Multi-agent game method considering uncertainty of demand response adjustment capability of data center

Abstract

The invention relates to the technical field of data center demand response, solves the technical problem of uncertainty of the load regulation capability of a data center, and particularly relates to a multi-agent game method considering the uncertainty of the demand response regulation capability of the data center, which comprises the following steps: constructing a data center user load optimization model; solving the micro-grid operator optimization model to obtain a micro-grid operator optimization strategy, and solving the non-data center user load optimization model to obtain a non-data center user load optimization strategy; and inputting the non-data center user optimization strategy and the micro-grid optimization strategy into the data center user load model for solving. According to the method, the data center is supported to participate in demand response through demand side resources of the microgrid park and non-data center user load adjustment, and the problem of uncertainty of the load adjustment capability of the data center is solved, so that the flexibility of a power system is improved, safe and stable operation of the power system is guaranteed, and renewable energy power consumption is promoted.

Description

Technical Field

[0001] The present invention relates to the technical field of data center demand response, and in particular to a multi-agent game method that takes into account the uncertainty of data center demand response regulation capability. Background Art

[0002] In the relevant research and application of data centers participating in demand response, the existing technology mainly analyzes the uncertainty factors that affect the load regulation capacity of data centers and models them, and uses random optimization, robust optimization, fuzzy optimization and other algorithms for control to ensure that the formulated load optimization strategy can still be carried out normally under the influence of uncertainty. Or by predicting the workload and load of the data center, improving the prediction accuracy, ensuring that the formulated load optimization strategy has a small error with the actual execution, and minimizing the deviation caused by uncertainty. Or use real-time methods to optimize and adjust the load of the data center. However, none of the above methods consider solving the problem of uncertainty in the load regulation capacity of the data center through resource coordination of the microgrid or region where the data center is located, resulting in the optimization strategy being unable to be carried out normally under the influence of uncertainty, seriously affecting the safe and stable operation of the power system. Summary of the invention

[0003] In view of the shortcomings of the existing technology, the present invention provides a multi-agent game method that takes into account the uncertainty of the data center's demand response regulation capability, solving the technical problem that traditional methods have not considered solving the uncertainty of the data center's load regulation capability through resource coordination of the microgrid or region where the data center is located.

[0004] In order to solve the above technical problems, the present invention provides the following technical solutions: a multi-agent game method considering the uncertainty of the demand response regulation capability of a data center, the method comprising the following steps:

[0005] S1. Construct a data center user load optimization model, which includes a data center optimization objective function, a data center new workload constraint, a data center workload model, and a data center server quantity constraint model;

[0006] S2. Construct a microgrid operator optimization model and solve the microgrid operator optimization model to obtain the microgrid operator optimization strategy, including the discharge power size of the energy storage device. Charging power of energy storage device and the output of the gas turbine

[0007] S3. Construct a non-data center user load optimization model to solve the non-data center user load optimization strategy, including adjustable power and whether to participate in demand response

[0008] S4. Construct a two-layer iterative multi-agent game solving algorithm to obtain the optimal unit subsidy and optimal unit subsidies Corresponding microgrid operator optimization strategy and non-data center user load optimization strategy.

[0009] Furthermore, in step S1, the data center optimization objective function is:

[0010]

[0011] In the above formula, C DC_DR is a variable, which indicates the additional operating cost incurred by the new workload during the response period; T DR is a constant, indicating the demand response period; is a variable, indicating the unit subsidy given by the data center operator to the microgrid operator; is a variable, which represents the energy consumption of the data center due to processing the new workload; Δt is a constant, which represents the time interval;

[0012] The new workload constraints added to the data center include: the new interactive workload processing delay constraints added to the data center, and the new batch workload processing time constraints added to the data center;

[0013] The data center workload model is:

[0014]

[0015] The data center server quantity constraint model is:

[0016]

[0017] In the above formula, N t is a constant, indicating the number of servers in the data center; is a variable, which indicates the number of servers that the data center plans to enable for the newly added batch workload in period t; is a variable, which indicates the number of servers that the data center plans to enable for the newly added interactive workload in period t;

[0018]

[0019] In the above formula, N MAX is a constant, indicating the total number of servers in the data center; The number of servers that a data center has enabled to handle the workload it was originally scheduled to handle.

[0020] Furthermore, the energy consumption of the data center due to processing the additional power demand The following relationship is satisfied:

[0021]

[0022] Where γ is a constant, which represents the energy consumption coefficient of the data center; N t is a variable, which indicates the number of servers that the data center plans to enable for the newly added workload in period t; t is a variable, which represents the additional workload that the data center needs to handle in period t; P PEAK is a constant, representing the peak power of the server in the data center; P IDLE is a constant, representing the idle power of the server in the data center; μ is a constant, representing the unit rate at which the data center processes workloads.

[0023] Furthermore, the newly added interactive workload processing latency of the data center meets the service level agreement constraints, namely:

[0024]

[0025] In the above formula, is a constant, representing the maximum processing delay that the data center can tolerate for the interactive workloads processed during the demand response period; T TOL is a variable, which represents the processing delay reserved by the data center for interactive workloads; is a variable, which indicates the number of servers that the data center plans to enable for the newly added interactive workload in period t; is a constant, which represents the interactive workload that the data center needs to handle after adding new workloads in period t;

[0026] The processing time of the newly added batch workload in the data center must meet the following constraints, namely:

[0027]

[0028] In the above formula, is a constant, representing the i-th batch workload added to the data center during the demand response period; is a variable, which indicates the amount of batch workload that the data center needs to process in time period t; is a constant, indicating the time when the i-th batch workload initiates a computation request; is a constant, indicating the time when the calculation of the i-th batch workload is completed; is a variable, which represents the batch workload that the data center needs to process after adding new workloads in time period t; i∈I, I represents the total amount of batch workload; is a variable, which indicates the number of servers that the data center plans to enable for the newly added batch workload in period t.

[0029] Furthermore, in step S2, the microgrid operator optimization model includes: a microgrid operator optimization objective function, a microgrid power balance model, an energy storage operation model, and a distributed power supply operation model.

[0030] Furthermore, the microgrid operator optimizes the objective function as follows:

[0031]

[0032] In the above formula, C TOTAL is a variable, which represents the increased operating cost of the microgrid operator during the demand response period; C(·) represents the cost function, where represents the cost incurred by non-data center users participating in demand response; Represents the cost of adjusting the output of the energy storage device; Represents the cost of adjusting the output of distributed generation;

[0033] The microgrid power balance model is:

[0034]

[0035] In the above formula, is a variable, which represents the energy consumption of the data center due to processing the new workload;

[0036] The energy storage operation model is:

[0037]

[0038] In the above formula, are all constants, representing the charging and discharging power of the energy storage to be discharged in time period t as originally planned; is a constant, indicating the maximum charge and discharge power of the energy storage in time period t as originally planned; ΔQ t is a variable, which indicates the change in the amount of electricity after the energy storage device adjusts the charging and discharging power in time period t; η CH is a constant, representing the charging efficiency of the energy storage device; η DCH is a constant, indicating the discharge efficiency of the energy storage device; Q MAX is a constant, indicating the maximum capacity of the energy storage device; Q MIN is a constant, indicating the minimum capacity of the energy storage device; is a constant, which represents the amount of electricity in time period t after the energy storage is charged and discharged according to the original plan;

[0039] The distributed power supply operation model is:

[0040]

[0041] In the above formula, PDMIN is a constant, indicating the minimum output of the gas turbine; P DMAX is a constant representing the maximum processing capacity of the gas turbine.

[0042] Further, in step S3, the non-data center user load optimization model includes a non-data center user optimization objective function and a non-data center user incentive model.

[0043] The optimization objective function for non-data center users is:

[0044]

[0045] In the above formula, C NDC is a variable, which indicates the additional operating cost of non-mobile users; θ is a parameter, which takes a value between 0 and 1, indicating the unit demand response subsidy ratio allocated by the microgrid operator to non-data center users; is a variable, indicating the unit subsidy given by the data center operator to the microgrid operator; is a variable, which represents the load regulation amount of non-data center users estimated by the microgrid operator in time period t; π is a constant, which represents the additional cost coefficient of non-data center users temporarily participating in demand response; is the auxiliary variable introduced;

[0046] The incentive model for non-data center users is:

[0047]

[0048]

[0049] In the above formula, x m,n,t is a 0-1 variable, indicating whether user m is in the nth gear during time period t; M is a constant, indicating a maximum value; is a constant, indicating the lower limit of adjustable power when user m is in the nth gear during time period t; is a constant, indicating the upper limit of adjustable power when user m is in the nth gear during time period t; is a constant, indicating the lower limit of the incentive given to user m in the nth gear during period t; is a constant, indicating the lower limit of the incentive given to user m in the nth gear during period t; is the introduced auxiliary variable.

[0050] Furthermore, in step S4, the specific process includes the following steps:

[0051] S41, initializing a variable k=0, where k is a variable used to find the optimal unit subsidy, indicating the kth iteration calculation;

[0052] S42, Randomly Initialized Unit Subsidy Let the variable Based on the data center user load optimization model, the data center optimization objective function is solved to obtain the energy consumption

[0053] S43. Subsidize the unit Substitute into step S3, solve the non-data center user optimization objective function based on the non-data center user load optimization model, and obtain the non-data center user load optimization strategy and

[0054] S44. Optimize strategy based on non-data center user load and Calculate Substitute into the microgrid operator optimization model of step S2, let As variables, solving the optimization objective function of the microgrid operator yields and microgrid operators’ optimization strategies And let k = k + 1,

[0055] in,

[0056]

[0057] S45. Determine unit subsidies and optimal unit subsidy Is the difference less than ∈, where ∈ is a minimum value?

[0058] If so, output the optimal unit subsidy and optimal unit subsidies Corresponding optimization strategies for microgrid operators and non-data center user load optimization strategies;

[0059] If not, return to step S42 and recalculate.

[0060] By means of the above technical solution, the present invention provides a multi-agent game method considering the uncertainty of the data center demand response regulation capability, which has at least the following beneficial effects:

[0061] 1. The present invention solves the problem of uncertainty in the load regulation capability of a data center by coordinating resources in the microgrid or region where the data center is located. It can eliminate the influence of uncertainty factors on the load regulation capability of a data center, and ensure that the optimization strategies of microgrid operators and the load optimization strategies of non-data center users can still be carried out normally under the influence of uncertainty, thereby optimizing the power consumption pattern, sustainably improving the stability of the power grid, and improving the level of utilization of renewable energy by the power system.

[0062] 2. The present invention supports data centers to participate in demand response through demand-side resources of microgrid parks and load regulation of non-data center users, solves the uncertainty problem of data center load regulation capabilities, thereby improving the flexibility of the power system, ensuring the safe and stable operation of the power system, and promoting the consumption of renewable energy power.

[0063] 3. The present invention can realize demand-side resource control of data centers, microgrids, and non-data center users in a specific area. On the one hand, it can solve the problem of changes in demand response regulation capabilities of data centers due to workload uncertainty, and support data centers to participate in demand response on a regular basis through load regulation of other non-data center users. On the other hand, it can model user load regulation capabilities and demand response unit subsidies. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0065] Figure 1 It is a flow chart of the multi-agent game method of the present invention;

[0066] Figure 2 It is a schematic diagram of the relationship between the load regulation amount and the incentive size of the non-data center user of the present invention. DETAILED DESCRIPTION

[0067] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific implementation methods, so that the implementation process of how the present application uses technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly.

[0068] In the context of East-West computing, data center computing power deeply participates in the planning of power system operation through the optimized configuration of computing resources, in order to improve the efficiency of computing power operation in different regions. Demand response is a key link in power grid planning and operation. Data centers participating in demand response can not only obtain additional economic subsidies to optimize electricity consumption patterns, but also continuously improve the stability of the power grid and enhance the utilization of renewable energy in the power system.

[0069] There are several modes of research and application related to data center participation in demand response: First, model the distribution characteristics of data center workloads, build workload scheduling models and load regulation models, and propose corresponding methods, devices, and systems to support data center load optimization in participating in demand response; second, model the distribution network and microgrid resources in the area where the data center is located, use planning theory to optimize the data center load within the day, and propose corresponding methods, devices, and systems to support data center load optimization in participating in demand response; third, predict the data center load and workload before demand response, build its prediction and analysis model, form a load demand response regulation capability evaluation model, and propose corresponding methods, devices, and systems to support data center load optimization in participating in demand response.

[0070] The above three modes all take into account the uncertainty of the load regulation capability of data centers in the process of demand response. The basic method is to analyze the uncertainty factors affecting the load regulation capability of data centers and model them, and use random optimization, robust optimization, fuzzy optimization and other algorithms for control.

[0071] When data center loads participate in demand response, there is uncertainty in the regulation capacity, which causes the load regulation amount to fail to meet the threshold specified by the power grid, affecting the normal implementation of power grid demand response.

[0072] The existing technology mainly analyzes the uncertainty factors that affect the load regulation capacity of the data center and models them, and uses random optimization, robust optimization, fuzzy optimization and other algorithms for control to ensure that the formulated load optimization strategy can still be carried out normally under the influence of uncertainty. Or by predicting the workload and load of the data center, improve the prediction accuracy, ensure that the error between the formulated load optimization strategy and the actual execution is small, and minimize the deviation caused by uncertainty. Or use real-time methods to optimize and adjust the data center load. However, none of the above methods consider solving the problem of uncertainty in the load regulation capacity of the data center through resource coordination of the microgrid or region where the data center is located.

[0073] In this embodiment:

[0074] Demand response refers to the response to short-term electricity supply and demand tensions, difficulties in absorbing renewable energy power, etc., through measures mainly based on economic incentives, guiding electricity users to voluntarily adjust their electricity usage behavior according to the needs of power system operation, achieve peak shaving and valley filling, improve the flexibility of the power system, ensure the safe and stable operation of the power system, and promote the absorption of renewable energy power.

[0075] Demand-side resources refer to adjustable loads, distributed power sources, new energy storage, and other power resources that are widely distributed on the user side and can be aggregated and optimized to participate in the operation and regulation of the power system.

[0076] In order to solve the problem of uncertainty in the load regulation capacity of the data center through the resource coordination of the microgrid or region where the data center is located. Figure 1 This embodiment proposes a multi-agent game method that takes into account the uncertainty of the data center's demand response regulation capability, which can eliminate the influence of uncertainty factors on the data center's load regulation capability, and ensure that the optimization strategy of microgrid operators and the load optimization strategy of non-data center users can still be carried out normally under the influence of uncertainty, thereby achieving power consumption mode optimization, and sustainably improving the stability of the power grid and the level of utilization of renewable energy by the power system. The method includes the following steps:

[0077] S1. Construct a data center user load optimization model, which includes a data center optimization objective function, a data center new workload constraint, a data center workload model, and a data center server quantity constraint model;

[0078] The data center optimization objective function is:

[0079] The data center optimization objective function that minimizes the additional operating cost is the optimization goal of the data center operator, that is:

[0080]

[0081] In the above formula, C DC_DR is a variable, which indicates the additional operating cost incurred by the new workload during the response period; T DR is a constant, indicating the demand response period; is a variable, indicating the unit subsidy given by the data center operator to the microgrid operator; is a variable, which represents the energy consumption of the data center due to processing the new workload; Δt is a constant, which represents the time interval;

[0082] Among them, the energy consumption generated by data centers due to processing new electricity demand The following relationship is satisfied:

[0083]

[0084] Where γ is a constant, which represents the energy consumption coefficient of the data center; N t is a variable, which indicates the number of servers that the data center plans to enable for the newly added workload in period t; t is a variable, which represents the additional workload that the data center needs to handle in period t; P PEAK is a constant, representing the peak power of the server in the data center; P IDLE is a constant, representing the idle power of the server in the data center; μ is a constant, representing the unit rate at which the data center processes workloads.

[0085] The new workload constraints of the data center include: the new interactive workload processing delay constraints of the data center, and the new batch workload processing time constraints of the data center, as follows:

[0086] The latency of the newly added interactive workload processing in the data center meets the service level agreement constraints, namely:

[0087]

[0088] In the above formula, is a constant, representing the maximum processing delay that the data center can tolerate for the interactive workloads processed during the demand response period; T TOL is a variable, which represents the processing delay reserved by the data center for interactive workloads; is a variable, which indicates the number of servers that the data center plans to enable for the newly added interactive workload in period t; is a constant, which represents the interactive workload that the data center needs to process after adding new workloads in period t.

[0089] The processing time of the newly added batch workload in the data center must meet the following constraints, namely:

[0090]

[0091] In the above formula, is a constant, representing the i-th batch workload added to the data center during the demand response period; is a variable, which indicates the amount of batch workload that the data center needs to process in time period t; is a constant, indicating the time when the i-th batch workload initiates a computation request; is a constant, indicating the time when the calculation of the i-th batch workload is completed; is a variable, which represents the batch workload that the data center needs to process after adding new workloads in time period t; i∈I, I represents the total amount of batch workload; is a variable, which indicates the number of servers that the data center plans to enable for the newly added batch workload in period t;

[0092] Data center workload model:

[0093] The data center can count the batch workload and interactive workload that need to be processed during the demand response period, and the amount of additional workload that the data center needs to process λ t The constraints are:

[0094]

[0095] The data center server quantity constraint model is:

[0096]

[0097] In the above formula, N t is a constant, indicating the number of servers in the data center; is a variable, which indicates the number of servers that the data center plans to enable for the newly added batch workload in period t; is a variable, which indicates the number of servers that the data center plans to enable for the newly added interactive workload in period t.

[0098]

[0099] In the above formula, N MAX is a constant, indicating the total number of servers in the data center; The number of servers that a data center has enabled to handle the workload it was originally scheduled to handle.

[0100] The present invention can realize demand-side resource control of data centers, microgrids, and non-data center users in a specific area. On the one hand, it can solve the problem of changes in the demand response regulation capability of the data center due to workload uncertainty, and support the data center's normalized participation in demand response through load regulation of other non-data center users. On the other hand, it can model the user's load regulation capability and demand response unit subsidy.

[0101] S2. Construct a microgrid operator optimization model and solve the microgrid operator optimization model to obtain the microgrid operator optimization strategy in, is a variable, which indicates the discharge power adjusted by the energy storage device in time period t; is a variable, which represents the charging power adjusted by the energy storage device in time period t; is the output size adjusted by the gas turbine in time period t; in this embodiment, the microgrid operator optimization objective function is used as the optimization target, and the microgrid power balance model, energy storage operation model and distributed power supply operation model are used as constraints. The conventional mixed integer linear programming method or other methods can be used to solve the microgrid operator optimization model. This method is a prior art means and will not be described in detail here.

[0102] Among them, the microgrid operator optimization model includes: microgrid operator optimization objective function, microgrid power balance model, energy storage operation model and distributed power supply operation model.

[0103] In this embodiment, the microgrid operator optimizes the objective function as follows:

[0104] The optimization goal of microgrid operators is to minimize additional operating costs, namely:

[0105]

[0106] In the above formula, C TOTAL is a variable, which represents the increased operating cost of the microgrid operator during the demand response period; C(·) represents the cost function, where represents the cost incurred by non-data center users participating in demand response; Represents the cost of adjusting the output of the energy storage device; Represents the cost of adjusting the output of distributed generation.

[0107] Specifically, cost is defined as follows:

[0108]

[0109] In the above formula, θ is a parameter with a value between 0 and 1, which represents the unit demand response subsidy ratio allocated by the microgrid operator to non-data center users; is a variable, indicating the unit subsidy given by the data center operator to the microgrid operator; is a variable, which represents the load regulation amount of non-data center users estimated by the microgrid operator in time period t; DCH is a constant, representing the unit discharge cost of the energy storage device; is a variable, indicating the discharge power adjusted by the energy storage device in time period t; q CH is a constant, representing the unit discharge cost of the energy storage device; is a variable, which represents the charging power adjusted by the energy storage device in time period t; is the output of the gas turbine adjusted in time period t; is the original planned output of the gas turbine in time period t; a and b are the cost coefficients of the gas turbine; Δt is a constant, representing the time interval.

[0110] The microgrid power balance model is:

[0111] The microgrid operator needs to ensure the power balance of the microgrid where the data center is located, so as to meet the constraints of the microgrid power balance model:

[0112]

[0113] In the above formula, is a variable, which represents the energy consumption of the data center due to processing the new workload.

[0114] The energy storage operation model is:

[0115] Energy storage in the microgrid park meets the following operating constraints:

[0116]

[0117] In the above formula, are all constants, representing the charging and discharging power of the energy storage to be discharged in time period t as originally planned; is a constant, indicating the maximum charge and discharge power of the energy storage in time period t as originally planned; ΔQ t is a variable, which indicates the change in the amount of electricity after the energy storage device adjusts the charging and discharging power in time period t; η CH is a constant, representing the charging efficiency of the energy storage device; η DCH is a constant, indicating the discharge efficiency of the energy storage device; Q MAX is a constant, indicating the maximum capacity of the energy storage device; Q MIN is a constant, indicating the minimum capacity of the energy storage device; It is a constant, which represents the amount of electricity in time period t after the energy storage is charged and discharged according to the original plan.

[0118] The distributed power supply operation model is:

[0119] Distributed power sources are mainly small gas turbines. The output constraints of gas turbines after adjusting the power generation are as follows:

[0120]

[0121] In the above formula, P DMIN is a constant, indicating the minimum output of the gas turbine; P DMAX is a constant representing the maximum processing capacity of the gas turbine.

[0122] This embodiment supports the data center's participation in demand response through demand-side resources of the microgrid park and load regulation of non-data center users, solves the uncertainty problem of the data center's load regulation capability, thereby improving the flexibility of the power system, ensuring the safe and stable operation of the power system, and promoting the consumption of renewable energy power.

[0123] S3. Construct a non-data center user load optimization model to solve the non-data center user load optimization strategy and in, is a variable, which represents the power that can be adjusted by user m during period t; It is a 0-1 variable, indicating whether user m participates in demand response in time period t. In this embodiment, the non-data center user optimization objective function is used as the optimization target, and the non-data center user incentive model is used as the constraint condition. The conventional mixed integer linear programming method or other methods can be used to solve the non-data center user load optimization model. This method is a prior art means and will not be described in detail here.

[0124] The non-data center user load optimization model includes a non-data center user optimization objective function and a non-data center user incentive model.

[0125] The optimization objective function for non-data center users is:

[0126] The optimization goal of non-data center users is to determine their own bids based on the data center bids fed back by the microgrid operator, and to determine whether to participate in demand response to reduce loads to help the data center complete the load reduction task. That is:

[0127]

[0128] In the above formula, C NDC is a variable, which indicates the additional operating cost of non-mobile users; θ is a parameter, which takes a value between 0 and 1, indicating the unit demand response subsidy ratio allocated by the microgrid operator to non-data center users; is a variable, indicating the unit subsidy given by the data center operator to the microgrid operator; is a variable, which represents the load regulation amount of non-data center users estimated by the microgrid operator in time period t; is a variable, which represents the power that can be adjusted by user m during period t; is a 0-1 variable, indicating whether user m participates in demand response in time period t; π is a constant, indicating the additional cost coefficient of non-data center users temporarily participating in demand response.

[0129] Since the objective function of non-data center users involves the multiplication of two variables, that is, It is easy to increase the difficulty of solving. Therefore, in order to reduce the difficulty of solving the load optimization strategy for non-data center users, it is necessary to transform the objective function of non-data center users. After the transformation, it becomes:

[0130]

[0131] In the above formula, M is a constant, representing a maximum value.

[0132] The optimization goal of non-data center users is to maximize C NDC , the final non-data center user optimization objective function is expressed as follows:

[0133]

[0134] In the above formula, C NDC is a variable, which indicates the additional operating cost of non-mobile users. θ is a parameter, which takes a value between 0 and 1, indicating the unit demand response subsidy ratio allocated by the microgrid operator to non-data center users; is a variable, indicating the unit subsidy given by the data center operator to the microgrid operator; is a variable, which represents the load regulation amount of non-data center users estimated by the microgrid operator in time period t; π is a constant, which represents the additional cost coefficient of non-data center users temporarily participating in demand response; is the introduced auxiliary variable.

[0135] The incentive model for non-data center users is:

[0136] The load adjustment amount of non-data center users is related to the incentive size. For details, refer to Figure 2 In the figure, the excitation is divided into multiple gears. When the excitation reaches a certain range, the load regulation capacity also reaches the corresponding range.

[0137] The relationship between data center user load regulation and incentives is as follows:

[0138]

[0139] In the above formula, x m,n,t is a 0-1 variable, indicating whether user m is in the nth gear during time period t; M is a constant, indicating a maximum value; is a constant, indicating the lower limit of adjustable power when user m is in the nth gear during time period t; is a constant, indicating the upper limit of the adjustable power when user m is in the nth gear during time period t; q m,t is a constant, indicating the incentive given to user m during period t; is a constant, indicating the lower limit of the incentive given to user m in the nth gear during period t; is a constant, indicating the lower limit of the incentive given to user m for the nth gear in period t.

[0140] S4. Construct a two-layer iterative multi-agent game solving algorithm to obtain the optimal unit subsidy and optimal unit subsidies The corresponding microgrid operator optimization strategy and non-data center user load optimization strategy. The specific process includes the following steps:

[0141] S41, initializing a variable k=0, where k is a variable used to find the optimal unit subsidy, indicating the kth iteration calculation;

[0142] S42, Randomly Initialized Unit Subsidy Let the variable Based on the data center user load optimization model, the data center optimization objective function is solved to obtain the energy consumption

[0143] S43. Subsidize the unit Substitute into step S3, solve the non-data center user optimization objective function based on the non-data center user load optimization model, and obtain the non-data center user load optimization strategy and

[0144] S44. Optimize strategy based on non-data center user load and Calculate Substitute into the microgrid operator optimization model of step S2, let As variables, solving the optimization objective function of the microgrid operator yields and microgrid operators’ optimization strategies And let k = k + 1,

[0145] in,

[0146]

[0147] S45. Determine unit subsidies and optimal unit subsidy Is the difference less than ∈, where ∈ is a minimum value; if so, output the optimal unit subsidy and optimal unit subsidies Corresponding microgrid operator optimization strategy and non-data center user load optimization strategy; if not, return to step S42 and recalculate.

[0148] This embodiment can achieve demand-side resource control for data centers, microgrids, and non-data center users in a specific area. On the one hand, it can solve the problem of changes in the demand response adjustment capability of the data center due to workload uncertainty, and support the data center's normalized participation in demand response through load adjustment of other non-data center users. On the other hand, it can model the user's load adjustment capability and demand response unit subsidy.

[0149] Those skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a program, so the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0150] Each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the above embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0151] The above implementation methods have been described in detail. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A multi-agent game method considering the uncertainty of data center demand response regulation capability, characterized in that: The method comprises the following steps: S1. Construct a data center user load optimization model, which includes a data center optimization objective function, a data center new workload constraint, a data center workload model, and a data center server quantity constraint model; S2. Construct a microgrid operator optimization model and solve the microgrid operator optimization model to obtain the microgrid operator optimization strategy, including the discharge power size of the energy storage device. Charging power of energy storage device and the output of the gas turbine S3. Construct a non-data center user load optimization model to solve the non-data center user load optimization strategy, including adjustable power and whether to participate in demand response S4. Construct a two-layer iterative multi-agent game solving algorithm to obtain the optimal unit subsidy and optimal unit subsidies Corresponding microgrid operator optimization strategy and non-data center user load optimization strategy.

2. The multi-agent game method according to claim 1, characterized in that: In step S1, the data center optimization objective function is: In the above formula, C DC _ DR is a variable, which indicates the additional operating cost incurred by the new workload during the response period; T DR is a constant, indicating the demand response period; is a variable, indicating the unit subsidy given by the data center operator to the microgrid operator; is a variable, which represents the energy consumption of the data center due to processing the new workload; Δt is a constant, which represents the time interval; The new workload constraints added to the data center include: the new interactive workload processing delay constraints added to the data center, and the new batch workload processing time constraints added to the data center; The data center workload model is: The data center server quantity constraint model is: In the above formula, N t is a constant, indicating the number of servers in the data center; is a variable, which indicates the number of servers that the data center plans to enable for the newly added batch workload in period t; is a variable, which indicates the number of servers that the data center plans to enable for the newly added interactive workload in period t; In the above formula, N MAX is a constant, indicating the total number of servers in the data center; The number of servers that a data center has enabled to handle the workload it was originally scheduled to handle.

3. The multi-agent game method according to claim 2, characterized in that: The energy consumption of the data center due to processing the additional electricity demand The following relationship is satisfied: Where γ is a constant, which represents the energy consumption coefficient of the data center; N t is a variable, which indicates the number of servers that the data center plans to enable for the newly added workload in period t; t is a variable, which represents the additional workload that the data center needs to handle in period t; P PEAK is a constant, representing the peak power of the server in the data center; P IDLE is a constant, representing the idle power of the server in the data center; μ is a constant, representing the unit rate at which the data center processes workloads.

4. The multi-agent game method according to claim 2, characterized in that: The newly added interactive workload processing latency of the data center meets the service level agreement constraints, namely: In the above formula, is a constant, representing the maximum processing delay that the data center can tolerate for the interactive workloads processed during the demand response period; T TOL is a variable, which represents the processing delay reserved by the data center for interactive workloads; is a variable, which indicates the number of servers that the data center plans to enable for the newly added interactive workload in period t; is a constant, which represents the interactive workload that the data center needs to handle after adding new workloads in period t; The processing time of the newly added batch workload in the data center must meet the following constraints, namely: In the above formula, is a constant, representing the i-th batch workload added to the data center during the demand response period; is a variable, which indicates the amount of batch workload that the data center needs to process in time period t; is a constant, indicating the time when the i-th batch workload initiates a computation request; is a constant, indicating the time when the calculation of the i-th batch workload is completed; is a variable, which represents the batch workload that the data center needs to process after adding new workloads in time period t; i∈I, I represents the total amount of batch workload; is a variable, which indicates the number of servers that the data center plans to enable for the newly added batch workload in period t.

5. The multi-agent game method according to claim 1, characterized in that: In step S2, the microgrid operator optimization model includes: a microgrid operator optimization objective function, a microgrid power balance model, an energy storage operation model, and a distributed power supply operation model.

6. The multi-agent game method according to claim 5, characterized in that: The microgrid operator optimization objective function is: In the above formula, C TOTAL is a variable, which represents the increased operating cost of the microgrid operator during the demand response period; C(·) represents the cost function, where represents the cost incurred by non-data center users participating in demand response; Represents the cost of adjusting the output of the energy storage device; Represents the cost of adjusting the output of distributed generation; The microgrid power balance model is: In the above formula, is a variable, which represents the energy consumption of the data center due to processing the new workload; The energy storage operation model is: In the above formula, are all constants, representing the charging and discharging power of the energy storage to be discharged in time period t as originally planned; is a constant, indicating the maximum charge and discharge power of the energy storage in time period t as originally planned; ΔQ t is a variable, which indicates the change in the amount of electricity after the energy storage device adjusts the charging and discharging power in time period t; η CH is a constant, representing the charging efficiency of the energy storage device; η DCH is a constant, indicating the discharge efficiency of the energy storage device; Q MAX is a constant, indicating the maximum capacity of the energy storage device; Q MIN is a constant, indicating the minimum capacity of the energy storage device; is a constant, which represents the amount of electricity in time period t after the energy storage is charged and discharged according to the original plan; The distributed power supply operation model is: In the above formula, P DMIN is a constant, indicating the minimum output of the gas turbine; P DMAX is a constant representing the maximum processing capacity of the gas turbine.

7. The multi-agent game method according to claim 1, characterized in that: In step S3, the non-data center user load optimization model includes a non-data center user optimization objective function and a non-data center user incentive model. The optimization objective function for non-data center users is: In the above formula, C NDC is a variable, which indicates the additional operating cost of non-mobile users; θ is a parameter, which takes a value between 0 and 1, indicating the unit demand response subsidy ratio allocated by the microgrid operator to non-data center users; is a variable, indicating the unit subsidy given by the data center operator to the microgrid operator; is a variable, which represents the load regulation amount of non-data center users estimated by the microgrid operator in time period t; π is a constant, which represents the additional cost coefficient of non-data center users temporarily participating in demand response; is the auxiliary variable introduced; The incentive model for non-data center users is: In the above formula, x m,n,t is a 0-1 variable, indicating whether user m is in the nth gear during time period t; M is a constant, indicating a maximum value; is a constant, indicating the lower limit of adjustable power when user m is in the nth gear during time period t; is a constant, indicating the upper limit of adjustable power when user m is in the nth gear during time period t; is a constant, indicating the lower limit of the incentive given to user m in the nth gear during period t; is a constant, indicating the lower limit of the incentive given to user m for the nth gear in period t.

8. The multi-agent game method according to claim 1, characterized in that: In step S4, the specific process includes the following steps: S41, initializing a variable k=0, where k is a variable used to find the optimal unit subsidy, indicating the kth iteration calculation; S42, Randomly Initialized Unit Subsidy Let the variable Based on the data center user load optimization model, the data center optimization objective function is solved to obtain the energy consumption S43. Subsidize the unit Substitute into step S3, solve the non-data center user optimization objective function based on the non-data center user load optimization model, and obtain the non-data center user load optimization strategy and S44. Optimize strategy based on non-data center user load and Calculate Substitute into the microgrid operator optimization model of step S2, let As variables, solving the optimization objective function of the microgrid operator yields and microgrid operators’ optimization strategies And let k = k + 1, in, S45. Determine unit subsidies and optimal unit subsidy Is the difference less than ∈, where ∈ is a minimum value? If so, output the optimal unit subsidy and optimal unit subsidies Corresponding optimization strategies for microgrid operators and non-data center user load optimization strategies; If not, return to step S42 and recalculate.

9. A computer-readable storage medium, characterized in that: It stores computer executable instructions, which, when executed by a processor, can implement the multi-agent game method described in any one of claims 1 to 8.

10. A computer device, characterized in that: It comprises a processor and a memory, wherein the memory is used to store a computer program, and when the computer program is executed by the processor, the multi-agent game method according to any one of claims 1 to 8 is implemented.

Images