Green data center multi-resource collaborative planning method, computing device and storage medium

Abstract

The application discloses a kind of green data center multi-resource collaborative planning method, computing device and storage medium, the method is executed in computing device, including: in combination with endogenous uncertainty and exogenous uncertainty, the operation benefit of green data center system is obtained, green data center system includes distribution network line, green data center and terminal user, endogenous uncertainty includes terminal user participation demand response willingness, exogenous uncertainty includes renewable energy power generation output and terminal user data demand;Based on the investment cost and operation benefit of green data center system, the objective function of green data center multi-resource collaborative planning is constructed;Constraint condition is generated, constraint condition includes planning phase constraint and running phase constraint;Through objective function, in combination with constraint condition, green data center is carried out multi-resource collaborative planning.

Description

Technical Field

[0001] This invention relates to the field of energy and power, and in particular to a method for multi-resource collaborative planning of green data centers, computing devices, and storage media. Background Technology

[0002] As a crucial component of information infrastructure, data centers have experienced rapid development. Along with the surge in the scale and number of data centers, their enormous electricity consumption has exacerbated carbon pollution. Therefore, the concept of Green Data Center (GDC) has been proposed, aiming to maximize energy efficiency and minimize environmental impact in IT (Information Technology) systems, cooling, lighting, and electrical systems within data centers.

[0003] Currently, there is extensive research on the internal energy management of green data centers. Operating strategies are determined based on key factors affecting the effectiveness of green data centers and the availability of internal resources (such as servers, cooling systems, and distributed renewable energy). However, most existing studies treat green data centers as demand response (DR) resources, modeling and exploring them from a power grid perspective. This often neglects the impact of various uncertainties in green data centers on planning decisions, or uses only simple probability distribution models for description, thus affecting the accuracy of the planning results.

[0004] Especially for end users, data center operations always assume a complete understanding of their needs. In practice, however, end-user willingness to participate in demand-side response for green data centers and the depth of their involvement become key factors in determining investment, construction, and operation. End-user data usage behavior is uncertain, not always following constant statistical patterns, and is influenced by external conditions such as subsidies and user-specific service quality requirements. In green data center planning, current investment decisions can significantly impact the temporal distribution of end-user data demands. Ignoring the dependence of end-user willingness to participate in demand response on planning decisions and policy incentives may lead to suboptimal planning outcomes.

[0005] Therefore, a new green data center multi-resource collaborative planning method is needed to optimize processing. Summary of the Invention

[0006] Therefore, this invention provides a green data center multi-resource collaborative planning scheme in an attempt to solve or at least alleviate the problems mentioned above.

[0007] According to one aspect of the present invention, a method for multi-resource collaborative planning of a green data center is provided. The method includes the following steps: First, obtaining the investment cost and operational benefits of a green data center system, which includes a power distribution network, a green data center, and end users; based on the investment cost and operational benefits, constructing an objective function for multi-resource collaborative planning of the green data center; generating constraints, including planning stage constraints and operational stage constraints; and performing multi-resource collaborative planning of the green data center using the objective function and the constraints.

[0008] Optionally, in the green data center multi-resource collaborative planning method according to the present invention, the step of obtaining the investment cost and operating benefits of the green data center system includes: obtaining the installation costs of the server, energy storage system power, energy storage system capacity, wind turbine, photovoltaic equipment, and cooling equipment respectively to calculate the investment cost of the green data center system; and calculating the operating benefits of the green data center system by combining endogenous uncertainty and exogenous uncertainty.

[0009] Optionally, in the green data center multi-resource collaborative planning method according to the present invention, the endogenous uncertainty includes the willingness of end users to participate in demand response, and the exogenous uncertainty includes renewable energy power generation output and end user data demand.

[0010] Optionally, in the green data center multi-resource collaborative planning method according to the present invention, the investment cost is determined by the following formula:

[0011] C Inv =k ser c ser n ser +k ess c ess P Ness +k ess_e c ess_e E Ness +

[0012] k wt c wt P Nwt +k pv c pv P Npv +k cool c cool P Ncool

[0013] Among them, C Inv k represents the investment cost. ser k ess k ess_e k wt k pv k coolc represents the capital recovery factor of the server, energy storage system power, energy storage system capacity, wind turbine, photovoltaic equipment, and refrigeration equipment, respectively. ser c ess c ess_e c wt c pv c cool These represent the unit installation costs of the server, energy storage system power, energy storage system capacity, wind turbine, photovoltaic equipment, and refrigeration equipment, respectively. ser P represents the number of servers installed. Ness E Ness P Ncool These represent the rated power of the energy storage system, the rated capacity of the energy storage system, and the rated installed capacity of the refrigeration equipment, respectively. P Nwt P Npv These represent the installed capacity of wind turbines and photovoltaic equipment, respectively.

[0014] Optionally, in the green data center multi-resource collaborative planning method according to the present invention, the operational efficiency is determined by the following formula:

[0015]

[0016] Among them, Λ Ope ρ represents operational efficiency. y,s This represents the probability of scenario s occurring during contract period y. Let Y represent the operating revenue, operating cost, carbon emission cost, and incentive cost under scenario s in contract period y, respectively. Y represents the set of contract periods in which the data load cluster participates in demand response, and S represents the set of different scenarios.

[0017] Optionally, in the green data center multi-resource collaborative planning method according to the present invention, the operational benefits under scenario s in the contract period y are... Determined by the following formula:

[0018]

[0019] Where θ represents the number of days in a contract period y. δ represents the total data load of the green data center after the demand response of cluster i within time t in scenario s during the contract period y. Data This represents the unit price of data services provided by the green data center, where T represents the set of times of day, and I represents the set of data load clusters.

[0020] Optionally, in the green data center multi-resource collaborative planning method according to the present invention, the operating cost under scenario s in the contract period y is... Determined by the following formula:

[0021]

[0022] Among them, c serm c essm c wtm c pvm c coolm Let n represent the unit annual operation and maintenance cost of servers, energy storage systems, wind turbines, photovoltaic equipment, and refrigeration equipment, respectively. ser E represents the number of servers installed. Ness P Ncool P represents the rated capacity of the energy storage system and the rated installed capacity of the refrigeration equipment, respectively. Nwt P Npv Let represent the installed capacity of wind turbines and photovoltaic equipment, respectively, and θ represent the number of days in a contract period y. This represents the power consumption of the green data center from the grid during time t in scenario s within the contract period y. Δt represents the electricity price in the distribution network at time t, and Δt represents time.

[0023] Optionally, in the green data center multi-resource collaborative planning method according to the present invention, the step of constructing the objective function of green data center multi-resource collaborative planning based on investment cost and operational benefits includes: calculating the difference between operational benefits and investment cost, and maximizing the difference as the objective function of green data center multi-resource collaborative planning.

[0024] Optionally, in the green data center multi-resource collaborative planning method according to the present invention, the planning stage constraints include equipment installation constraints and incentive price constraints. The equipment installation constraints include server quantity constraints, energy storage system power and capacity constraints, distributed renewable energy installation capacity constraints, and cooling equipment installation constraints.

[0025] Optionally, in the green data center multi-resource collaborative planning method according to the present invention, the operational constraints include equipment operation constraints, grid-side constraints, and demand-side constraints. The equipment operation constraints include energy storage system operation constraints, distributed renewable energy operation constraints, cooling equipment operation constraints, and power balance constraints.

[0026] According to another aspect of the present invention, a computing device is provided, comprising: at least one processor; and a memory storing program instructions, wherein the program instructions are configured to be executed by the at least one processor, the program instructions including instructions for performing the green data center multi-resource collaborative planning method as described above.

[0027] According to another aspect of the present invention, a readable storage medium storing program instructions is provided, which, when read and executed by a computing device, causes the computing device to perform the green data center multi-resource collaborative planning method as described above.

[0028] According to the green data center multi-resource collaborative planning scheme of the present invention, based on the correlation between system planning and operation, and comprehensively considering the impact of endogenous and exogenous uncertainties on the effectiveness of decision-making and planning, the investment cost and operational benefits of the green data center system are obtained. An objective function for the green data center multi-resource collaborative planning is constructed, and constraints are generated. Through the objective function and the constraints, multi-resource collaborative planning of the green data center is performed. In the above technical solution, multi-resource collaborative planning of the green data center optimizes the economy and low-carbon nature of green data center planning and operation, and explores the potential value of end-user flexibility in promoting the efficiency improvement of green data centers from a medium- and long-term planning perspective. Attached Figure Description

[0029] To achieve the foregoing and related objectives, certain illustrative aspects are described herein in conjunction with the following description and accompanying drawings. These aspects indicate various ways in which the principles disclosed herein may be practiced, and all aspects and their equivalents are intended to fall within the scope of the claimed subject matter. The foregoing and other objectives, features, and advantages of this disclosure will become more apparent from the following detailed description, taken in conjunction with the accompanying drawings. Throughout this disclosure, the same reference numerals generally refer to the same parts or elements.

[0030] Figure 1 A structural block diagram of a computing device 100 according to an embodiment of the present invention is shown; and

[0031] Figure 2 A flowchart of a green data center multi-resource collaborative planning method 200 according to an embodiment of the present invention is shown. Detailed Implementation

[0032] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0033] Figure 1 A structural block diagram of a computing device 100 according to an embodiment of the present invention is shown.

[0034] like Figure 1 As shown, in the basic configuration 102, the computing device 100 typically includes system memory 106 and one or more processors 104. Memory bus 108 can be used for communication between processor 104 and system memory 106.

[0035] Depending on the desired configuration, processor 104 can be any type of processor, including but not limited to: microprocessor (UP), microcontroller (UC), digital information processor (DSP), or any combination thereof. Processor 104 may include one or more levels of cache such as L1 cache 110 and L2 cache 112, processor core 114, and registers 116. Example processor core 114 may include an arithmetic logic unit (ALU), floating-point unit (FPU), digital signal processing core (DSP core), or any combination thereof. Example memory controller 118 may be used with processor 104, or in some implementations, memory controller 118 may be an internal part of processor 104.

[0036] Depending on the desired configuration, system memory 106 can be any type of memory, including but not limited to volatile memory (such as RAM), non-volatile memory (such as ROM, flash memory, etc.), or any combination thereof. System memory 106 may include operating system 120, one or more applications 122, and program data 124. In some embodiments, application 122 may be arranged to execute instructions on the operating system using program data 124 by one or more processors 104.

[0037] The computing device 100 also includes a storage device 132, which includes a removable storage device 136 and a non-removable storage device 138.

[0038] The computing device 100 may also include a storage interface bus 134. The storage interface bus 134 enables communication from storage devices 132 (e.g., removable storage 136 and non-removable storage 138) to the basic configuration 102 via a bus / interface controller 130. At least a portion of the operating system 120, application 122, and program data 124 may be stored on the removable storage 136 and / or the non-removable storage 138, and loaded into system memory 106 via the storage interface bus 134 when the computing device 100 is powered on or when application 122 is to be executed, and executed by one or more processors 104.

[0039] The computing device 100 may also include an interface bus 140 that facilitates communication from various interface devices (e.g., output devices 142, peripheral interfaces 144, and communication devices 146) to the basic configuration 102 via a bus / interface controller 130. Example output devices 142 include a graphics processing unit 148 and an audio processing unit 150. They may be configured to facilitate communication with various external devices such as displays or speakers via one or more A / V ports 152. Example peripheral interfaces 144 may include a serial interface controller 154 and a parallel interface controller 156, which may be configured to facilitate communication with external devices such as input devices (e.g., keyboards, mice, pens, voice input devices, touch input devices) or other peripherals (e.g., printers, scanners, etc.) via one or more I / O ports 158. Example communication devices 146 may include a network controller 160, which may be arranged to facilitate communication with one or more other computing devices 162 via a network communication link through one or more communication ports 164.

[0040] A network communication link can be an example of a communication medium. A communication medium can typically be embodied in a modulated data signal, such as a carrier wave or other transmission mechanism, and can include any information delivery medium. A “modulated data signal” can be a signal whose data set, or its modifications, can be encoded as information within the signal. As a non-limiting example, a communication medium can include wired media such as wired networks or leased lines, and various wireless media such as sound, radio frequency (RF), microwave, infrared (IR), or other wireless media. The term “computer-readable medium” as used herein can include both storage media and communication media.

[0041] The computing device 100 can be implemented as a personal computer, including desktop and laptop computer configurations. Of course, the computing device 100 can also be implemented as part of a small-sized portable (or mobile) electronic device, such as a cellular phone, digital camera, personal digital assistant (PDA), personal media player device, wireless network browsing device, personal head-mounted device, application-specific device, or a hybrid device that may include any of the above functions. It can even be implemented as a server, such as a file server, database server, application server, and web server. The embodiments of the present invention do not limit this.

[0042] In an embodiment of the invention, computing device 100 is configured to execute a green data center multi-resource collaborative planning method 200 according to the invention. An application 122, arranged on an operating system, includes multiple program instructions for executing method 200. These instructions can instruct processor 104 to execute method 200 of the invention, so that computing device 200 can implement multi-resource collaborative planning of a green data center by executing method 200 of the invention.

[0043] Figure 2 A flowchart of a green data center multi-resource collaborative planning method 200 according to an embodiment of the present invention is shown. The green data center multi-resource collaborative planning method 200 can be executed in a computing device (e.g., the aforementioned computing device 100).

[0044] like Figure 2 As shown, method 200 begins with step S210. In step S210, the investment cost and operational benefits of a green data center system are obtained. The green data center system includes power distribution lines, a green data center, and end users. According to an embodiment of the present invention, the green data center includes distributed renewable energy (such as wind turbines and photovoltaic equipment), energy storage systems, IT equipment represented by servers, and auxiliary equipment such as cooling systems.

[0045] This system exhibits significant differences from traditional data centers in the following aspects:

[0046] First, there is a deep integration of physical and information elements. Green data centers require physical devices such as servers to process data tasks in the information domain, while the nature of the information domain data load (such as type, capacity, and distribution) will conversely affect the status of various physical devices.

[0047] Secondly, there is the intertwining of information and social elements. Data center operations design optimal strategies to incentivize end-user demand responses based on the overall system's operational goals, while end-users, in turn, respond with actions that best serve their own interests based on relevant incentive information. Because the interactions between different stakeholders contain personalized preferences and desires, necessary social elements are introduced into the collaborative planning and design of the system, resulting in a close intertwining of the information and social domains.

[0048] Third, social and physical systems are interdependent. Users' social domain needs (information needs) are not only the core elements affecting the operation of physical domain equipment in data centers and power distribution systems, but also the starting point and end point driving their integrated utilization; while physical domain resources are the foundation of social domain interaction activities, and their functional attributes (equipment type, capacity, etc.) will in turn affect the activity mode of the social domain (users' data usage behavior).

[0049] Based on the above analysis, the overall performance of a green data center system is constrained and influenced by multiple factors, including information, physical, and social factors. Therefore, the multi-resource collaborative planning of a green data center must be based on the CPSS (Cyber-Physical-Social System) framework to reflect the actual operating conditions of the system.

[0050] In a green data center system, the data center operator is responsible for the investment planning and operation management of the green data center, as well as the pricing and incentive design for data services. Based on the data load demand within a certain area, the system coordinates and schedules the operation of multiple resources, including power sources, loads, and storage, to fully leverage the cleanliness of distributed renewable energy generation and the flexibility of end-user data needs. This achieves optimal system configuration and promotes carbon emission reduction while maintaining optimal economic efficiency.

[0051] For planning elements, the multi-domain coupling characteristics of CPSS bring diverse constraints to the collaborative planning model, requiring comprehensive consideration of constraints such as system security, data service quality, and end-user satisfaction. Finally, for green data center systems, due to the randomness and volatility of renewable energy power generation, end-user data demands, and differences in end-user willingness to participate in demand response, the aforementioned endogenous and exogenous uncertainties need to be fully considered during the modeling process to ensure the effectiveness and feasibility of the final collaborative planning scheme.

[0052] The multi-resource collaborative planning problem of green data centers is constructed as a two-stage stochastic programming model to achieve optimal decisions in investment planning and operation scheduling. In the first stage, the goal is to minimize the annual system investment cost by optimizing system equipment capacity and data service pricing and incentive design. The second stage, considering internal and external uncertainties, aims to maximize the annual system operating efficiency by optimizing the operation control strategies for the green data center at different time periods (including information load data processing volume, system power purchase, server start / stop status, system equipment output, etc.). To assess the environmental friendliness of green data centers, carbon emission costs are incorporated into the quantification process of operating costs.

[0053] When modeling the characteristics of green data centers, three aspects should be considered comprehensively: equipment energy consumption (physical domain), data load (information domain), and incentive-based data demand response (social domain). The energy consumption of a green data center mainly consists of servers, cooling equipment, and other auxiliary equipment, with the former two typically accounting for over 90% of the total power consumption. To simplify the problem, the sum of the power consumption of servers and cooling equipment is approximated as the total power consumption of the green data center.

[0054] Server energy consumption is primarily determined by server utilization, taking into account server power parameters and the number of servers currently active. For simplicity, it is assumed that all servers in the green data center have the same model and configuration. Under the scheduling of the cloud operation management platform, the information load at any given time will be evenly distributed among the active servers. The server power consumption characteristics under different operating conditions in the green data center can be expressed as follows:

[0055]

[0056]

[0057]

[0058]

[0059] in, P represents the server's power consumption, CPU (Central Processing Unit) utilization, and number of servers started within time t under scenario s in the contract period y. ser_peak P ser _idle These represent the peak power consumption and silent power consumption of a single server, respectively. μ represents the total data load (i.e., data load arrival rate) of the green data center after the demand response of load cluster i within time t under scenario s in the contract period y. ser U represents the service rate of a single server. P,max θ represents the upper limit of server utilization. ser n represents the server's redundancy factor. ser This indicates the number of servers installed, and I represents the set of data load clusters.

[0060] To ensure ambient temperature and meet the operating requirements of servers and other equipment in green data centers, a large number of cooling devices are required, resulting in significant power consumption. A "load factor" is used to model the power consumption of cooling devices. This parameter is defined as the ratio of the server operating power to the power consumed by the cooling devices in a green data center, reflecting the overall energy efficiency of the cooling equipment.

[0061]

[0062] In the formula, λ represents the power consumption of the cooling device within time t under scenario s in the contract period y. t LF This indicates the overall load factor of the refrigeration equipment.

[0063] The demand responsiveness of green data centers primarily stems from the flexibility of data load scheduling. Based on response methods, data loads in green data centers can be categorized into three types: Interruptible Workload (IW), Shiftable Workload (SW), and Rigid Data Load (RDL).

[0064] Interruptible workloads refer to data loads whose data processing volume can be flexibly reduced within a certain time range. For green data centers, interruptible workloads mainly include video software, etc. Equations (6) to (8) respectively represent the capacity ratio of interruptible workloads, the demand balance before and after response, and the total demand interruption constraint, as shown below:

[0065]

[0066]

[0067]

[0068] in, These represent the load arrival rate of IW before demand response, the load arrival rate of IW after demand response, the total load arrival rate before demand response, and the IW load reduction amount, respectively, within time t of scenario s in contract period y. This indicates the willingness of load cluster i to participate in demand response under scenario s during contract period y. This represents the IW percentage of load cluster i.

[0069] Portable workloads refer to data loads with a fixed total workload within a certain time frame, but flexible and variable data processing volume in different time periods. For green data centers, portable workloads mainly include batch processing loads such as image processing and offline big data analysis. Based on the characteristics of portable workloads, data center operations can postpone the processing time of portable workload tasks to a certain extent within the allowable scheduling period of the load. Equations (9) to (12) respectively represent the capacity ratio of portable workloads, demand balance before and after response, time transfer characteristic constraints, and total demand transfer constraints, as shown below:

[0070]

[0071]

[0072]

[0073]

[0074] in, These represent the load arrival rates of load cluster i SW before and after demand response within time t in scenario s during contract period y. These represent the SW transfer amounts of load cluster i under scenario s during contract period y, between time t and t′, and between time t″ and t, respectively. This represents the cutoff time of load cluster i SW within time t. This indicates the percentage of SWs in load cluster i.

[0075] Furthermore, some critical data loads have strict requirements on both the time and quality of information processing; these are referred to here as rigid loads. Such loads lack operational adjustability and therefore cannot participate in scheduling operations as demand response resources. In practical applications, typical rigid loads include remote real-time medical care and power security control signal processing. The operating characteristics of rigid loads can be expressed as follows:

[0076]

[0077] in, This represents the load arrival rate of the load cluster iRW within time t under scenario s in the contract period y.

[0078] At this point, the total data load of the green data center in each time period after demand response can be calculated. Represented as:

[0079]

[0080] Users' data usage choices are influenced by multiple personalized factors, including economic factors, environmental factors, habitual preferences, and subjective intentions. The concept of "user engagement" represents the willingness of users in different data load clusters to participate in demand response under external incentives. For simplicity, it is assumed that within the same load cluster i, IW and SW have the same willingness to participate.

[0081] According to one embodiment of the present invention, the investment cost and operational benefits of a green data center system can be obtained in the following manner. First, the installation costs of the server, energy storage system power, energy storage system capacity, wind turbine, photovoltaic equipment, and cooling equipment are obtained respectively to calculate the investment cost of the green data center system.

[0082] In this implementation, the investment cost is determined using the following formula:

[0083]

[0084] Among them, C Inv k represents the investment cost. ser k ess k ess_e k wt k pvk cool c represents the capital recovery factor of the server, energy storage system power, energy storage system capacity, wind turbine, photovoltaic equipment, and refrigeration equipment, respectively. ser c ess c ess_e c wt c pv c cool P represents the unit installation cost of the server, energy storage system power, energy storage system capacity, wind turbine, photovoltaic equipment, and refrigeration equipment, respectively. Ness E Ness P Ncool P represents the rated power of the energy storage system, the rated capacity of the energy storage system, and the rated installed capacity of the refrigeration equipment, respectively. Nwt P Npv These represent the installed capacity of wind turbines and photovoltaic equipment, respectively. The capital recovery factor can be calculated from the equipment's lifespan and the discount rate.

[0085] Then, combining endogenous and exogenous uncertainties, the operational benefits of the green data center system are calculated. Endogenous uncertainties include end-user willingness to participate in demand response, while exogenous uncertainties include renewable energy power generation output and end-user data demands. The uncertainty of data parameters leads to the dynamic nature of relevant decision boundary conditions. Therefore, to achieve multi-resource collaborative planning for green data centers, the potential impact of these uncertainties on the effectiveness of multi-resource collaborative planning needs to be fully considered during the modeling process.

[0086] Considering geographical factors and the maturity of relevant technologies, this paper uses wind power and photovoltaic power generation as examples to illustrate the modeling process for the uncertainty of renewable energy power generation output. Wind power output is mainly affected by environmental factors such as wind speed and air density. Wind speed exhibits randomness; in long-term planning, a two-parameter Weibull distribution model is typically used to describe this randomness, thereby determining the predicted output power of the wind turbine as follows:

[0087]

[0088] P Nwt =n wt P Nwt_Sta (17)

[0089] in, v y,t,s V represents the predicted output power and wind speed of the wind turbine at time t within scenario s during the contract period y. in v out v rate These represent the inlet velocity, outlet velocity, and rated velocity of the fan, respectively. wt P represents the number of wind turbines installed. Nwt_Sta"or" indicates the unit installed capacity of the wind turbine.

[0090] Photovoltaic power generation is mainly affected by environmental factors such as solar radiation and temperature. The Beta distribution is widely used in this study to simulate the probability distribution of solar irradiance, thereby determining the output power of the photovoltaic power generation unit as follows:

[0091]

[0092] P Npv =n pv P Npv_Sta (19)

[0093] in, ν represents the predicted output power and irradiance of the photovoltaic equipment at time t in scenario s within the contract period y. pv_Cap ν pv_Sta n represents standard illuminance and unit illuminance of photovoltaic equipment, respectively. pv P represents the number of photovoltaic devices installed. Npv_Sta This indicates the unit installed capacity of photovoltaic equipment.

[0094] The data demands of end users are poorly predictable and highly uncertain. Generally, the randomness of various data load clusters in different time periods can be described by the truncated Gaussian distribution.

[0095] Whether end-users are willing to participate in demand-side response is a key factor determining the investment and operational effectiveness of green data centers, and it is influenced by factors such as user preferences and policy incentives. In practice, because this uncertainty depends on the planning decisions for green data centers, end-user willingness to participate in demand response is an endogenous uncertainty of the system, which can be described using the regret-matching (RM) mechanism.

[0096] The benefit to end-user load cluster i participating in the data center operation demand response plan can be considered as the difference between the subsidy for participating in the demand response plan and the user benefit loss, expressed as:

[0097]

[0098] IW's participation in demand response provides compensation benefits (i.e., subsidy incentives). This is a tiered incentive compensation system with multiple segments to achieve differentiated pricing in segmented markets and reasonably subsidize different end users. The compensation benefits of SW's participation in demand response (i.e., subsidy incentives) The actual delay time is linearly positively correlated with the user benefits of IW and SW participating in demand response. The user benefits loss is positively correlated with the amount of IW reduction and SW transfer, respectively.

[0099]

[0100]

[0101]

[0102]

[0103] Among them, W y,i,s This represents the revenue of participating in demand response load cluster i under scenario s during contract period y. Let represent the subsidy incentives for load cluster iIW, the user benefit deduction for IW, the subsidy incentives for SW, the user benefit deduction for SW, and the actual latency of SW within time t under scenario s in the contract period y, respectively. These represent the unit prices for the 1st, 2nd, ..., nth tiers of the IW subsidy for green data centers. δ represents the data volume boundary value for each step interval. SW,Inc α represents the unit price of the green data center subsidy for SW. IW,Loss α SW,Loss These represent the user benefit reduction coefficients for IW and SW, respectively. θ represents the maximum delay time of the load cluster iSW, θ represents the number of days in a contract period y, for example, 90 days, and T represents the set of all times in a day.

[0104] In demand response (RM), users repeatedly adjust their behavior based on a transition probability distribution. This distribution depends on the "regret" of not adopting a new strategy earlier, following a "reflex-response" behavioral paradigm. When a green data center is operational, end-users have virtually no way of knowing in advance the potential benefits of participating in demand response. Therefore, to maximize benefits, users can only estimate the profitability of demand response based on their past strategy choices and the resulting gains, and use this information to select their next stage of participation. Based on these behavioral characteristics, the RM mechanism can appropriately describe users' willingness to participate in demand response.

[0105] Using the RM method, with finite set Ω B Let Ω represent the set of assumed policies for data load cluster I, where each policy represents a possible user intention during operation. Normalizing user intentions into discrete values ​​at equal intervals of 0.1, then Ω... B= {0, 0.1, ..., 0.9, 1}. Here, strategy "1" indicates that load cluster i will utilize its full potential in demand response during the contract period y, while strategy "0" indicates that load cluster i will completely withdraw from the demand response plan and will not participate in any demand response. Each contract period y corresponds to one contract signing, and the load cluster i can change its own demand response strategy.

[0106] According to the definition of the RM mechanism, for any contract period y, the strategy adopted belongs to Ω. B The probability depends on the user's regret score, as shown below:

[0107]

[0108]

[0109] Among them, M y,i,s (·), G y,i,s (·) represent the regret function and regret quantification function of load cluster i under scenario s during contract period y, respectively. Let b and b' represent two different strategies for load cluster i under scenario s (i.e., two different intentions of load cluster i to participate in demand response, distinguished by the subscripts b and b'), and max represents finding the maximum value. W represents the selection strategy (i.e., historical decision) of load cluster i within a historical time τ before the contract period y under scenario s. i (·) represents the user benefit function of load cluster i, which can be calculated with reference to equation (20).

[0110] The "regret" of end-user participation in demand response is defined as the degree of regret that arises after all outcomes have occurred within a certain time period, if the previous choice... (This has already been achieved) being replaced by a different option, resulting in a potential increase in benefits. If the individual's strategy... The expected profit is higher than the previously selected strategy. If the profit is too high, then users will regret their decision and quantify their regret; otherwise, they will have no regrets, i.e., the regret level is 0.

[0111] Based on the above definitions, assuming that the load cluster i adopts a strategy during the contract period y under scenario s... Then, during the contract period y+1, the strategy is switched to the new strategy. probability distribution function This can be described using a decision dependency model as follows:

[0112]

[0113] in, These represent the load cluster i selection strategies under scenario s in contract period y+1. The probability, λ i It represents a predefined constant for load cluster i, which guarantees a total probability sum of 1.

[0114] Equation (27) shows that, for each runtime segment, the end user can retain their previous policy. Or from Ω B Choose a new strategy The probability of each strategy being chosen is a function M of regret. y,i,s (·), the greater the regret degree, the higher the probability of being selected, which better describes the behavior of information load under planning decisions and contract incentives.

[0115] In practical optimization, given the historical decisions of the data load, the probability distribution of each load cluster i∈I under scenario s at contract period y can be determined using the following method.

[0116] 1) Potential options for loading data loads into clusters to participate in demand response Ω B Historical decision User benefit function of load cluster i The model parameters, and from Ω B Randomly select the initial strategy Initialize τ = 1;

[0117] 2) Quantitative calculation based on formula (20) And record, for Similarly, quantitative calculation is performed according to formula (20). And record it;

[0118] 3) Determine if τ = y. If yes, continue; otherwise, let τ = τ + 1 and return to step 2).

[0119] 4) By aggregating previous output results According to equations (25) and (26), calculate M. y,i,s (·);

[0120] 5) Determine the probability distribution function according to equation (27). And terminated.

[0121] Based on the above description, the uncertainties associated with green data centers can generally be categorized into two aspects: exogenous uncertainty and endogenous uncertainty. The former has fixed statistical characteristics and can be represented by a predetermined probability model, while the latter's probability evolves continuously with data center operational decisions. Therefore, the relevant scenarios regarding uncertainties in green data center planning can be represented by a matrix as follows:

[0122]

[0123] Among them, Ψ y,s This represents a hypothetical scenario under scenario s within the contract period y, comprising various uncertainties, indicating the various states that the green data center may experience during operation. These represent the actual output power of the wind turbine and photovoltaic equipment at time t within scenario s during the contract period y.

[0124] Considering the probability distribution of uncertain inputs, the Monte Carlo simulation method was used to generate the original scenario set for green data center planning and operation. Due to the large number of generated scenarios and their similar statistical characteristics, K-means clustering was further used to reduce the model size and complexity, improving the solution efficiency of the planning model while fully preserving the statistical characteristics of the original scenario set. The specific implementation process involving the RM mechanism and the construction of anticipated probability scenarios can be implemented with reference to conventional algorithms and will not be elaborated here.

[0125] According to one embodiment of the present invention, the operational efficiency is determined by the following formula:

[0126]

[0127] Among them, Λ Ope ρ represents operational efficiency. y,s This represents the probability of scenario s occurring during contract period y. Let Y represent the operating revenue, operating cost, carbon emission cost, and incentive cost under scenario s in contract period y, respectively. Y represents the set of contract periods in which the data load cluster participates in demand response, and S represents the set of different scenarios.

[0128] In this implementation, operational revenue refers to the revenue obtained from providing data services to end users. Therefore, the operational revenue under scenario s during contract period y is... Determined by the following formula:

[0129]

[0130] Where, δ Data This indicates the unit price of data services provided by green data centers.

[0131] Operating costs mainly include the maintenance costs of servers, energy storage systems, distributed renewable energy, and cooling equipment in the green data center, as well as the electricity purchase costs paid by the green data center to the power distribution network. Therefore, the operating costs under scenario s during contract period y are... Determined by the following formula:

[0132]

[0133] Among them, cserm c essm c wtm c pvm c coolm These represent the unit annual operation and maintenance costs of servers, energy storage systems, wind turbines, photovoltaic equipment, and refrigeration equipment, respectively. This represents the power consumption of the green data center from the grid during time t in scenario s within the contract period y. Δt represents the electricity price in the distribution network at time t, and Δt represents time.

[0134] Leveraging the operational flexibility of green data centers can promote the integration of distributed renewable energy and reduce carbon emissions on the generation side. If the low-carbon benefits are characterized by the carbon emission costs on the generation side resulting from the system purchasing electricity from the distribution network, then the carbon emission cost under scenario s during contract period y is... Determined by the following formula:

[0135]

[0136] Among them, c cm e represents the unit cost of carbon tax. TPG This indicates the carbon emission rate of thermal power generation.

[0137] During system operation, some end users will participate in demand response based on factors such as their information load characteristics and data service quality requirements, which will inevitably reduce user satisfaction. Therefore, economic incentives will be provided to them. The incentive cost under scenario s during contract period y is then calculated. Determined by the following formula:

[0138]

[0139] Subsequently, in step S220, an objective function for the multi-resource collaborative planning of the green data center is constructed based on investment costs and operational benefits. According to one embodiment of the present invention, the objective function can be constructed by calculating the difference between operational benefits and investment costs, and maximizing this difference as the objective function for the multi-resource collaborative planning of the green data center.

[0140] Based on equations (15) and (29), the objective function for multi-resource collaborative planning of green data centers can be determined as follows:

[0141] Maximize{profit DC}=Λ Ope -C Inv (34)

[0142] Among them, profit DC This represents the final profit, and Maximize represents finding the maximum value.

[0143] In step S230, constraints are generated, including planning stage constraints and operation stage constraints. According to one embodiment of the present invention, planning stage constraints include equipment installation constraints and incentive price constraints. Equipment installation constraints include server quantity constraints, energy storage system power and capacity constraints, distributed renewable energy installation capacity constraints, and refrigeration equipment installation constraints.

[0144] In this implementation, the server quantity constraint is determined by the following formula:

[0145] 0≤n ser ≤n ser，max (35)

[0146] Where, n ser，max This indicates the maximum installation limit for the server.

[0147] The power and capacity constraints of the energy storage system are determined by the following formula:

[0148] 0≤P Ness ≤P Ness,max (36)

[0149] 0≤E Ness ≤E Ness,max (37)

[0150] Among them, P Ness,max E Ness,max These represent the maximum installation limits for the power and capacity of the energy storage system, respectively.

[0151] Distributed renewable energy includes wind turbines and photovoltaic equipment, and its installation capacity constraints are determined by the following formula:

[0152] 0≤n wt ≤n wt,max (38)

[0153] 0≤n pv ≤n pv,max (39)

[0154] Where, n wt,max n pv,max These represent the maximum installation limits for wind turbines and photovoltaic equipment, respectively.

[0155] The installation constraints of refrigeration equipment include the installation capacity constraint and the power constraint. The installation capacity constraint is determined by the following formula:

[0156] 0≤n cool ≤n cool,max (40)

[0157] P Ncool =n cool PNcool_Sta (41)

[0158] Where, n cool n represents the installed capacity of the refrigeration equipment. cool,max P indicates the maximum installation limit for refrigeration equipment. Ncool_Sta This indicates the rated capacity of a unit of refrigeration equipment.

[0159] For cooling equipment, its rated power must meet the power constraints of the server running at full load. The power constraints of the cooling equipment are determined by the following formula:

[0160]

[0161] For the pricing and incentive design of green data center data services, the basic price and subsidy incentives need to meet the following constraints to ensure the rationality of the design, namely, the incentive price constraint is determined by the following formula:

[0162] 0≤δ IW,Inc ≤δ IW,max (43)

[0163] 0≤δ SW,Inc ≤δ SW,max (44)

[0164] Where, δ IW,Inc δ SW,Inc These represent the unit prices of IW and SW subsidies for green data centers, respectively, δ IW,max δ SW,max These represent the upper limit of the unit price for green data center incentives for IW and SW, respectively.

[0165] According to one embodiment of the present invention, the operational constraints include equipment operation constraints, grid-side constraints, and demand-side constraints. The equipment operation constraints include energy storage system operation constraints, distributed renewable energy operation constraints, refrigeration equipment operation constraints, and power balance constraints.

[0166] In this implementation, the operating constraints of the energy storage system are determined by the following formula:

[0167]

[0168]

[0169]

[0170]

[0171]

[0172]

[0173] Equations (45), (46), (47), and (48) to (50) represent the power balance constraint, energy constraint, sustainability constraint, and charge / discharge power constraint of the energy storage system, respectively. These represent the energy storage system's electricity, charging power, discharging power, charging coefficient, and discharging coefficient at time t within scenario s during contract period y, respectively. ε ess η represents the self-discharge rate of the energy storage system. ess_c η ess_d These represent the charging efficiency and discharging efficiency of the energy storage system, respectively. These represent the energy storage system's power output at times t-1, 0, and T within scenario s during the contract period y, respectively, and the State of Charge (SOC). ess,min SOC ess,max These represent the minimum and maximum values ​​of the state of charge of the energy storage system, respectively.

[0174] Distributed renewable energy operating constraints typically require consideration of the actual output power of wind turbines and photovoltaic equipment. None of them exceed the corresponding predicted output power The operating constraints of refrigeration equipment are determined by the following formula:

[0175]

[0176] The power balance constraint is determined by the following formula:

[0177]

[0178] Due to the limitation of distribution transformer capacity, the grid-side constraints are determined by the following formula:

[0179]

[0180] Among them, P tr This indicates the capacity of the distribution transformer that connects the green data center to the external power grid.

[0181] Assuming all servers in a green data center are of the same model, and that during operation, data processing tasks are evenly distributed across all active servers, then the demand-side constraints can be determined by the following formula:

[0182]

[0183]

[0184]

[0185] in, T represents the data task queuing time, data processing time, and data load arrival rate within time t under scenario s in contract period y. maxThis indicates the maximum tolerable latency for the data task.

[0186] Finally, step S240 is executed, which uses the objective function and constraints to perform multi-resource collaborative planning for the green data center.

[0187] According to one embodiment of the present invention, a multi-resource collaborative planning of a green data center is performed using the objective function shown in equation (34) and the constraints shown in equations (35) to (56). The specific planning process described above can be implemented using heuristic algorithms such as genetic algorithms and particle swarm optimization algorithms, which will not be elaborated here.

[0188] According to embodiments of the present invention, the green data center multi-resource collaborative planning scheme, based on the relationship between system planning and operation, comprehensively considers the impact of endogenous and exogenous uncertainties on the effectiveness of decision-making and planning, obtains the investment cost and operational benefits of the green data center system, constructs an objective function for green data center multi-resource collaborative planning, and generates constraints. Through the objective function and the constraints, multi-resource collaborative planning is performed on the green data center. In the above technical solution, endogenous uncertainty refers to the willingness of end users to participate in demand response planning, and exogenous uncertainty includes the source-load uncertainty of renewable energy power generation output and end-user data demand. Through multi-resource collaborative planning of green data centers, the economic efficiency and low-carbon nature of green data center planning and operation are optimized, and the potential value of end-user flexibility in promoting the efficiency improvement of green data centers is explored from a medium- to long-term planning perspective.

[0189] The various techniques described herein can be implemented in combination with hardware or software, or a combination thereof. Thus, the methods and apparatus of the present invention, or certain aspects or portions thereof, can take the form of program code (i.e., instructions) embedded in a tangible medium, such as a removable hard disk, USB flash drive, floppy disk, CD-ROM, or any other machine-readable storage medium, wherein when the program is loaded into and executed by a machine such as a computer, the machine becomes an apparatus for practicing the present invention.

[0190] When the program code is executed on a programmable computer, the computing device generally includes a processor, a processor-readable storage medium (including volatile and non-volatile memory and / or storage elements), at least one input device, and at least one output device. The memory is configured to store program code; the processor is configured to execute the green data center multi-resource collaborative planning method of the present invention according to instructions in the program code stored in the memory.

[0191] By way of example, and not limitation, readable media include readable storage media and communication media. Readable storage media stores information such as computer-readable instructions, data structures, program modules, or other data. Communication media generally embodies computer-readable instructions, data structures, program modules, or other data in the form of modulated data signals such as carrier waves or other transmission mechanisms, and includes any information delivery medium. Any combination of the above is also included within the scope of readable media.

[0192] In the specification provided herein, the algorithms and displays are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used with the examples of this invention. The required structure for constructing such systems is apparent from the above description. Furthermore, this invention is not directed to any particular programming language. It should be understood that the contents of the invention described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of the invention.

[0193] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0194] Similarly, it should be understood that, in order to streamline this disclosure and aid in understanding one or more of the various aspects of the invention, in the above description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof. However, this method of disclosure should not be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as reflected in the following claims, inventive aspects lie in fewer than all features of a single foregoing disclosed embodiment. Therefore, the claims following the detailed description are hereby expressly incorporated into this detailed description, wherein each claim itself is a separate embodiment of the invention.

[0195] Those skilled in the art will understand that modules, units, or components of the devices disclosed in the examples herein can be arranged in the devices described in this embodiment, or alternatively, can be located in one or more devices different from the devices in this example. The modules in the foregoing examples can be combined into a single module or, in addition, can be divided into multiple sub-modules.

[0196] Those skilled in the art will understand that modules in the device of the embodiments can be adaptively changed and placed in one or more devices different from that embodiment. Modules, units, or components in the embodiments can be combined into a single module, unit, or component, and further, they can be divided into multiple sub-modules, sub-units, or sub-components. Except where at least some of such features and / or processes or units are mutually exclusive, any combination can be used to combine all features disclosed in this specification (including the accompanying claims, abstract, and drawings) and all processes or units of any method or device so disclosed. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) may be replaced by an alternative feature that serves the same, equivalent, or similar purpose.

[0197] Furthermore, those skilled in the art will understand that although some embodiments described herein include certain features but not others included in other embodiments, combinations of features from different embodiments are intended to be within the scope of the invention and form different embodiments. For example, in the following claims, any of the claimed embodiments can be used in any combination.

[0198] Furthermore, some of the embodiments described herein are methods or combinations of method elements that can be implemented by a processor of a computer system or by other means of performing the functions. Therefore, a processor having the necessary instructions for implementing the methods or method elements forms means for implementing the methods or method elements. Furthermore, the elements described herein in the apparatus embodiments are examples of means for implementing the functions performed by elements for the purposes of carrying out the invention.

[0199] As used herein, unless otherwise specified, the use of ordinal numbers such as “first,” “second,” “third,” etc., to describe ordinary objects merely indicates different instances of similar objects and is not intended to imply that the objects being described must have a given order in time, space, ordering, or any other manner.

[0200] Although the invention has been described with reference to a limited number of embodiments, those skilled in the art will understand from the foregoing description that other embodiments are conceivable within the scope of the invention described herein. Furthermore, it should be noted that the language used in this specification has been chosen primarily for readability and instructional purposes, and not for the purpose of interpreting or limiting the subject matter of the invention. Therefore, many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the appended claims. The disclosure of the invention is illustrative and not restrictive, and the scope of the invention is defined by the appended claims.

Claims

1. A method for green data center multi-resource collaborative planning, executed in a computing device, the method comprising: obtaining operation benefits of a green data center system, in combination with endogenous uncertainty and exogenous uncertainty, the green data center system comprising a power distribution network line, a green data center, and an end user, the green data center including a distributed renewable energy source, an energy storage system, a server, and a refrigeration system, the endogenous uncertainty including an end user participation demand response willingness, the exogenous uncertainty including a renewable energy generation output and an end user data demand, wherein the end user participation demand response willingness is described using a regret degree matching mechanism; constructing an objective function of the green data center multi-resource collaborative planning based on an investment cost of the green data center system and the operation benefits; generating constraint conditions, the constraint conditions including planning phase constraints and operation phase constraints, wherein the planning phase constraints include device installation constraints and incentive price constraints, the device installation constraints including a server quantity constraint, an energy storage system power and capacity constraint, a distributed renewable energy installation capacity constraint, and a refrigeration device installation constraint, and the operation phase constraints including device operation constraints, grid side constraints, and demand side constraints, the device operation constraints including an energy storage system operation constraint, a distributed renewable energy operation constraint, a refrigeration device operation constraint, and a power balance constraint; performing multi-resource collaborative planning on the green data center by the objective function in combination with the constraint conditions, to achieve optimal configuration of the green data center, including information load data processing amount, power purchase amount, server start-stop state, and device output of each period.

2. The method of claim 1, wherein, The operation benefits are determined by the following formula: wherein, Λ Ope represents the operating benefit, p y,s represents the probability of the scenario s in the contract period y, respectively represent the operating revenue, operating cost, carbon emission cost, and incentive cost in the scenario s in the contract period y, Y represents a contract period set in which the data load cluster participates in demand response, and S represents a different scenario set.

3. The method of claim 2, wherein, The running revenue under the scenario s in the contract period y is determined as follows: where θ represents the number of days in a contract period y, represents the total data load of the green data center at time t in scenario s in contract period y after the demand response of load cluster i, δ Data represents the unit price of data service provided by the green data center, T represents the set of each time in a day, and I represents the set of data load clusters.

4. The method of claim 2 or 3, wherein, Operating cost in contract period y under scenario s is determined as follows: wherein c serm , c essm , c wtm , c pvm , c coolm respectively represent the unit annual operation and maintenance cost of the server, the energy storage system, the fan, the photovoltaic device, and the refrigeration device, n ser represents the number of servers installed, E Ness , P Ncool respectively represent the rated capacity of the energy storage system and the rated installed capacity of the refrigeration device, P Nwt , P Npv respectively represent the installed capacity of the fan and the photovoltaic device, θ represents the number of days in a contract period y, represents the power purchased from the power grid by the green data center at time t in scenario s in contract period y, represents the power grid price at time t, and Δt represents time. 5.The method of any one of claims 1-3, further comprising: respectively obtaining installation costs of the server, the energy storage system power, the energy storage system capacity, the fan, the photovoltaic device, and the refrigeration device, to calculate the investment cost of the green data center system.

6. The method of claim 5, wherein, The investment cost is determined by the following formula: C Inv = k ser c ser n ser + k ess c ess P Ness + k ess_e c ess_e E Ness + k wt c wt P Nwt + k pv c pv P Npv + k cool c cool P Ncool wherein C Inv represents the investment cost, k ser , k ess , k ess_e , k wt , k pv , k cool respectively represent the capital recovery coefficient of the server, the energy storage system power, the energy storage system capacity, the fan, the photovoltaic device, and the refrigeration device, c ser , c ess , c ess_e , c wt , c pv , c cool respectively represent the unit installation cost of the server, the energy storage system power, the energy storage system capacity, the fan, the photovoltaic device, and the refrigeration device, n ser represents the server installation quantity, P Ness , E Ness , P Ncool respectively represent the energy storage system rated power, the energy storage system rated capacity, and the refrigeration device rated installation capacity, P Nwt , P Npv respectively represent the installed capacity of the fan and the photovoltaic device.

7. The method of any one of claims 1-3, wherein, The step of constructing the objective function of the green data center multi-resource collaborative planning based on the investment cost of the green data center system and the operation benefits, comprises: calculating a difference between the operation benefits and the investment cost of the green data center system, and maximizing the difference as the objective function of the green data center multi-resource collaborative planning. 8.A computing device, comprising: at least one processor; and a memory storing program instructions configured to be executed by the at least one processor, the program instructions comprising instructions for performing the method of any one of claims 1-7. 9.A readable storage medium storing program instructions, when the program instructions are read and executed by a computing device, causing the computing device to perform the method of any one of claims 1-7.

Images