CN110084444B - Cloud data center power load scheduling method considering randomness of natural resources - Google Patents
Cloud data center power load scheduling method considering randomness of natural resources Download PDFInfo
- Publication number
- CN110084444B CN110084444B CN201910445540.3A CN201910445540A CN110084444B CN 110084444 B CN110084444 B CN 110084444B CN 201910445540 A CN201910445540 A CN 201910445540A CN 110084444 B CN110084444 B CN 110084444B
- Authority
- CN
- China
- Prior art keywords
- time
- cloud data
- data center
- load
- 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
Images
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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
-
- 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- 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
Abstract
The invention discloses a cloud data center power load scheduling method considering randomness of natural resources, which comprises the following steps of; predicting the total power load to be scheduled and the output of renewable energy; scheduling time-shifting power loads from time and space scales, if the power load processing capacity under the current scheduling strategy is smaller than the bearing capacity, calculating the electric energy demand and waste heat utilization of each cloud data center, simultaneously judging whether a power system and a thermodynamic system are balanced at the same time, and recording a power load scheduling scheme and the total operation cost under the balanced condition, thereby obtaining all possible scheduling schemes, and taking the scheduling scheme with the lowest total operation cost as a final scheduling scheme; the power load scheduling method can simultaneously ensure the balance of the power system and the balance of the thermodynamic system of each cloud data center, obviously reduce the operation cost of cloud service providers, greatly increase the profit margin and provide effective basis for the sustainable development of enterprises.
Description
Technical Field
The invention relates to the technical field of cloud data center power load scheduling, in particular to a cloud data center power load scheduling method considering randomness of natural resources.
Background
As is well known, cloud computing is developed at a high speed in all countries around the world, and data can be transmitted in the same cloud, which also promotes the expansion and development of cloud data centers. In general, there are hundreds of thousands of servers in a cloud data center, and each server needs 24 × 7 to ensure stable continuous operation, which necessarily requires a large amount of energy consumption. Statistically, the amount of electricity used by the global internet technology industry in 2016 accounts for 3% of the total amount of electricity worldwide, and it is estimated that this proportion doubles every 4 years. A large amount of power consumption brings huge power cost for each internet company in the internet technology industry, and the too high power consumption cost influences the enterprise benefit. Statistically, the energy cost of a large-scale data center in 2016 to 2018 is 1850 hundred million dollars, which accounts for more than 40% of the total cost. How to reduce the energy cost of the cloud data center is very important.
Although the prior art provides some solutions for reducing energy consumption cost of a cloud data center, such as a bvt (rounded Virtual time) scheduling algorithm, a Credit scheduling algorithm, and the like, these solutions may improve energy utilization efficiency to a certain extent, but cannot simultaneously satisfy balance of all cloud data center power systems, so problems that the prior art often has a single power consumption load scheduling mode consideration factor, strong subjectivity, and cannot perform load scheduling from the global consideration of the whole power system.
Therefore, how to objectively and comprehensively schedule the power load of the cloud data center to improve the energy utilization efficiency of the data center and reduce the energy cost of each cloud data center becomes a key point for the technical problem to be solved and the research of the technical problem to be performed all the time by the technical staff in the field.
Disclosure of Invention
The cloud data center power load scheduling method can guarantee balance of supply and demand of the whole power system and can guarantee balance of supply and demand of the whole thermodynamic system at the same time, so that the cloud data center power load can be scheduled comprehensively and objectively, and a plurality of problems of the existing power load scheduling method are thoroughly solved.
In order to achieve the technical purpose, the invention discloses a cloud data center power load scheduling method considering randomness of natural resources, which comprises the following steps;
step 1, forecasting total power utilization loads to be dispatched in a future preset time period by utilizing historical power utilization load data of cloud data centers acquired in advance, and forecasting renewable energy output in the future preset time period by utilizing historical renewable energy output data of the cloud data centers acquired in advance;
step 2, screening out time-shifting power utilization loads from the total power utilization loads to be scheduled; the time-shifting power utilization load represents a power utilization load with the delayed processed time length being greater than a first preset time length;
step 3, on the premise that the time-shifting electric loads are processed in the respective delay processed time, adjusting the time distribution and the regional distribution of the time-shifting electric loads in the future preset time period, so that the processing amount of the time-shifting electric loads is in direct proportion to the real-time output of renewable energy of each cloud data center, and the processing amount of the time-shifting electric loads is in inverse proportion to the real-time electricity price of the region where each cloud data center is located in the future preset time period;
step 4, judging whether the power load handling capacity of each cloud data center is smaller than the respective carrying capacity under the conditions of time distribution and regional distribution of each current time-shifting power load; if yes, executing step 5; if not, returning to the step 3;
step 5, respectively calculating the electric energy demand and the waste heat utilization of each cloud data center in the future preset time period;
step 6, simultaneously judging whether the power supply quantity of the power system of each cloud data center meets the electric energy demand quantity of each cloud data center and whether the waste heat utilization quantity of each cloud data center meets the heat productivity of the thermal system in which each cloud data center is located; if the two conditions are simultaneously satisfied, executing the step 7, otherwise, returning to the step 3;
step 7, recording the current power load scheduling scheme, calculating the total operation cost under the scheduling scheme, and then storing the current power load scheduling scheme and the total operation cost thereof into a scheduling strategy set;
step 8, judging whether the distribution adjusting mode of each time-shiftable power utilization load in the future preset time period is exhausted; if yes, executing step 9; if not, returning to the step 3;
and 9, taking the power load scheduling scheme with the lowest operation total cost in the scheduling strategy set as a final power load scheduling scheme of the cloud data center.
Based on the technical scheme, the power load of the cloud data center is newly redistributed on the space-time scale, namely the power load distribution on the time scale and the power load distribution on the space scale can be simultaneously met, and the power load distribution can be effectively matched with renewable energy sources and electricity prices with peak-valley characteristics, so that the energy utilization efficiency is greatly improved on the premise of ensuring the supply and demand balance of a power system and a thermodynamic system.
Further, in step 3, when the time distribution of each time-shiftable electric load is adjusted, the time-shiftable electric load is transferred to a time period in which the output of the renewable energy source is greater than a first preset value and/or a time period in which the electricity price is lower than a second preset value; and when the regional distribution of the time-shifting electric loads is adjusted, the time-shifting electric loads are transferred to the cloud data center with the renewable energy output larger than the third preset value and/or the cloud data center with the electricity price lower than the fourth preset value.
Further, in step 3, when adjusting the time distribution of the time-shiftable electrical loads, the method further includes the step of completing the processing of each time-shiftable electrical load in the first half of the delay processed time period.
Further, step 2 further comprises the step of screening out non-time-shifting electric loads from the total electric loads to be scheduled;
step 3, adjusting the regional distribution of each non-time-shifting power load in the future preset time period to enable the processing capacity of the non-time-shifting power load to be in direct proportion to the real-time output of the renewable energy of each cloud data center and enable the processing capacity of the non-time-shifting power load to be in inverse proportion to the real-time electricity price of the region where each cloud data center is located in the future preset time period; and the non-time-shifting power utilization load represents a power utilization load with the processing time length less than a second preset time length.
Further, in step 3, when the regional distribution of each non-time-shifting electric load is adjusted, the non-time-shifting electric load is transferred to the cloud data center with the renewable energy output greater than the fifth preset value and/or the cloud data center with the electricity price lower than the sixth preset value.
Further, in step 3, before transferring the time-shiftable electric loads and/or the non-time-shiftable electric loads, each electric load to be transferred is decomposed into a plurality of sub-loads having the same load scale, and the electric load transfer is performed with one sub-load as the minimum scheduling unit.
Further, in step 6, the power system includes a power grid subsystem, a renewable energy power generation subsystem, a non-renewable energy power generation subsystem, and an electric energy storage subsystem.
Further, in step 6, the thermodynamic system includes a cloud data center heat release subsystem, a generator set heat release subsystem and an electric energy storage heat release subsystem.
Further, in step 7, the operation total cost includes a power generation cost, a power purchase cost, a power load scheduling cost, a cloud data center service cost, and a penalty cost of load shedding.
Further, before step 1, the method further includes the step of simultaneously acquiring historical electricity load data and historical renewable energy output data of each cloud data center within a past preset time period.
The invention has the beneficial effects that: the cloud data center power load scheduling method considering the randomness of natural resources provided by the invention can simultaneously ensure the balance of the power system and the balance of the thermodynamic system of each cloud data center, obviously reduce the operation cost of a cloud service provider, greatly increase the profit margin, obviously improve the service quality of the cloud service provider, contribute to enhancing the enterprise strength of the cloud service provider and provide effective basis for the sustainable development of enterprises.
Drawings
Fig. 1 is a flowchart illustrating a method for scheduling time-shifting power consumption load of a cloud data center in consideration of randomness of natural resources.
Fig. 2 is a schematic flow chart of a method for scheduling a total power load to be scheduled in a cloud data center in consideration of randomness of natural resources.
Detailed Description
The following explains and explains the method for scheduling the electrical load of the cloud data center in consideration of randomness of natural resources in detail with reference to the drawings of the specification.
A cloud service provider (such as google, hundredths, Tencent, etc.) may typically manage multiple cloud data centers at different geographic locations, and each cloud data center may perform data transmission with other cloud data centers through an agent; the invention provides a cloud data center power load scheduling method considering randomness of natural resources based on the principle that data exchange and transmission can be carried out among different cloud data centers, and the method is used for newly reallocating power loads on two aspects of time and space so as to obtain a global optimal result, for example, the cost is reduced to the maximum extent; as shown in fig. 1 and 2, the scheduling method specifically includes the following steps.
Step 1, predicting total power utilization load to be scheduled in a future preset time period by using historical power utilization load data of each cloud data center obtained in advance, and predicting renewable energy output in the future preset time period by using historical renewable energy output data of each cloud data center obtained in advance. In order to improve the accuracy of the cloud data center power load scheduling, before step 1, the present embodiment further includes a step of simultaneously acquiring historical power load data and historical renewable energy output data of each cloud data center within a past preset time period.
Step 2, screening out time-shifting power utilization loads from the total power utilization loads to be scheduled; the time-shiftable power consumption load represents a power consumption load with a delayed processed time length longer than a first preset time length, and in this embodiment, the first preset time length may be several hours or one day. Step 2, screening non-time-shifting power utilization loads from the total power utilization loads to be scheduled; the non-time-shifting electric load represents an electric load with a processed time length less than a second preset time length, and in this embodiment, the second preset time length may be several milliseconds or several seconds.
And 3, on the premise that the time-shifting electric loads are processed in the respective delay processed time, adjusting the time distribution and the region distribution of the time-shifting electric loads in the future preset time period, so that the processing amount of the time-shifting electric loads is in direct proportion to the real-time output of the renewable energy of each cloud data center, and the processing amount of the time-shifting electric loads is in inverse proportion to the real-time electricity price of the region where each cloud data center is located in the future preset time period. The process of adjusting the time distribution and the regional distribution of the time-shifting power consumption correspondingly represents the time shifting process and the space shifting process of the time-shifting load, the shifting on the spatial scale means that the power consumption load can be processed on a local data center of an agent device, or can be transmitted to a cloud data center of other regions for processing through network connection by the agent device, and the shifting of the power consumption load in the process can generate corresponding power consumption load shifting cost; the fact that the load can be shifted on a time scale means that the shift can be made within a processable time period because the processable time of the timeshiftable load is long. It should be understood that the time-shiftable power load of the present embodiment may be shifted in time and shifted in space completely at the same time, that is, the time-shiftable power load may be shifted to other time periods for processing, and may also be processed on cloud data centers in other regions.
More specifically, in step 3, when the time distribution of each time-shiftable electric load is adjusted, the time-shiftable electric load is transferred to a time period in which the output of the renewable energy is greater than a first preset value and/or a time period in which the electricity price is lower than a second preset value, that is, a time period in which the output of the renewable energy is higher and/or the electricity price is lower; for example, if the time-varying characteristic of the renewable energy supply amount is used to shift the electric load (or called the workload), assuming that the processing time of the time-shiftable electric load is one day, if the time-shiftable electric load is 0: 00, arriving at a data center, which can be transferred to any time period in one day for processing, in order to better utilize renewable energy and reduce cost, the embodiment transfers time-shifting electricity utilization load to time with more wind power output (early morning time period and late night time period) or time with more photovoltaic output (daytime time period, for example, 8: 00-16: 00) for processing; for another example, if the time-variable characteristic of the spot market electricity price is used to transfer the electricity load, assuming that the processing time of the time-variable electricity load is one day, if the time-variable load is 0: 00 arrive at the data center, it can be transferred to any time period in one day for processing, in order to better utilize the characteristic that the electricity price changes in real time in one day and reduce the cost, the invention transfers the time-shifting electricity load to the time period with relatively low electricity charge for processing.
When the regional distribution of each time-shiftable power load is adjusted, the time-shiftable power load is transferred to a cloud data center with the output of renewable energy larger than a third preset value and/or a cloud data center with the price of electricity lower than a fourth preset value, for example, the power load is transferred by using the change characteristic between renewable energy supply volumes and areas, in order to better utilize renewable energy and reduce cost, the time-shiftable power load is transferred to an area where a data center with more wind power output is located or an area where a data center with more photovoltaic output is located for processing; for another example, the electricity load is transferred by using the characteristics of change between the spot market electricity prices, and in order to better utilize the characteristics of change in different regions and reduce the cost, the time-shiftable electricity load is transferred to the data center in the region with relatively low electricity charge for processing.
As an optimized technical solution, in step 3, when adjusting the time distribution of the time-shiftable power loads, this embodiment further includes a step of completing each time-shiftable power load in the first half of the delayed processed time, and processing in the first 50% of the time significantly reduces the service cost of the data center, and this way can meet the needs of the user as early as possible, i.e., complete the processing of the corresponding task as early as possible.
As shown in fig. 2, in step 3, the regional distribution of each non-timeshiftable power consumption load in the future preset time period is further adjusted, so that the processing amount of the non-timeshiftable power consumption load is directly proportional to the real-time output of the renewable energy source of each cloud data center, and the processing amount of the non-timeshiftable power consumption load is inversely proportional to the real-time electricity price of the region where each cloud data center is located in the future preset time period, in step 3, when the regional distribution of each non-timeshiftable power consumption load is adjusted, the non-timeshiftable power consumption load is transferred to a cloud data center whose renewable energy output is greater than a fifth preset value and/or a cloud data center whose electricity price is lower than a sixth preset value; for example, the electrical load is transferred by using the change characteristic between the renewable energy supply volumes and the areas, in order to better utilize renewable energy and reduce cost, the non-time-shifting electrical load is transferred to the area where the data center with more wind power output is located or the area where the data center with more photovoltaic output is located for processing; for another example, the power load is transferred by using the characteristics of change between the spot market power prices, and in order to better utilize the characteristics of change in different regions and reduce the cost, the embodiment transfers the non-timeshiftable power load to the data center of the region with relatively low power charge for processing.
As an optimized technical solution, in step 3 of this embodiment, before transferring time-shiftable and/or non-time-shiftable electric loads, each electric load to be transferred is decomposed into a plurality of sub-loads having the same load scale, and the electric load is transferred with one sub-load as a minimum scheduling unit, that is, the time-shiftable load is decomposed into small-scale jobs in equal portions, which not only can fully utilize computing resources and reduce computing resource waste on the time and space level, but also can accurately calculate bandwidth occupation, thereby realizing accurate calculation of load transfer cost according to accurate bandwidth occupation.
In addition, in specific implementation, after a user submits a load calculation request, the present embodiment completes power load scheduling by controlling agents in an area where the user is located, collects power loads through the agents, and sends the collected power loads to corresponding cloud data centers under a scheduling policy through the agents.
It should be understood that, on the basis of the disclosure of the present invention, the first preset value, the second preset value, the third preset value, the fourth preset value, the first preset duration, the second preset duration, etc. of the present embodiment can be reasonably and judiciously adjusted according to specific situations.
Step 4, judging whether the power load handling capacity of each cloud data center is smaller than the respective carrying capacity under the conditions of time distribution and regional distribution of each current time-shifting power load, namely judging whether each cloud data center can normally process the distributed power load capacity under the current power load scheduling condition; if so, the distributed electricity utilization load can be normally processed, and then the step 5 is executed; if not, namely the distributed electric load cannot be normally processed, returning to the step 3.
And 5, respectively calculating the electric energy demand and the waste heat utilization of each cloud data center in a future preset time period.
Step 6, simultaneously judging whether the power supply quantity of the power system of each cloud data center meets the electric energy demand quantity of each cloud data center and whether the waste heat utilization quantity of each cloud data center meets the heat productivity of the thermal system in which each cloud data center is located; if both conditions are true, step 7 is executed, otherwise step 3 is returned to. The essence of the judging process in the step is as follows: the power supply of the data center is guaranteed, and the heat utilization requirement in a micro-grid area is met; in the embodiment, when judging whether the power supply amount of the power system of each cloud data center can meet the respective electric energy demand, the method further comprises the step of taking the non-IT electric load as a consideration factor, and particularly, the method also takes the electric power required by a refrigeration system, a lighting system and other switching equipment which are arranged for preventing the server from overheating into consideration, so that the method has the outstanding advantages of more comprehensive consideration of influence factors and the like; in addition, when specific calculation is carried out, the scheme for ensuring the balance of the supply and the demand of the power system is called a feasible solution, and the points for ensuring the balance of the supply and the demand of the thermodynamic system form a feasible domain.
The power system comprises a power grid subsystem, a renewable energy power generation subsystem, a non-renewable energy power generation subsystem (comprising a traditional generator set) and an electric energy storage subsystem, and is used for ensuring stable and uninterrupted electric quantity supply of the cloud data center.
The thermodynamic system comprises a cloud data center heat release subsystem, a generator set heat release subsystem, an electric energy storage heat release subsystem, an electric boiler and the like, wherein the generator set heat release subsystem can comprise a CHP (cyclic redundancy protocol) generator set and the like, the cloud data center heat release subsystem is used for recycling waste heat generated by a data center and improving the energy utilization rate, and particularly, the collected heat and the heat of other heat release subsystems supply heat for surrounding residential areas together.
The power system, the thermodynamic system and the information transmission system of the embodiment jointly form a cloud data center interconnection system in geographic distribution, and the invention can ensure the overall optimal energy utilization condition of the whole cloud data center interconnection system.
And 7, recording the current power load scheduling scheme, calculating the total operation cost under the scheduling scheme, and then storing the current power load scheduling scheme and the total operation cost thereof in a scheduling strategy set. The total operation cost comprises power generation cost, power purchasing cost (from a power grid), power load scheduling cost, cloud data center service cost and penalty cost of load shedding, and the penalty cost of load shedding can comprise penalty cost of load shedding of an electric power system and penalty cost of load shedding of a thermodynamic system.
Step 8, judging whether the distribution adjusting mode of each time-shiftable power utilization load in the future preset time period is exhausted; if yes, all possible power load scheduling schemes are obtained, and step 9 is executed; if not, the method indicates that an available power load scheduling scheme still exists, and then the method returns to the step 3.
And 9, determining a feasible solution (namely, a power system supply and demand balance scheme) with the lowest total cost under a feasible domain (namely, a point of heat power system supply and demand balance), and taking the power load scheduling scheme with the lowest operation total cost in the scheduling strategy set as a final power load scheduling scheme of the cloud data center.
In specific implementation, the embodiment is implemented by means of model division, where the first model is a demand-side model, the second model is an energy supply-side model, the third model is an energy balance model, and the fourth model calculates the running cost of the whole system. The model II supplies energy for the model I, the model III ensures that the energy (including electric energy and heat energy) of the model I and the energy of the model II are balanced in real time, and the model IV restricts the model I and the model II to achieve the optimal cost.
In the description herein, references to the description of the term "the present embodiment," "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "a plurality" means at least two, e.g., two, three, etc., unless specifically limited otherwise.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and simplifications made in the spirit of the present invention are intended to be included in the scope of the present invention.
Claims (10)
1. A cloud data center power load scheduling method considering randomness of natural resources is characterized by comprising the following steps: the scheduling method comprises the following steps;
step 1, forecasting total power utilization loads to be dispatched in a future preset time period by utilizing historical power utilization load data of cloud data centers acquired in advance, and forecasting renewable energy output in the future preset time period by utilizing historical renewable energy output data of the cloud data centers acquired in advance;
step 2, screening out time-shifting power utilization loads from the total power utilization loads to be scheduled; the time-shifting power utilization load represents a power utilization load with the delayed processed time length being greater than a first preset time length;
step 3, on the premise that the time-shifting electric loads are processed in the respective delay processed time, adjusting the time distribution and the regional distribution of the time-shifting electric loads in the future preset time period, so that the processing amount of the time-shifting electric loads is in direct proportion to the real-time output of renewable energy of each cloud data center, and the processing amount of the time-shifting electric loads is in inverse proportion to the real-time electricity price of the region where each cloud data center is located in the future preset time period;
step 4, judging whether the power load handling capacity of each cloud data center is smaller than the respective carrying capacity under the conditions of time distribution and regional distribution of each current time-shifting power load; if yes, executing step 5; if not, returning to the step 3;
step 5, respectively calculating the electric energy demand and the waste heat utilization of each cloud data center in the future preset time period;
step 6, simultaneously judging whether the power supply quantity of the power system of each cloud data center meets the electric energy demand quantity of each cloud data center and whether the waste heat utilization quantity of each cloud data center meets the heat productivity of the thermal system in which each cloud data center is located; if the two conditions are simultaneously satisfied, executing the step 7, otherwise, returning to the step 3;
step 7, recording the current power load scheduling scheme, calculating the total operation cost under the scheduling scheme, and then storing the current power load scheduling scheme and the total operation cost thereof into a scheduling strategy set;
step 8, judging whether the distribution adjusting mode of each time-shiftable power utilization load in the future preset time period is exhausted; if yes, executing step 9; if not, returning to the step 3;
and 9, taking the power load scheduling scheme with the lowest operation total cost in the scheduling strategy set as a final power load scheduling scheme of the cloud data center.
2. The method for scheduling the electrical load of the cloud data center considering the randomness of the natural resources according to claim 1, wherein:
in step 3, when the time distribution of each time-shiftable electric load is adjusted, the time-shiftable electric load is transferred to a time period when the output of the renewable energy source is greater than a first preset value and/or a time period when the electricity price is lower than a second preset value; and when the regional distribution of the time-shifting electric loads is adjusted, the time-shifting electric loads are transferred to the cloud data center with the renewable energy output larger than the third preset value and/or the cloud data center with the electricity price lower than the fourth preset value.
3. The method for scheduling the electrical load of the cloud data center considering the randomness of the natural resources according to claim 2, wherein:
in step 3, when adjusting the time distribution of the time-shiftable electrical loads, the method further includes the step of completing the processing of each time-shiftable electrical load in the first half of the delayed processed time period.
4. The method for scheduling the electrical load of the cloud data center in consideration of the randomness of the natural resources according to claim 3, wherein:
step 2, screening non-time-shifting power utilization loads from the total power utilization loads to be scheduled;
step 3, adjusting the regional distribution of each non-time-shifting power load in the future preset time period to enable the processing capacity of the non-time-shifting power load to be in direct proportion to the real-time output of the renewable energy of each cloud data center and enable the processing capacity of the non-time-shifting power load to be in inverse proportion to the real-time electricity price of the region where each cloud data center is located in the future preset time period; and the non-time-shifting power utilization load represents a power utilization load with the processing time length less than a second preset time length.
5. The method for scheduling the electrical load of the cloud data center in consideration of the randomness of the natural resources according to claim 4, wherein:
in step 3, when the regional distribution of each non-time-shifting electric load is adjusted, the non-time-shifting electric load is transferred to the cloud data center with the renewable energy output larger than the fifth preset value and/or the cloud data center with the electricity price lower than the sixth preset value.
6. The method for scheduling the electrical load of the cloud data center in consideration of the randomness of the natural resources according to claim 5, wherein:
in step 3, before the time-shiftable electric load and/or the non-time-shiftable electric load is transferred, each electric load to be transferred is decomposed into a plurality of sub-loads with the same load scale, and the electric load transfer is performed by taking one sub-load as a minimum scheduling unit.
7. The method for scheduling the electrical load of the cloud data center in consideration of the randomness of the natural resources according to claim 6, wherein:
in step 6, the power system comprises a power grid subsystem, a renewable energy power generation subsystem, a non-renewable energy power generation subsystem and an electric energy storage subsystem.
8. The method for scheduling the electrical load of the cloud data center in consideration of the randomness of the natural resources according to claim 7, wherein:
in step 6, the thermodynamic system comprises a cloud data center heat release subsystem, a generator set heat release subsystem and an electric energy storage heat release subsystem.
9. The method for scheduling the electrical load of the cloud data center considering the randomness of the natural resources according to claim 8, wherein:
in step 7, the total operation cost comprises power generation cost, power purchase cost, power load scheduling cost, cloud data center service cost and penalty cost of load shedding.
10. The method for scheduling the electrical load of the cloud data center in consideration of the randomness of the natural resources according to claim 1, 2 or 9, wherein:
before the step 1, the method further comprises the step of simultaneously acquiring historical electricity load data and historical renewable energy output data of each cloud data center in a past preset time period.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910445540.3A CN110084444B (en) | 2019-05-27 | 2019-05-27 | Cloud data center power load scheduling method considering randomness of natural resources |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910445540.3A CN110084444B (en) | 2019-05-27 | 2019-05-27 | Cloud data center power load scheduling method considering randomness of natural resources |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110084444A CN110084444A (en) | 2019-08-02 |
| CN110084444B true CN110084444B (en) | 2021-05-14 |
Family
ID=67421939
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910445540.3A Active CN110084444B (en) | 2019-05-27 | 2019-05-27 | Cloud data center power load scheduling method considering randomness of natural resources |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110084444B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111550915A (en) * | 2020-05-14 | 2020-08-18 | 郑瀚韬 | Air conditioner intelligent control system based on non-invasive measurement |
| CN112103997B (en) * | 2020-09-04 | 2022-11-04 | 天津大学 | Active power distribution network operation flexibility improving method considering data center adjustment potential |
| CN114967900B (en) * | 2022-06-06 | 2023-07-04 | 北京亿安天下科技股份有限公司 | Method, system, equipment and medium for reducing electricity consumption of data center |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106228258A (en) * | 2016-07-11 | 2016-12-14 | 浙江工业大学 | A kind of meter and the home energy source LAN energy optimal control method of dsm |
| CN108898282A (en) * | 2018-06-06 | 2018-11-27 | 华北电力大学 | Data center resource Optimization Scheduling and computer storage medium |
| CN109742812A (en) * | 2019-03-11 | 2019-05-10 | 长沙理工大学 | A kind of source-lotus-storage coordinated scheduling method improving new energy consumption |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108183500B (en) * | 2017-11-24 | 2020-07-10 | 国网甘肃省电力公司电力科学研究院 | Multi-energy complementary rural micro-energy network capacity optimization configuration method and device |
-
2019
- 2019-05-27 CN CN201910445540.3A patent/CN110084444B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106228258A (en) * | 2016-07-11 | 2016-12-14 | 浙江工业大学 | A kind of meter and the home energy source LAN energy optimal control method of dsm |
| CN108898282A (en) * | 2018-06-06 | 2018-11-27 | 华北电力大学 | Data center resource Optimization Scheduling and computer storage medium |
| CN109742812A (en) * | 2019-03-11 | 2019-05-10 | 长沙理工大学 | A kind of source-lotus-storage coordinated scheduling method improving new energy consumption |
Non-Patent Citations (1)
| Title |
|---|
| "Integrated Stochastic Energy Management for Data Center Microgrid Considering Waste Heat Recovery";Zhaohao Ding等;《2018 IEEE Industry Application Society Annual Meeting（IAS）》;20181129;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110084444A (en) | 2019-08-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Guo et al. | Integrated planning of internet data centers and battery energy storage systems in smart grids | |
| Chen et al. | Fair demand response with electric vehicles for the cloud based energy management service | |
| Liu et al. | A scalable and robust approach to demand side management for smart grids with uncertain renewable power generation and bi-directional energy trading | |
| Deng et al. | Harnessing renewable energy in cloud datacenters: opportunities and challenges | |
| CN110084444B (en) | Cloud data center power load scheduling method considering randomness of natural resources | |
| US8793365B2 (en) | Environmental and computing cost reduction with improved reliability in workload assignment to distributed computing nodes | |
| Naqvi et al. | Metaheuristic optimization technique for load balancing in cloud-fog environment integrated with smart grid | |
| Cioara et al. | Exploiting data centres energy flexibility in smart cities: Business scenarios | |
| US20150081374A1 (en) | Client-selectable power source options for network-accessible service units | |
| JP2015531585A (en) | Method and system for distributing and / or controlling energy flow taking into account constraints on electrical networks | |
| Liu et al. | Restoration of power distribution systems with multiple data centers as critical loads | |
| Wang et al. | Stochastic programming for cost optimization in geographically distributed internet data centers | |
| Camus et al. | Self-Consumption Optimization of Renewable Energy Production in Distributed Clouds | |
| Borowski | Digital Transformation and Prosumers Activities in the Energy Sector | |
| Wang et al. | Day-ahead emission-aware resource planning for data center considering energy storage and batch workloads | |
| Kumar et al. | A survey on demand response in smart power distribution systems | |
| Abaza et al. | Demand-side management-based dynamic pricing within smart grid environment | |
| Chen et al. | Proliferation of small data networks for aggregated demand response in electricity markets | |
| Fonseca et al. | Flexigy Smart-grid Architecture. | |
| Masood et al. | Congestion management in local market considering the payback effect | |
| VC et al. | Management of Unprioritized Loads in an Islanded Microgrid: Algorithms for Scheduling of Historic Load Data and Real-Time Load Management | |
| Zhang et al. | A Cooperation Model of Data Center Cluster and Electricity Retailer Base on Demand Response | |
| Koronen | Demand-response potential of large-scale data centers in Europe 2030 | |
| Huang et al. | A Day-Ahead Renewables-Based Power Scheduling System for Internet of Energy | |
| Atif et al. | Demand side Management for SMART Grid |
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 | ||
| GR01 | Patent grant |