of the u-th data center cluster.
In embodiment 3, the risk cost consideration is increased by the model in embodiment 2, so that the safety and reliability of the cluster operation of the data center are further improved. That is, in the daily cost of the data center cluster, further consideration is given to the fact that the standby discharging and charging power just meeting the confidence level requirement is based, so that the energy use demand response can be flexibly adjusted, a scheduling model based on the opportunity constraint target planning is established, and the standby storage power is allowed to deviate from the standby discharging and charging power value just meeting the opportunity constraint to a certain extent. Risk costs come from unsatisfied confidence level requirements.
Model 13: the scheduling model based on the opportunity constraint goal planning is as follows:
Model 13 improves the two probability constraints of model 12 to the following:
on the basis of satisfying the remaining constraints of the model 12, the following constraints are added:
wherein:
l load and l wind are the unit load loss cost and the unit wind abandon cost respectively; And The absence between the actual standby charge/discharge power and the standby charge/discharge power that just meets the confidence level requirements, respectively. CR u represents the risk cost of the u-th data center cluster.
Further, according to the following reasoning:
For the opportunity constraint g n (x, k) expressed in Pr { g n(x,k)≤bn}≥βn, if and only if it can be expressed as g n (x, k) =h (x) -k, it can be converted into deterministic form h (x) -b n≤Vβ, Wherein: x is a decision variable; kappa is a random variable; b n is the value of target n, which is constant; beta n is the confidence level of constraint n; is the inverse cumulative distribution function of the random variable.
The opportunistic constraints of the two probability forms can be converted into two deterministic forms to reduce the solving difficulty of the model 11, and the converted result is as follows:
wherein: Is the electric power with the corresponding confidence level (1-beta) obtained by the inverse cumulative distribution function.
The same parameters referred to in examples 1-3 have the same meaning.
In embodiment 4, a data center cluster DCC and an integrated shared energy storage system SIESS are included. DCC contains three DCs, hereinafter denoted DC1, DC2 and DC 3. A day is divided into 24 time periods. The energy price of the integrated energy system and the energy service price of the integrated shared energy storage system are shown in fig. 2 and 3. The wind power output prediction curves of the wind turbine generators in the three DCs are shown in FIG. 4. Each DC contains 10 batch servers, each aggregating 10000 sub-servers with a CPU frequency of 3.4GHz. Details of the sub-servers are shown in tables 1-3. The upper limit of server utilization is set to 90%. There are three types of computing tasks assigned to the DC, the details of which are shown in Table 4. The energy consumption component parameters in DC are shown in Table 5. The confidence level β is set to 0.95. Table 6 shows the day costs of DCC operators in five cases.
Table 1 server information at DC1
Table 2 server information at DC2
Table 3 server information at DC3
Table 4 data center calculation job information
Table 5 data center energy element information
Table 6 day cost of DCC operators in five cases
Case 1: the method does not participate in demand response, does not purchase shared energy storage service, does not perform internal energy balance, and takes the scheduling result as a reference value. DCC is the most costly day when it does not participate in demand responses. Case 2: participate in demand response, do not purchase shared energy storage service, do not perform internal energy balancing. The scheduling results indicate that the computing tasks are redistributed by participating in the demand response. More calculation tasks are distributed in the period with larger wind power generation, so that the air discarding quantity is reduced. The purchase energy cost and the abandoned wind cost of the data center cluster are respectively reduced by 6.2 percent and 32.5 percent. Case 3: and participating in demand response, not purchasing shared energy storage service, and performing internal energy balance. Case 4: and participating in demand response, purchasing shared energy storage service, and not performing internal energy balance. In cases 3 and 4, the wind curtailment cost is further reduced to zero, which indicates that both the internal energy balance mechanism and the purchase of shared energy storage service contribute to improved wind power utilization. Especially in case 4, the data center cluster operator starts to conduct energy transactions with the integrated shared energy storage system operator. Thereby reducing the energy purchasing cost and increasing the energy storage service cost. In addition, by purchasing the shared energy storage service, the data center cluster operator is able to discharge to the integrated shared energy storage system during peak wind power generation hours to obtain revenue. Case 5: and participating in demand response, purchasing shared energy storage service and balancing internal energy. The data center cluster has the lowest operation daily cost, which means that the shared energy storage method provided by the invention can effectively improve the economic benefit of the data center cluster.
Fig. 5 depicts power supply and power demand conditions in a 24 hour data center cluster. The amount of unbalance in the power supply and the power demand corresponds to the risk of load loss based on the application of the opportunity constraint goal planning. When the data center cluster does not participate in the demand response, the initial computing task allocation of the three DCs is the same. Since the power consumption of the server in DC3 is maximum, the power consumption of the server in DC3 is maximum in all periods without the DCC participating in the demand response. In contrast, after participating in DR, most computing tasks are allocated to DC2 because the servers in DC2 consume the least power. In addition, more computing tasks may be allocated to periods of lower electricity prices, i.e., 0:00 am to 7:00 am. Since the integrated energy system and the integrated shared energy storage system are both power suppliers in case 5, DCC always selects a lower price power supplier for maximum reduction of daily costs. Thus, DCC selects charging from the integrated shared energy storage system in 7:00 a.m., 9:00 a.m. During the remaining period, DCC purchases electrical energy from the integrated energy system.
Fig. 6 compares the leased power of DCC in case of internal energy balance. Each DC independently performs energy transactions with the integrated shared energy storage system operator without internal energy balancing. In the same period, some DCs have surplus wind power, while other DCs have insufficient wind power supply. In order to maintain the internal power balance of the DC, the DC with the surplus wind power is discharged to the integrated shared energy storage system, and the other DC purchases power from the integrated energy storage system or is charged from the integrated shared energy storage system. In fig. 6, at 7 a.m.: 00, DC1 discharges and DC2 and DC3 charge. Therefore, the total leased power of DCC is very large without internal energy balancing. When the DCC performs internal energy balance, the surplus wind power is supplied to the DC where wind power is not supplied enough. DCC operators lease charge and discharge power from the comprehensive shared energy storage system after internal energy complementation. The DCC of fig. 7 has significantly reduced charging lease power compared to fig. 6. The calculation result shows that the proposed internal energy balance mechanism can effectively reduce the leasing power of DCC and promote the in-situ digestion of renewable energy sources.
From the foregoing description of the embodiments, it will be apparent to those skilled in the art that all and/or portions of the methods, systems, and apparatus of the present disclosure may be implemented in software plus necessary general purpose hardware, or may be implemented in special purpose hardware, including application specific integrated circuits, special purpose CPUs, special purpose memories, special purpose components, and the like. Generally, functions performed by computer programs can be easily implemented by corresponding hardware, and specific hardware structures for implementing the same functions can be varied, such as analog circuits, digital circuits, or dedicated circuits. But for the purposes of this disclosure a software program implementation is a preferred embodiment in many more cases.
Although the embodiments of the present invention have been described above with reference to the accompanying drawings, the present invention is not limited to the above-described specific embodiments and application fields, and the above-described specific embodiments are merely illustrative, and not restrictive. Those skilled in the art, having the benefit of this disclosure, may effect numerous forms of the invention without departing from the scope of the invention as claimed.
Claims (10)
1. A data center cluster shared energy storage method, the method comprising the steps of:
Establishing a server model, a hybrid cooling system model, a renewable energy model, a back pressure turbine model and a gas boiler model, thereby obtaining the energy demand of each data center in each period;
Establishing a demand response model, and carrying out energy scheduling in a data center cluster according to the predicted value of the renewable energy generating capacity device of each data center in each period and the energy demand of each data center;
According to the scheduling result, carrying out energy storage planning based on a shared energy storage model of the data center cluster;
The shared energy storage model is a model which aims at minimizing daily energy cost of the data center cluster; the daily energy cost comprises energy purchasing cost, renewable energy cost reduction, energy storage service cost and energy selling benefit; the energy purchasing cost comprises electricity purchasing cost, heat purchasing cost and gas purchasing cost;
The demand response model is expressed as follows:
If it is
If it is
If it is
If it is
Wherein:
U is a data center cluster set, and U is an element in U; t represents a set of scheduling periods; the ith data center of the ith data center cluster in the ith scheduling period is shown to transfer out the calculation task number; i -i represents the set of data centers other than the ith data center, I' being an element in I -i; the ith' data center of the ith data center cluster in the ith scheduling period is transferred to the computing task number; a is a computing task set, a is an element therein; representing the completion time of task a; Arrival time of task a; representing the deferrable time of task a; Representing the processing time length of the task a; representing the maximum deferrable time of task a; a del represents a deferrable task set, a non represents a non-deferrable task set, a int represents an interruptible task set, and a con represents a continuity task set; Representing the time required for task a to calculate; a start instruction amount for the task a representing the t scheduling period; Indicating the completion calculation indicating quantity of the task a in the t scheduling period; a state variable representing execution of task a for the t-th scheduling period; A state variable representing execution of task a for the t-1 th scheduling period; u i,s,t represents the utilization of the S-th server in the I-th data center in the t-th scheduling period, S ser represents a server set, and I represents a data center set;
The shared energy storage model is expressed as follows:
min DCu＝CPIu+CWCu+CESu-BSSu
s.t.CPIu＝CPEIu+CPHIu+CPGIu
CESu＝Ccau+Cpou
wherein:
DC u represents the energy cost per day for the u-th data center cluster; CPI u represents the purchase cost of the nth data center cluster; CWC u represents the wind curtailment cost of the u-th data center cluster; CES u represents the energy storage service cost of the u-th data center cluster; CPEI u represents the purchase cost of the ith data center cluster; CPHI u represents the purchase cost of the u-th data center cluster; CPHI u represents the gas purchase cost of the u-th data center cluster; The power purchasing power of the ith data center cluster of the t scheduling period is represented; EPI t represents the electricity purchase price at the t scheduling period; The heat purchasing power of the ith data center cluster of the t scheduling period is represented; HPI t represents the t-th scheduling period purchase price; The gas purchasing power of a ith data center cluster in the t scheduling period is represented; GPI t gas purchase price in the t scheduling period; l wc represents the unit cost of the waste wind; representing the waste wind power of the ith data center in the t scheduling period; cca u represents the cost of leased storage capacity for the nth data center cluster; cpo u represents the leased energy storage power cost of the u-th data center cluster; Representing the load electric power of the ith data center of the t-th scheduling period; representing the predicted wind power output power value of the ith data center in the t scheduling period; representing the electric power output by the c4 th element in the ith data center of the t scheduling period; The charging power of the nth data center cluster representing the nth scheduling period; the discharge power of the nth data center cluster representing the nth scheduling period; Representing a power value at a t-th scheduling period in an i-th data center; Representing the electric power required by the ith data center cooling system in the t-th scheduling period; representing the power supply system electric power of the ith data center of the t scheduling period; Representing the lighting system electric power of the ith data center of the t-th scheduling period; representing the load thermal power of the ith data center of the t scheduling period; Representing the thermal power output by the c4 th element in the ith data center of the t scheduling period; the endothermic power of the ith data center cluster representing the ith scheduling period; a heat release power of a kth data center cluster representing a kth scheduling period; Representing the thermal power required by the ith data center cooling system in the t scheduling period; BSS u represents the economic benefit of the u-th data center cluster; CE dis represents a price of a discharge acquisition contract signed by the data center cluster and the comprehensive shared energy storage system; CH rel represents the price of an exothermic acquisition contract signed by the data center cluster and the comprehensive shared energy storage system; Representing the fuel gas power input by a c5 th element in an ith data center of a t scheduling period; Δt represents the time interval of two adjacent scheduling periods; c GB represents a collection of gas boilers, C5 is an element in C GB; c BPST represents the back pressure turbine set and C4 is the element in C BPST.
2. The method according to claim 1, characterized in that:
The daily energy costs further include backup costs; the standby costs include standby discharge costs and standby charge costs.
3. The method according to claim 2, characterized in that:
The daily energy costs further include risk costs; the risk costs include the loss of load costs, relinquishing renewable energy costs, meeting confidence level requirements; the load loss cost is equal to the product of the unit load loss cost and the absence between the actual standby charging power and the standby charging power which just meets the confidence level requirement; the relinquishing renewable energy cost is equal to the product of the unit relinquishing renewable energy cost and the absence between the actual standby discharge power and the standby discharge power that just meets the confidence level requirement.
4. The method of claim 1, wherein the energy scheduling is achieved by computing task reassignment; the influence factors of the computing task reassignment include: calculating the delay characteristic and the interruptible characteristic of the task, whether the renewable energy device is generating in a peak period, whether the comprehensive shared energy storage system is in a low price energy price and/or an energy service price.
5. The method of claim 1, wherein the renewable energy generation device comprises a wind turbine, mounted in a data center.
6. The method of claim 5, wherein the wind turbine generator meets the following model when not participating in energy scheduling:
wherein:
Representing the wind power prediction error of the data center i in the t period; n (0, σ 2) represents a normal distribution with a mean of 0 and a variance of σ 2; representing a wind power predicted value of an ith data center in a t scheduling period; Representing the wind power utilization value of the ith data center in the t scheduling period; the wind power wind curtailment value of the ith data center in the t scheduling period is represented; i represents a data center set; t represents a set of scheduling periods;
When internal energy is scheduled, the output model of the wind turbine generator meets the following conditions:
7. The method of claim 1, wherein the energy storage plan comprises: and determining whether to acquire directly used energy, store energy from the comprehensive shared energy storage system and release energy to the comprehensive shared energy storage system according to the price, the energy use condition and the capacity condition of the renewable energy device.
8. The utility model provides a data center cluster sharing energy storage system, the system includes data center cluster, synthesizes shared energy storage system, data center cluster includes data center, there is renewable energy device in the data, its characterized in that:
Establishing a server model, a hybrid cooling system model, a renewable energy model, a back pressure turbine model and a gas boiler model, thereby obtaining the energy demand of each data center in each period;
Establishing a demand response model, and carrying out energy scheduling in a data center cluster according to the predicted value of the renewable energy generating capacity device of each data center in each period and the energy demand of each data center;
According to the scheduling result, carrying out energy storage planning based on a shared energy storage model of the data center cluster;
The shared energy storage model is a model which aims at minimizing daily energy cost of the data center cluster; the daily energy cost comprises energy purchasing cost, renewable energy cost reduction, energy storage service cost and energy selling benefit; the energy purchasing cost comprises electricity purchasing cost, heat purchasing cost and gas purchasing cost;
The energy storage and retrieval plan comprises: determining whether to purchase directly used energy, store energy from the comprehensive shared energy storage system and release energy to the comprehensive shared energy storage system according to the price, the energy use condition and the capacity condition of the renewable energy device; the demand response model is expressed as follows:
If it is
If it is
If it is
If it is
Wherein:
U is a data center cluster set, and U is an element in U; t represents a set of scheduling periods; Expressing the ith data center of the ith data center cluster in the ith scheduling period to transfer out the calculation task number; i -i represents the set of data centers other than the ith data center, I' being an element in I -i; the ith' data center of the ith data center cluster in the ith scheduling period is transferred to the computing task number; a is a computing task set, a is an element therein; representing the completion time of task a; Arrival time of task a; representing the deferrable time of task a; Representing the processing time length of the task a; representing the maximum deferrable time of task a; a del represents a deferrable task set, a non represents a non-deferrable task set, a int represents an interruptible task set, and a con represents a continuity task set; Representing the time required for task a to calculate; a start instruction amount for the task a representing the t scheduling period; Indicating the completion calculation indicating quantity of the task a in the t scheduling period; a state variable representing execution of task a for the t-th scheduling period; A state variable representing execution of task a for the t-1 th scheduling period; u i,s,t represents the utilization of the S-th server in the I-th data center in the t-th scheduling period, S ser represents a server set, and I represents a data center set;
The shared energy storage model is expressed as follows:
min DCu＝CPIu+CWCu+CESu-BSSu
s.t.CPIu＝CPEIu+CPHIu+CPGIu
CESu＝Ccau+Cpou
wherein:
DC u represents the energy cost per day for the u-th data center cluster; CPI u represents the purchase cost of the nth data center cluster; CWC u represents the wind curtailment cost of the u-th data center cluster; CES u represents the energy storage service cost of the u-th data center cluster; CPEI u represents the purchase cost of the ith data center cluster; CPHI u represents the purchase cost of the u-th data center cluster; CPGI u represents the gas purchase cost of the u-th data center cluster; The power purchasing power of the ith data center cluster of the t scheduling period is represented; EPI t represents the electricity purchase price at the t scheduling period; The heat purchasing power of the ith data center cluster of the t scheduling period is represented; HPI t represents the t-th scheduling period purchase price; The gas purchasing power of a ith data center cluster in the t scheduling period is represented; GPI t gas purchase price in the t scheduling period; l wc represents the unit cost of the waste wind; representing the waste wind power of the ith data center in the t scheduling period; cca u represents the cost of leased storage capacity for the nth data center cluster; cpo u represents the leased energy storage power cost of the u-th data center cluster; Representing the load electric power of the ith data center of the t-th scheduling period; representing the predicted wind power output power value of the ith data center in the t scheduling period; representing the electric power output by the c4 th element in the ith data center of the t scheduling period; The charging power of the nth data center cluster representing the nth scheduling period; the discharge power of the nth data center cluster representing the nth scheduling period; Representing a power value at a t-th scheduling period in an i-th data center; Representing the electric power required by the ith data center cooling system in the t-th scheduling period; representing the power supply system electric power of the ith data center of the t scheduling period; Representing the lighting system electric power of the ith data center of the t-th scheduling period; representing the load thermal power of the ith data center of the t scheduling period; Representing the thermal power output by the c4 th element in the ith data center of the t scheduling period; the endothermic power of the ith data center cluster representing the ith scheduling period; a heat release power of a kth data center cluster representing a kth scheduling period; Representing the thermal power required by the ith data center cooling system in the t scheduling period; BSS u represents the economic benefit of the u-th data center cluster; CE dis represents a price of a discharge acquisition contract signed by the data center cluster and the comprehensive shared energy storage system; CH rel represents the price of an exothermic acquisition contract signed by the data center cluster and the comprehensive shared energy storage system; Representing the fuel gas power input by a c5 th element in an ith data center of a t scheduling period; Δt represents the time interval of two adjacent scheduling periods; c GB represents a collection of gas boilers, C5 is an element in C GB; c BPST represents the back pressure turbine set and C4 is the element in C BPST.
9. A data center cluster sharing energy storage device, characterized in that: comprising a memory and a processor, said memory having stored thereon a computer program capable of being loaded by the processor and performing the method according to any of claims 1 to 7.
10. A computer-readable storage medium, characterized by: a computer program being stored which can be loaded by a processor and which performs the method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210322060.XA CN114841510B (en) | 2022-03-29 | 2022-03-29 | Data center cluster shared energy storage method, system, device and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210322060.XA CN114841510B (en) | 2022-03-29 | 2022-03-29 | Data center cluster shared energy storage method, system, device and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114841510A CN114841510A (en) | 2022-08-02 |
| CN114841510B true CN114841510B (en) | 2024-09-13 |
Family
ID=82563851
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210322060.XA Active CN114841510B (en) | 2022-03-29 | 2022-03-29 | Data center cluster shared energy storage method, system, device and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114841510B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107086587A (en) * | 2017-05-24 | 2017-08-22 | 天津大学 | A kind of data center's dominant eigenvalues control method based on Demand Side Response |
| CN108898282A (en) * | 2018-06-06 | 2018-11-27 | 华北电力大学 | Data center resource Optimization Scheduling and computer storage medium |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013126800A1 (en) * | 2012-02-22 | 2013-08-29 | Viridity Energy, Inc. | Facilitating revenue generation from data shifting by data centers |
| US10671509B1 (en) * | 2015-06-02 | 2020-06-02 | Amazon Technologies, Inc. | Simulating storage server configurations |
| CN106485352B (en) * | 2016-09-30 | 2019-06-25 | 国电南瑞科技股份有限公司 | A kind of multiclass power supply generation schedule combination decision-making method |
| CN110363353A (en) * | 2019-07-16 | 2019-10-22 | 厦门大学 | The optimization design and dispatching method and system of a kind of Distributed Integration energy resource system |
| CN111882105B (en) * | 2020-06-15 | 2024-05-28 | 东南大学 | Micro-grid group containing shared energy storage system and day-ahead economic optimization scheduling method thereof |
| CN113867618A (en) * | 2020-06-30 | 2021-12-31 | 华为技术有限公司 | Resource scheduling method, device, equipment, system and medium for energy storage equipment pool |
| CN112966857B (en) * | 2021-02-10 | 2022-09-30 | 合肥工业大学 | Data center multifunctional collaborative optimization method and system |
| CN113722895A (en) * | 2021-08-18 | 2021-11-30 | 国网上海市电力公司 | Comprehensive energy system optimal configuration method based on multi-station fusion |
-
2022
- 2022-03-29 CN CN202210322060.XA patent/CN114841510B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107086587A (en) * | 2017-05-24 | 2017-08-22 | 天津大学 | A kind of data center's dominant eigenvalues control method based on Demand Side Response |
| CN108898282A (en) * | 2018-06-06 | 2018-11-27 | 华北电力大学 | Data center resource Optimization Scheduling and computer storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114841510A (en) | 2022-08-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ding et al. | Integrated stochastic energy management for data center microgrid considering waste heat recovery | |
| Huang et al. | Minimizing electricity costs by sharing energy in sustainable microgrids | |
| CN111008739B (en) | Optimal regulation and control and income distribution method and system for cogeneration virtual power plant | |
| Han et al. | A shared energy storage business model for data center clusters considering renewable energy uncertainties | |
| CN107194514B (en) | Demand response multi-time scale scheduling method for wind power prediction error | |
| Dashtdar et al. | Probabilistic planning for participation of virtual power plants in the presence of the thermal power plants in energy and reserve markets | |
| CN113610311A (en) | Comprehensive energy service provider cooperation operation optimization method considering carbon emission reduction under double-layer cooperative architecture | |
| Cao et al. | UPS node-based workload management for data centers considering flexible service requirements | |
| CN110649594A (en) | Industrial park comprehensive demand response scheduling method based on multi-energy cooperation | |
| CN111445154B (en) | Power market resource source scheduling optimization method, system and equipment | |
| CN111030101B (en) | Clean energy consumption linkage regulation and control method and system based on diversified big data | |
| JP2005102357A (en) | Method and device for preparing start/stop plan of generator, recording medium for recording process program thereof | |
| Han et al. | Waste heat reutilization and integrated demand response for decentralized optimization of data centers | |
| Hayati et al. | A two-stage stochastic optimization scheduling approach for integrating renewable energy sources and deferrable demand in the spinning reserve market | |
| Ali | Development and Improvement of Renewable Energy Integrated with Energy Trading Schemes based on Advanced Optimization Approaches | |
| Liu et al. | P2H Modeling and operation in the microgrid under coupled electricity–hydrogen markets | |
| CN112531785B (en) | Photo-thermal-heat-storage active power distribution network main and auxiliary combined scheduling method under multiple time scales | |
| Peng et al. | Energy-efficient management of data centers using a renewable-aware scheduler | |
| Wang et al. | Day-ahead emission-aware resource planning for data center considering energy storage and batch workloads | |
| CN114841510B (en) | Data center cluster shared energy storage method, system, device and medium | |
| CN110298505A (en) | Integrated energy system dispatching method, device, computer equipment and storage medium | |
| CN116228461A (en) | P2P transaction-based multi-virtual power plant random optimization scheduling method | |
| CN115330015A (en) | Data center participation comprehensive demand response distributed optimization method considering waste heat | |
| CN112102047A (en) | Virtual power plant optimization combination bidding method, device, equipment and storage medium | |
| CN110429662A (en) | A kind of cogeneration units unit commitment formulating method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |