Power configuration method and device based on power and computing power cooperation and electronic equipment

Abstract

The application discloses a power configuration method and device based on power and computing power cooperation and electronic equipment. The method comprises the following steps: obtaining target computing power tasks and power operation data in a power configuration period; determining a plurality of computing power devices based on the target computing power tasks; determining a power supply topology based on the power operation data and the plurality of computing power devices; determining a power deviation cost parameter corresponding to a first computing power task and a power variation cost parameter corresponding to a second computing power task; and determining a target power configuration based on a power demand parameter corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameter and the power variation cost parameter. The application solves the technical problem of high power configuration cost in the related art when the power configuration is performed on the computing power system based on the cooperation between the power and the computing power.

Description

Technical Field

[0001] This invention relates to the field of computer collaboration, and more specifically, to a power configuration method, apparatus, and electronic device for the collaboration of power and computing power. Background Technology

[0002] In related technologies, to achieve coordinated scheduling and stable operation of the power system and the computing system, meet the execution requirements of computing tasks, and adapt to the power supply capacity and operating status of the power side, power configuration of the computing system is required. However, in related technologies, power configuration of the computing system based on the coordination of power and computing power presents a technical problem of high power configuration cost.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This invention provides a power configuration method, apparatus, and electronic device that coordinates power and computing power, in order to at least solve the technical problem of high power configuration cost when configuring the power of a computing system based on the coordination of power and computing power in related technologies.

[0005] According to one aspect of the present invention, a power configuration method for power and computing power coordination is provided, comprising: acquiring a target computing power task corresponding to a computing power system and power operation data corresponding to a power system during a power configuration period; determining a plurality of computing power devices corresponding to the computing power system based on the target computing power task; determining a power supply topology corresponding to the computing power system based on the power operation data and the plurality of computing power devices, wherein the power supply topology is used to characterize the power supply association relationship between the plurality of computing power devices and a plurality of power devices in the power system; determining a power deviation cost parameter corresponding to a first computing power task and a power variation cost parameter corresponding to a second computing power task, wherein the first computing power task is a computing power task executed during a fixed period, the second computing power task is a computing power task executed during a non-fixed period, and the target computing power task includes the first computing power task and the second computing power task; determining a target power configuration corresponding to the computing power system based on the power demand parameter corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameter, and the power variation cost parameter, for configuring the operating power of the plurality of computing power devices.

[0006] Optionally, determining the target power configuration corresponding to the computing power system based on the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters includes: determining power interaction cost parameters corresponding to the plurality of computing power devices respectively based on the power operation data and the power supply topology, wherein the corresponding power interaction cost parameters are cost parameters generated when the corresponding computing power device interacts with the corresponding power device; and determining the target power configuration corresponding to the computing power system based on the power interaction cost parameters, the power demand parameters corresponding to the target computing power task, the power deviation cost parameters, and the power variation cost parameters.

[0007] Optionally, determining the target power configuration corresponding to the computing power system based on the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameter, and the power variation cost parameter includes: determining the device operation data corresponding to each of the plurality of computing power devices; determining a first power association constraint corresponding to each of the plurality of target devices and a second power association constraint corresponding to the computing power system based on the device operation data corresponding to each of the plurality of computing power devices, the power operation data, and the power supply topology, wherein the first power association constraint is a constraint corresponding to the power association between the input power and output power of the corresponding target device, and the second power association constraint is a constraint corresponding to the power association between the power supply power of the power system and the power consumption power of the computing power system; and determining the target power configuration corresponding to the computing power system based on the first power association constraint, the second power association constraint, the power demand parameters corresponding to the target computing power task, the power deviation cost parameter, and the power variation cost parameter corresponding to each of the plurality of target devices.

[0008] Optionally, determining the target power configuration corresponding to the computing power system based on the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters includes: determining the computing power energy efficiency parameters corresponding to the target computing power task, wherein the computing power energy efficiency parameters characterize the task progress completed by the corresponding computing power device per unit of power consumed; determining the task execution period corresponding to the target computing power task; determining the progress execution constraints corresponding to the target computing power task based on the task execution period, the computing power energy efficiency parameters, and the current task progress corresponding to the target computing power task; and determining the target power configuration corresponding to the computing power system based on the progress execution constraints, the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters.

[0009] Optionally, determining the target power configuration corresponding to the computing system based on the power demand parameters corresponding to the target computing task, the power operation data, the power supply topology, the power deviation cost parameter, and the power variation cost parameter includes: determining the cooling energy efficiency parameter and environmental parameters corresponding to the computing system, wherein the cooling energy efficiency parameter represents the degree of cooling of the computing system per unit of power consumed; determining the cooling power constraint corresponding to the computing system based on the cooling energy efficiency parameter and the environmental parameters; and determining the target power configuration corresponding to the computing system based on the cooling power constraint, the power demand parameters corresponding to the target computing task, the power operation data, the power supply topology, the power deviation cost parameter, and the power variation cost parameter.

[0010] Optionally, determining the target power configuration corresponding to the computing power system based on the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters includes: when the plurality of power devices include a first device and a second device, determining the carbon emission constraints corresponding to the computing power system based on the carbon emission index corresponding to the first device and the target carbon emission of the power system, wherein the first device is a power device with a carbon emission index greater than or equal to a carbon emission threshold, and the second device is a power device with a carbon emission index less than a carbon emission threshold; determining the power supply stability characteristics corresponding to the second device; determining the power supply stability constraints corresponding to the computing power system based on the power supply stability characteristics; and determining the target power configuration corresponding to the computing power system based on the carbon emission constraints, the power supply stability constraints, the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters.

[0011] Optionally, determining the target power configuration corresponding to the computing power system based on the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters includes: determining a first security constraint corresponding to the computing power system and a second security constraint corresponding to the power system; and determining the target power configuration corresponding to the computing power system based on the first security constraint, the second security constraint, the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters.

[0012] According to one aspect of the present invention, a power configuration device for power and computing power coordination is provided, comprising: an acquisition module, configured to acquire a target computing power task corresponding to a computing power system and power operation data corresponding to a power system during a power configuration period; a first determination module, configured to determine a plurality of computing power devices corresponding to the computing power system based on the target computing power task; a second determination module, configured to determine a power supply topology corresponding to the computing power system based on the power operation data and the plurality of computing power devices, wherein the power supply topology is used to characterize the power supply association relationship between the plurality of computing power devices and a plurality of power devices in the power system; and a third determination module. A first computing power task is used to determine the power deviation cost parameter corresponding to the first computing power task and the power variation cost parameter corresponding to the second computing power task, wherein the first computing power task is a computing power task executed during a fixed period, and the second computing power task is a computing power task executed during a non-fixed period, and the target computing power task includes the first computing power task and the second computing power task; a fourth determining module is used to determine the target power configuration corresponding to the computing power system based on the power demand parameter corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameter, and the power variation cost parameter, so as to configure the operating power of the multiple computing power devices.

[0013] According to one aspect of the present invention, an electronic device is provided, comprising: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to execute the instructions to implement the power configuration method for power and computing power coordination as described in any of the preceding claims.

[0014] According to one aspect of the present invention, a computer-readable storage medium is provided, wherein when the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the power configuration method of power and computing power coordination as described above.

[0015] In this embodiment of the invention, by acquiring target computing power tasks and power operation data, a foundation can be provided for subsequent matching of computing power devices and power supply correlation analysis; determining the corresponding computing power devices based on the target computing power tasks can accurately pinpoint the core objects of power regulation; determining the power supply topology based on power operation data and computing power devices can clearly depict the power supply correlation between computing power devices and power equipment, providing a physical constraint framework for power configuration; determining the power deviation cost parameters of the first computing power task and the power variation cost parameters of the second computing power task can quantify the power adjustment costs of different types of computing power tasks; finally, determining the target power configuration based on multi-dimensional parameters can achieve refined configuration of computing power device operating power under the premise of meeting the execution requirements of computing power tasks and power system operation constraints, avoiding unnecessary power loss, reducing the overall power configuration cost, and thus solving the technical problem of high power configuration cost when configuring the power of computing power systems based on the collaboration of power and computing power in related technologies. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings:

[0017] Figure 1 This is a flowchart of a power configuration method for coordinating power and computing power according to an embodiment of the present invention;

[0018] Figure 2 This is a schematic diagram of the power configuration process in an optional embodiment of the present invention;

[0019] Figure 3 This is a schematic diagram of the power configuration process of fusion constraints in an optional embodiment of the present invention;

[0020] Figure 4 This is a schematic diagram of the regional association relationship in an optional embodiment of the present invention;

[0021] Figure 5 This is a structural block diagram of a power configuration device that coordinates power and computing power according to an embodiment of the present invention. Detailed Implementation

[0022] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0023] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0024] Example 1

[0025] According to an embodiment of the present invention, an embodiment of a power configuration method that coordinates power and computing power is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0026] Figure 1 This is a flowchart of a power configuration method for power and computing power coordination according to an embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0027] S102, Obtain the target computing power task corresponding to the computing power system and the power operation data corresponding to the power system during the power configuration period;

[0028] This involves power configuration time periods, which are specific time intervals for power configuration. These time periods can be divided into different time scales, such as hours or minutes, depending on actual application needs. For example, the power configuration time period could be from 8:00 AM to 8:00 PM daily as the daytime power configuration period, and from 8:00 PM to 8:00 AM the next day as the nighttime power configuration period. Alternatively, it could be set to a power configuration time period every 30 minutes according to scheduling precision.

[0029] This involves a computing power system, which is a hardware and software collaborative system that receives, processes, and executes various computing power tasks. It can undertake various computing tasks such as artificial intelligence training, big data analysis, and cloud services. It includes hardware such as server clusters, computing power terminals, network switching equipment, and cooling equipment, as well as computing power task scheduling software, forming an overall system, such as a computing power center (i.e., an intelligent computing center).

[0030] This involves target computing power tasks, which are specific computing power tasks that need to be completed during the power configuration period, including rigid computing power tasks (i.e., the first computing power task) and adjustable computing power tasks (i.e., the second computing power task).

[0031] This involves a power system, which is a system used to provide electricity to the power system. Specifically, it can be composed of power generation equipment (photovoltaic, wind power, thermal power), power transformation / transmission equipment, energy storage equipment, grid interface equipment, etc., and is an overall energy system that can provide continuous power supply to the computing power system and realize power transmission and dispatch.

[0032] This includes power operation data, which consists of the real-time operating status, power parameters, and related monitoring data of each power device and the power system as a whole during the power configuration period.

[0033] By acquiring the target computing power tasks corresponding to the computing power system during the power configuration period, we can clarify the specific computing needs of the computing power system during that period. Furthermore, by acquiring the power operation data corresponding to the power system during the same period, we can comprehensively grasp the real-time power supply capacity and operating status of the power system, providing data support for analyzing the power system's power supply support capability for the computing power system.

[0034] S104, Based on the target computing power task, determine multiple computing power devices corresponding to the computing power system;

[0035] This involves multiple computing devices, which are the devices that actually perform computing tasks in the computing system, including but not limited to servers, storage devices, network devices, accelerators, etc.

[0036] Based on the target computing task, multiple computing devices corresponding to the computing system can be determined, which can accurately locate the physical resources required to execute the computing task, ensure that the computing demand is allocated to the appropriate devices for processing, avoid resource idleness or overload, and achieve efficient operation of the computing system.

[0037] S106. Based on power operation data and multiple computing devices, determine the power supply topology corresponding to the computing system. The power supply topology is used to characterize the power supply relationship between the multiple computing devices and multiple power devices in the power system.

[0038] This involves the power supply topology, which is a structure used to characterize the power supply connection relationship between each computing device in the computing power system and each power device in the power system. Specifically, it is a topology diagram that reflects the core topology of the relationship between power supply and power consumption of computing devices in a computer-aided collaborative scenario. It clarifies how each computing device obtains power from the power system and how each device in the power system works collaboratively to meet the power consumption needs of the computing devices.

[0039] This involves multiple power devices, which are various types of equipment in the power system used to provide power supply, realize power transmission and dispatch, including but not limited to power generation equipment, power transformation equipment, power transmission equipment, energy storage equipment, grid interface equipment, etc.

[0040] By combining power operation data with multiple identified computing devices, the power supply topology of the computing system can be clearly defined. This allows for a clear understanding of the power supply relationships between each computing device and each power device in the power system, and accurately presents the power acquisition path of the computing devices and the power supply coordination logic of the power devices.

[0041] S108, determine the power deviation cost parameter corresponding to the first computing power task and the power variation cost parameter corresponding to the second computing power task, wherein the first computing power task is a computing power task executed in a fixed period of time, the second computing power task is a computing power task executed in a non-fixed period of time, and the target computing power task includes the first computing power task and the second computing power task.

[0042] Among them, the first computing power task is a rigid computing power task that needs to be executed within a fixed time period and whose execution time cannot be adjusted. It is a computing power task that the computing power system must prioritize to execute.

[0043] This includes the power deviation cost parameter, which quantifies the various losses and costs incurred when the actual operating power of the first computing power task deviates from the planned power, reflecting the correlation between the degree of power deviation and the corresponding cost.

[0044] This involves a second computing power task, which is an adjustable computing power task in the target computing power task with no fixed execution time requirement and adjustable execution time (such as a delayed adjustable computing power task, which has the flexibility of power configuration).

[0045] This includes the power variation cost parameter, which quantifies the various losses and costs incurred by the second computing power task due to adjustments in execution time and changes in operating power, reflecting the correlation between the power variation magnitude and the corresponding cost.

[0046] By distinguishing between the first computing task executed at fixed time periods and the second computing task executed at non-fixed time periods, and by determining the power deviation cost parameters for the first computing task and the power variation cost parameters for the second computing task, we can quantify the losses and costs incurred by different types of computing tasks during power adjustment. This accurately reflects the correlation between power changes and costs for the two types of tasks, ensuring that when setting target power configurations, we can achieve differentiated power allocation based on the cost characteristics of different tasks.

[0047] S110 determines the target power configuration corresponding to the computing power system based on the power demand parameters, power operation data, power supply topology, power deviation cost parameters, and power variation cost parameters corresponding to the target computing power task, so as to configure the operating power of multiple computing power devices.

[0048] This includes power requirement parameters, which are power-related indicators required by the computing system and various computing devices to complete the corresponding computing tasks during the execution of the target computing task, reflecting the various power requirements of the computing task.

[0049] This involves the target power configuration, which is used for the power allocation and operation of the computing system and its various computing devices.

[0050] Based on the power supply topology, the actual power supply relationship between computing power and power equipment is clearly defined, so that the power configuration fits the actual power supply architecture to avoid ineffective losses. The power adjustment cost of two types of computing power tasks is accurately quantified by power deviation and power change cost parameters. Then, the target power configuration is formulated by combining computing power demand parameters and power operation data, so that the operating power of computing equipment is adapted to the task and power system status, while avoiding unnecessary power loss and reducing power configuration costs.

[0051] Through the aforementioned steps S102-S110, by acquiring target computing power tasks and power operation data, a foundation can be provided for subsequent matching of computing power equipment and power supply correlation analysis. Determining the corresponding computing power equipment based on the target computing power tasks allows for precise identification of the core objects of power regulation. Determining the power supply topology based on power operation data and computing power equipment clearly depicts the power supply correlation between computing power equipment and power equipment, providing a physical constraint framework for power configuration. Determining the power deviation cost parameters of the first computing power task and the power variation cost parameters of the second computing power task quantifies the power adjustment costs of different types of computing power tasks. Finally, determining the target power configuration based on multi-dimensional parameters enables refined configuration of computing power equipment operating power while meeting the requirements of computing power task execution and power system operation constraints. This avoids unnecessary power loss, reduces overall power configuration costs, and thus solves the technical problem of high power configuration costs when configuring power for computing power systems based on power and computing power collaboration in related technologies.

[0052] As an optional embodiment, the target power configuration corresponding to the computing power system is determined based on the power demand parameters, power operation data, power supply topology, power deviation cost parameters, and power variation cost parameters corresponding to the target computing power task. This includes: determining the power interaction cost parameters corresponding to multiple computing power devices based on the power operation data and power supply topology, wherein the corresponding power interaction cost parameters are the cost parameters generated when the corresponding computing power device interacts with the corresponding power equipment; and determining the target power configuration corresponding to the computing power system based on the power interaction cost parameters, the power demand parameters, power deviation cost parameters, and power variation cost parameters corresponding to the target computing power task.

[0053] This includes power interaction cost parameters, which are various cost-related parameters generated during the power transmission and interaction between the corresponding computing power device and the corresponding power supply equipment in the power system. These parameters quantify the power interaction cost between the corresponding computing power device and the corresponding power equipment.

[0054] Specifically, based on power operation data and power supply topology, power interaction cost parameters corresponding to multiple computing devices are determined. This further includes: based on the regional correlation between the first region where the computing system is located and the second region where the power system is located, power operation data, and power supply topology, power interaction cost parameters corresponding to multiple computing devices are determined. The regional correlation includes: computing power demand and supply correlation, computing power task scheduling correlation, power consumption demand correlation, energy cost complementarity correlation, cross-regional power grid transmission correlation, and carbon emission cost control correlation.

[0055] Regarding the supply and demand relationship of computing power, the first region, as the place where computing power demand is generated, cannot meet the local explosive computing power development demand due to resource constraints such as land, energy and environment. Therefore, it is necessary to transfer massive computing power tasks across regions to the second region, and the power system of the second region supports the operation of the computing power system to achieve cross-regional supply matching of computing power demand.

[0056] Regarding the scheduling of computing power tasks, the first region has the right to initiate, allocate, and schedule computing power tasks. It defines the execution priority and time period requirements (rigid / adjustable) of computing power tasks according to its own business needs, and distributes the target computing power tasks to the computing power system of the second region. The power system of the second region then carries out targeted power configuration based on the scheduling requirements.

[0057] Regarding the correlation of power consumption demand, namely the demand for clean and low-cost power consumption in the first region, the computing power of the first region is transformed into a computer-based collaborative model for power calculation in the second region, which transforms the demand for new energy power consumption in the second region into a targeted demand for new energy power consumption in the second region. Relying on the abundant new energy resources such as photovoltaic and wind power in the second region, the green power demand behind the computing power task in the first region is met.

[0058] Regarding the complementary relationship of energy costs, the first region faces the current situation of high electricity prices and high energy usage costs. Its computing power industry's need to reduce costs complements the advantages of the second region's low electricity prices and low new energy power generation costs. Through cross-regional computing collaboration, the computing power tasks of the first region can achieve low-cost operation by leveraging the power system of the second region.

[0059] For cross-regional power grid transmission linkage, the first region and the second region can achieve power interconnection through ultra-high voltage AC / DC power grids. The first region can coordinate with the power system of the second region to carry out cross-regional power dispatch according to the real-time power demand of the computing power system. If the output of new energy in the second region is insufficient, the first region can supplement the power through the cross-regional power grid to ensure the stable operation of the computing power system.

[0060] Regarding the correlation of carbon emission cost control, the carbon emission reduction control requirements of the first region are higher than those of the second region. As a high-energy-consuming industry, the computing power industry in the first region needs to reduce its own carbon emission intensity through cross-regional computing collaboration and take advantage of the low-carbon emission advantages of new energy power generation in the second region, so as to achieve coordinated control of carbon emission costs between the first and second regions.

[0061] By combining power operation data and power supply topology to determine the power interaction cost parameters of each computing device, the power transmission cost between a single computing device and its corresponding power equipment can be accurately quantified. This parameter is then combined with the power requirement parameters of computing tasks and the power cost parameters of the two types of computing tasks to formulate a target power configuration. This allows the power allocation scheme to take into account both the actual power interaction loss of computing devices and the power adjustment cost of different computing tasks, ensuring that the power configuration of the computing system conforms to the actual transmission characteristics of cross-regional power supply. This achieves refined and accurate power configuration, thereby further avoiding unnecessary power loss and making the power configuration of the computing system more in line with the actual operating scenario.

[0062] For example, based on power interaction cost parameters, power demand parameters corresponding to the target computing power task, power deviation cost parameters, and power variation cost parameters, the target power configuration corresponding to the computing power system can be determined. This allows for the establishment of a Model Predictive Control (MPC) optimization problem, which can be applied in the prediction time domain. Minimize the total cost (i.e., power configuration cost) within the power configuration period, specifically achieved using the following objective function:

[0063]

[0064] in, The parameter representing the power interaction cost at time k includes the interaction power at time k. Interaction power cost weighting coefficient at time k ; This represents the penalty weighting coefficient for energy storage devices deviating from the target value of their state of charge (SOC). This represents the state of charge of the energy storage device at time k; This represents the power deviation cost parameter at time k, including the penalty weighting coefficient for the power deviation of the rigid computing power task (i.e., the first computing power task) at time k. The actual operating power of the i-th rigid computing task at time k and the reference operating power of the i-th rigid computing task , where R is the total number of rigid computing power tasks; This represents the power variation cost parameter, including the penalty weighting coefficient for power deviation in adjustable computing power tasks (second computing power tasks). The actual operating power of the j-th adjustable computing task at time k The reference operating power of the j-th adjustable computing power task at time k. Penalty weighting coefficient for adjustable computing power task execution latency The total computational load / total workload of the j-th adjustable computing power task , In order to delay duration The following is the amount of computation completed for the j-th adjustable computing power task, where M is the total number of adjustable computing power tasks; The penalty weighting coefficient is used to constrain violations; Let be the norm of all constraint violations at time k, representing the degree to which the constraints are broken.

[0065] Among them, power equipment includes energy storage equipment.

[0066] As an optional embodiment, based on the power demand parameters, power operation data, power supply topology, power deviation cost parameters, and power variation cost parameters corresponding to the target computing power task, the target power configuration corresponding to the computing power system is determined. This includes: determining the equipment operation data corresponding to each of the multiple computing power devices; based on the equipment operation data, power operation data, and power supply topology corresponding to each of the multiple computing power devices, determining the first power association constraint corresponding to each of the multiple target devices, and the second power association constraint corresponding to the computing power system. The first power association constraint is the constraint corresponding to the power association between the input power and output power of the corresponding target device, and the second power association constraint is the constraint corresponding to the power association between the power supply power of the power system and the power consumption power of the computing power system; based on the first power association constraint, the second power association constraint, the power demand parameters, power deviation cost parameters, and power variation cost parameters corresponding to each of the multiple target devices, the target power configuration corresponding to the computing power system is determined. The multiple target devices include multiple computing power devices and multiple power devices.

[0067] This includes equipment operation data, which is used to characterize the data generated by each computing device (such as server, storage device, network device, etc.) in the computing power system during operation. This includes, but is not limited to, the device's input power, output power, operating voltage, operating current, load rate, running time, etc., and is used to reflect the power consumption characteristics and operating status of a single device.

[0068] This involves a first power correlation constraint, which is a power balance constraint for the corresponding target device, used to limit the correspondence between the device's input power and output power.

[0069] This involves a second power correlation constraint, which is an overall power balance constraint between the power system and the computing system. It is used to limit the matching relationship between the total power supply provided by the power system and the total power consumption of the computing system.

[0070] For example, the first power correlation constraint and the second power view constraint can be represented by DC network power flow and power balance constraints, where the first power correlation constraint can be expressed as:

[0071]

[0072] in, Inject a power vector into the node (i.e., the target device); This is the conductance matrix of a DC network, which consists of the conductance of DC lines and the connection relationships between nodes, and describes the topological conductance characteristics of a DC power grid. The node voltage vector; This is the power loss vector, which can specifically be the converter power loss vector; This is the branch current vector.

[0073] The following formula can be used to determine this:

[0074]

[0075] in, The converter power loss of the i-th node in time period k; Let be the loss coefficient of the converter corresponding to the i-th node; Let be the DC voltage of the i-th node during time period k; Let be the current of the i-th node during time period k.

[0076] The second power correlation constraint can be expressed as:

[0077]

[0078] in, Let k be the power generation capacity of the photovoltaic equipment during time period k; Let k be the discharge power of the energy storage device during time period k. Let k be the rigid computing load power corresponding to the rigid computing task in time period k. The adjustable computing power load power corresponds to the adjustable computing power task in time period k; Let k be the load power of the refrigeration system (i.e., the cooling equipment) during time period k; Let k be the charging power of the energy storage device during time period k.

[0079] Among them, multiple power equipment include photovoltaic equipment, energy storage equipment, and computing systems include cooling systems.

[0080] By acquiring the operational data of each computing device, we can accurately grasp the power consumption characteristics and operating status of the corresponding computing devices. Based on the device operation data, power operation data, and power supply topology, we can determine the first and second power correlation constraints. This allows us to accurately characterize the input-output power balance relationship of the corresponding computing devices, as well as the overall power matching relationship between the power system and the computing system. This provides physical-level security and balance constraints for power configuration, enabling the target power configuration to meet both device-level and system-level power constraints while taking into account the power adjustment costs of different computing tasks. This ensures that the operating power of computing devices is finely allocated within a safe and compliant framework, avoiding power overload or system power imbalance, and guaranteeing the stable and coordinated operation of the computing system and the power system.

[0081] As an optional embodiment, the target power configuration corresponding to the computing power system is determined based on the power demand parameters, power operation data, power supply topology, power deviation cost parameters, and power variation cost parameters corresponding to the target computing power task. This includes: determining the computing power energy efficiency parameters corresponding to the target computing power task, wherein the computing power energy efficiency parameters characterize the task progress completed by the corresponding computing power device per unit of power consumed; determining the task execution period corresponding to the target computing power task; determining the progress execution constraints corresponding to the target computing power task based on the task execution period, computing power energy efficiency parameters, and the current task progress corresponding to the target computing power task; and determining the target power configuration corresponding to the computing power system based on the progress execution constraints, the power demand parameters, power operation data, power supply topology, power deviation cost parameters, and power variation cost parameters corresponding to the target computing power task.

[0082] This includes computing power energy efficiency parameters, which are parameters that characterize the progress of computing tasks completed by computing power equipment per unit of power consumed, and are used to reflect the correspondence between the power consumption of computing power equipment and the efficiency of task completion.

[0083] This involves the task execution period, which is the actual time interval during which the target computing power task is executed, and is used to limit the execution time range of the computing power task.

[0084] This involves schedule execution constraints, which are constraints used to ensure that the target computing power task completes its schedule as required, thereby limiting the matching relationship between the execution progress and power input of the target computing power task within a specified period.

[0085] By determining the computing power energy efficiency parameters, the correlation between the power consumption of computing equipment and the task progress can be quantified. By combining the task execution period with the current task progress, a progress execution constraint can be constructed. Then, by combining this constraint with power demand, power operation, power supply topology, and two types of power cost parameters, the target power configuration can be determined. This can avoid power waste or progress delays while ensuring the execution progress of computing tasks, thereby achieving a precise match between power input and task completion efficiency, and improving the rationality and reliability of the computing power system power configuration.

[0086] For example, schedule execution constraints can be represented as:

[0087]

[0088] in, Let $\frac{i}{i}$ be the cumulative computational amount completed by the i-th rigid computing task at time step $k$, with an initial value of $\frac{i}{i}$. ; Let be the computational energy efficiency factor (i.e., the computational energy efficiency parameter of the rigid computing task) for the i-th rigid computing task. Let be the running power of the i-th rigid computing task at time step k; The time interval between two power adjustments; For the i-th rigid computing task at the deadline The cumulative amount of calculations completed at any given time; This represents the total computational requirements / total workload for the i-th rigid computing task.

[0089]

[0090] in, Let $\frac{j}{j}$ be the cumulative computational amount completed by the j-th adjustable computing power task at time step $k$, with an initial value of $\frac{j}{j}$. ; Let be the computational energy efficiency factor (i.e., the computational energy efficiency parameter of the adjustable computing power task) for the j-th adjustable computing power task. The computing power allocated to the j-th adjustable computing power task at time step k; This represents the cumulative computational load of the j-th adjustable computing power task at the expected completion time and maximum delay time. The total computational requirement / total workload for the j-th adjustable computing power task; Let be the expected completion time of the j-th adjustable computing power task; This represents the maximum allowed latency for the j-th adjustable computing power task.

[0091] As an optional embodiment, based on the power demand parameters corresponding to the target computing power task, power operation data, power supply topology, power deviation cost parameters, and power variation cost parameters, the target power configuration corresponding to the computing power system is determined, including: determining the cooling energy efficiency parameters and environmental parameters corresponding to the computing power system, wherein the cooling energy efficiency parameters represent the degree of cooling of the computing power system per unit of power consumed; determining the cooling power constraints corresponding to the computing power system based on the cooling energy efficiency parameters and environmental parameters; and determining the target power configuration corresponding to the computing power system based on the cooling power constraints, the power demand parameters corresponding to the target computing power task, power operation data, power supply topology, power deviation cost parameters, and power variation cost parameters.

[0092] This involves cooling energy efficiency parameters, which are used to characterize the cooling effect of the computing power system that can be achieved per unit of power consumed by the cooling equipment. Specifically, it can be expressed by the energy efficiency ratio of the cooling equipment, reflecting the correspondence between cooling power and cooling effect.

[0093] This involves environmental parameters, which are parameters that characterize the operating environment of computing equipment, including ambient temperature (such as reference temperature, actual ambient temperature, and predicted ambient temperature).

[0094] This includes a cooling power constraint, which is a constraint used to limit the matching relationship between the operating power of the cooling equipment and the total heat generation of the computing system and environmental parameters (such as ambient temperature).

[0095] By determining the cooling energy efficiency parameters and environmental parameters, the correspondence between cooling power and cooling effect and environmental conditions can be quantified. The cooling power constraints determined can limit the matching relationship between cooling power and computing system heat generation and ambient temperature. This further enables the coordinated optimization of computing task power and cooling power while ensuring the temperature safety of the computing system, avoiding cooling power redundancy or overheating of computing equipment, and improving the rationality of power configuration and system operation stability.

[0096] For example, the cooling power constraint can be expressed as:

[0097]

[0098] in, This is a function of the energy efficiency ratio of the refrigeration system (i.e., cooling equipment); The predicted ambient temperature for time period k; The power utilization efficiency of computing equipment; Let k be the ambient temperature during time period k; Let k be the total input power of the computing devices during time period k. For the refrigeration system (at the reference temperature) (Below) Rated operating condition energy efficiency ratio; This is a temperature correction factor; This represents the actual ambient temperature.

[0099] As an optional embodiment, the target power configuration corresponding to the computing power system is determined based on the power demand parameters corresponding to the target computing power task, power operation data, power supply topology, power deviation cost parameters, and power variation cost parameters. This includes: in the case of multiple power devices, including a first device and a second device, determining the carbon emission constraints corresponding to the computing power system based on the carbon emission index corresponding to the first device and the target carbon emission of the power system, wherein the first device is a power device with a carbon emission index greater than or equal to a carbon emission threshold, and the second device is a power device with a carbon emission index less than a carbon emission threshold; determining the power supply stability characteristics corresponding to the second device; determining the power supply stability constraints corresponding to the computing power system based on the power supply stability characteristics; and determining the target power configuration corresponding to the computing power system based on the carbon emission constraints, power supply stability constraints, power demand parameters corresponding to the target computing power task, power operation data, power supply topology, power deviation cost parameters, and power variation cost parameters.

[0100] Among them, the first piece of equipment is high-carbon emission power equipment, specifically power generation / supply equipment with a carbon emission index greater than or equal to a preset carbon emission threshold (such as coal-fired power plants, oil-fired power units, etc.).

[0101] Among them, a second type of equipment is involved, which is a low-carbon / zero-carbon emission power equipment, referring to power generation / supply equipment with a carbon emission index less than a preset carbon emission threshold (such as photovoltaic equipment, wind power equipment, hydropower equipment and other new energy power generation equipment).

[0102] This includes the carbon emission index, which is a parameter that characterizes the carbon emissions generated per unit of electricity supplied by power equipment and is used to quantify the carbon emission intensity of different power equipment.

[0103] This includes the target carbon emission limit, which is the upper limit of the total carbon emissions allowed by the power system within a specified period. It is the target value for carbon emission control of the power system.

[0104] This includes carbon emission constraints, which are used to ensure that the overall carbon emissions of the system do not exceed the target carbon emission level.

[0105] This involves a carbon emission threshold, which is a pre-set critical value for carbon emission intensity, used to delineate the boundary between high-carbon emission equipment and low-carbon / zero-carbon emission equipment.

[0106] This involves power supply stability characteristics, which are parameters characterizing the power supply reliability and output fluctuation characteristics of the second device (low-carbon / zero-carbon power equipment), including output power fluctuation rate, power outage probability, voltage / frequency stability, etc.

[0107] This includes power supply stability constraints, which are constraints determined based on the power supply stability characteristics of the second device. These constraints are used to limit the power supply ratio or call range of low-carbon / zero-carbon power equipment, ensuring the overall power supply stability of the power system and preventing power shortages or excessive fluctuations.

[0108] By distinguishing between high-carbon-emission first equipment and low-carbon / zero-carbon-emission second equipment, and setting carbon emission constraints based on their carbon emission index and target carbon emission, while formulating power supply stability constraints based on the power supply stability characteristics of the second equipment, it is possible to limit the use of high-carbon-emission equipment and ensure the reliability of power supply for low-carbon equipment while meeting the power requirements of computing tasks. This reduces the overall carbon emissions of the system and avoids additional power configuration costs caused by fluctuations in low-carbon equipment.

[0109] For example, carbon emission constraints can be expressed as:

[0110]

[0111] in, The total load power during time period k; This represents the upper limit of carbon emission intensity corresponding to time period k; Let be the load power of the i-th time period k; Let be the predicted carbon emission intensity corresponding to the unit power of the i-th node in time period k; For the set of load nodes; The hour index corresponding to time period k.

[0112] The constraints corresponding to the power supply stability characteristics, that is, the power supply stability constraints, can be expressed as:

[0113]

[0114] in, Probability symbols represent the probability of an event occurring. The actual photovoltaic power generation during time period k; For time period k, the predicted photovoltaic power generation is given. This represents the error boundary for photovoltaic power prediction. The allowed probability threshold.

[0115] As an optional embodiment, the target power configuration corresponding to the computing power system is determined based on the power demand parameters corresponding to the target computing power task, power operation data, power supply topology, power deviation cost parameters, and power variation cost parameters. This includes: determining a first security constraint corresponding to the computing power system and a second security constraint corresponding to the power system; and determining the target power configuration corresponding to the computing power system based on the first security constraint, the second security constraint, the power demand parameters corresponding to the target computing power task, power operation data, power supply topology, power deviation cost parameters, and power variation cost parameters.

[0116] This involves a first security constraint, which is a device-level security operation constraint on the computing power system side. It is used to limit the operating parameters of computing power equipment (such as servers, storage devices, network devices, etc.) within a safe threshold. For example, it includes the upper limit of device input / output power, operating voltage / current range, operating temperature range, etc., to avoid device overload, overheating or electrical failure, and to ensure the stable and reliable operation of computing power equipment.

[0117] This involves a second safety constraint, which is a system-level safety operation constraint on the power system side. It is used to limit the operating parameters of power equipment, power supply lines and nodes within safety specifications. For example, it includes the upper limit of line transmission power, the range of node voltage / current fluctuations, transformer capacity limits, etc., to avoid power exceeding limits, short circuits or power outages, and to ensure the overall power supply safety and stability of the power system.

[0118] By defining the first security constraint on the computing power system side and the second security constraint on the power system side, the operating parameters of the computing power equipment and the power system can be limited within safe thresholds to avoid equipment overload, overheating, or power exceeding limits and interruption. Then, by combining the two types of security constraints with power demand, power operation, power supply topology, and the two types of power cost parameters to determine the target power configuration, it is possible to ensure that both the computing power equipment and the power system are in a safe and stable operating state while meeting the requirements for computing power task execution and the cost of power adjustment, thus achieving reliable operation of computing power and power in synergy.

[0119] For example, the first safety constraint can be expressed as:

[0120]

[0121] in, The minimum operating power for the target computing task; The actual operating power of the i-th rigid computing task at time k; The actual operating power of the j-th adjustable computing power task at time k; The maximum operating power of the target computing task.

[0122]

[0123] in, The total input power of computing devices during time period k; Let k be the rigid computing load power corresponding to the rigid computing task in time period k. The adjustable computing power load power corresponds to the adjustable computing power task in time period k; Let be the running power of the i-th rigid computing task at time step k; The computing power allocated to the j-th adjustable computing power task at time step k.

[0124] The second safety constraint can be expressed as:

[0125]

[0126] in, It represents the minimum instantaneous feasible power at time k (i.e., time period k); Let be the interaction power at time k; The maximum instantaneous feasible power at time k; Let k be the minimum discharge power of the energy storage device during time period k. Let k be the discharge power of the energy storage device during time period k. Let k be the maximum discharge power of the energy storage device during time period k. Let k be the minimum charging power of the energy storage device during time period k. The charging power of the energy storage device during time period k; Let k be the maximum charging power of the energy storage device during time period k. Let k be the minimum power generation capacity of the photovoltaic equipment during time period k. Let k be the power generation capacity of the photovoltaic equipment during time period k; Let k be the maximum power generation capacity of the photovoltaic equipment during time period k.

[0127]

[0128] in, The minimum DC voltage during time period k; Let be the DC voltage during time period k; This represents the maximum DC voltage during time period k.

[0129] Based on the above embodiments and optional embodiments, an optional implementation method is provided, which is described in detail below.

[0130] In related technologies, to achieve coordinated scheduling and stable operation of power systems and computing systems, meet the execution requirements of computing tasks, and adapt to the power supply capacity and operating status of the power side, power configuration of the computing system is necessary. However, in related technologies, power configuration of the computing system based on power and computing collaboration suffers from the technical problem of high power configuration costs.

[0131] There is currently no effective solution to the above problems.

[0132] In view of this, an optional embodiment of the present invention provides a power configuration method that coordinates power and computing power, which can effectively solve the above-mentioned technical problems.

[0133] Figure 2 This is a schematic diagram of the power configuration process in an optional embodiment of the present invention. Figure 3 This is a schematic diagram of the power configuration process of fusion constraints in an optional embodiment of the present invention. Figure 4 This is a schematic diagram of the regional association relationship in an optional embodiment of the present invention, such as... Figure 2 , Figure 3 and Figure 4 As shown, a detailed description follows.

[0134] Target time period under receive power configuration period ( The system includes real-time boundary constraints, current system status (energy storage state of charge, bus voltage), and ultra-short-term forecast data (PV output, load, and temperature for the next 30 minutes). This is the electricity cost coefficient. Carbon emission intensity.

[0135] Establish the Model Predictive Control (MPC) optimization problem, including:

[0136] Construct the objective function in the prediction time domain Minimize the total cost (i.e., power configuration cost) during the power configuration period:

[0137]

[0138] in, The parameter representing the power interaction cost at time k includes the interaction power at time k. Interaction power cost weighting coefficient at time k , For can be by Interpolation is obtained; This represents the penalty weighting coefficient (i.e., the energy storage SOC deviation penalty coefficient) indicating that the energy storage device's state of charge (SOC) deviates from the target value. This represents the state of charge of the energy storage device at time k; This represents the power deviation cost parameter at time k, including the penalty weighting coefficient for the power deviation of the rigid computing power task (i.e., the first computing power task) at time k. The actual operating power (i.e., the actual allocated power) of the i-th rigid computing task at time k. and the reference operating power of the i-th rigid computing task Where R is the total number of rigid computing power tasks; This represents the power variation cost parameter, including the penalty weighting coefficient for power deviation in adjustable computing power tasks (i.e., the second computing power task). The actual operating power (i.e., the actual allocated power) of the j-th adjustable computing power task at time k. The reference operating power of the j-th adjustable computing power task at time k. Penalty weighting coefficient for adjustable computing power task execution latency The total computational load / total workload of the j-th adjustable computing power task , In order to delay duration The following is the amount of computation completed for the j-th adjustable computing power task, where M is the total number of adjustable computing power tasks; The penalty weighting coefficient is used to constrain violations; Let be the norm of all constraint violations at time k, representing the degree to which the constraints are broken; Let k be the vector of constrained slack variables.

[0139] Among them, power equipment includes energy storage equipment, and k is the discrete-time index in rolling optimization.

[0140] Further, it includes: based on the regional correlation between the first region where the computing power system is located and the second region where the power system is located, power operation data, and power supply topology, determining the power interaction cost parameters corresponding to multiple computing power devices respectively. Among them, the regional correlation includes: computing power demand and supply correlation, computing power task scheduling correlation, power consumption demand correlation, energy cost complementarity correlation, cross-regional power grid transmission correlation, and carbon emission cost control correlation.

[0141] Regarding the supply and demand relationship of computing power, the first region, as the place where computing power demand is generated, cannot meet the local explosive computing power development demand due to resource constraints such as land, energy and environment. Therefore, it is necessary to transfer massive computing power tasks across regions to the second region, and the power system of the second region supports the operation of the computing power system to achieve cross-regional supply matching of computing power demand.

[0142] Regarding the scheduling of computing power tasks, the first region has the right to initiate, allocate, and schedule computing power tasks. It defines the execution priority and time period requirements (rigid / adjustable) of computing power tasks according to its own business needs, and distributes the target computing power tasks to the computing power system of the second region. The power system of the second region then carries out targeted power configuration based on the scheduling requirements.

[0143] Regarding the correlation of power consumption demand, namely the demand for clean and low-cost power consumption in the first region, the computing power of the first region is transformed into a computer-based collaborative model for power calculation in the second region, which transforms the demand for new energy power consumption in the second region into a targeted demand for new energy power consumption in the second region. Relying on the abundant new energy resources such as photovoltaic and wind power in the second region, the green power demand behind the computing power task in the first region is met.

[0144] Regarding the complementary relationship of energy costs, the first region faces the current situation of high electricity prices and high energy usage costs. Its computing power industry's need to reduce costs complements the advantages of the second region's low electricity prices and low new energy power generation costs. Through cross-regional computing collaboration, the computing power tasks of the first region can achieve low-cost operation by leveraging the power system of the second region.

[0145] For cross-regional power grid transmission linkage, the first region and the second region can achieve power interconnection through ultra-high voltage AC / DC power grids. The first region can coordinate with the power system of the second region to carry out cross-regional power dispatch according to the real-time power demand of the computing power system. If the output of new energy in the second region is insufficient, the first region can supplement the power through the cross-regional power grid to ensure the stable operation of the computing power system.

[0146] Regarding the correlation of carbon emission cost control, the carbon emission reduction control requirements of the first region are higher than those of the second region. As a high-energy-consuming industry, the computing power industry in the first region needs to reduce its own carbon emission intensity through cross-regional computing collaboration and take advantage of the low-carbon emission advantages of new energy power generation in the second region, so as to achieve coordinated control of carbon emission costs between the first and second regions.

[0147] Next, we construct the constraint set, including:

[0148] DC network power flow and power balance constraints:

[0149]

[0150] in, Inject a power vector (kW) into the node (i.e., the target device); S is the conductance matrix of a DC network, which is composed of the conductance of DC lines and the connection relationship of nodes, and describes the topological conductance characteristics of a DC power grid. The node voltage vector (kV); This is the power loss vector (kW), which can specifically be the converter power loss vector; Let A be the branch current vector (A).

[0151] The following formula can be used to determine this:

[0152]

[0153] in, The converter power loss of the i-th node in time period k; Let be the loss coefficient of the converter corresponding to the i-th node; Let be the DC voltage of the i-th node during time period k; Let be the current of the i-th node during time period k.

[0154]

[0155] in, Let k be the power generation capacity of the photovoltaic equipment during time period k; Let k be the discharge power of the energy storage device during time period k. Let k be the rigid computing load power corresponding to the rigid computing task in time period k. The adjustable computing power load power corresponds to the adjustable computing power task in time period k; Let k be the load power of the refrigeration system (i.e., the cooling equipment) during time period k; Let k be the charging power of the energy storage device during time period k.

[0156] Schedule execution constraints can be represented by computer-computer coupling constraints. Suppose the intelligent computing center (i.e., the computing power system) has R rigid computing tasks, where the total computational requirement of rigid computing task i is... The deadline is Its calculation progress The formula is:

[0157]

[0158] in, Let $\frac{i}{i}$ be the cumulative computational amount completed by the i-th rigid computing task at time step $k$, with an initial value of $\frac{i}{i}$. ; Let be the computational energy efficiency factor (i.e., the computational energy efficiency parameter of the rigid computing task) for the i-th rigid computing task. Let be the running power of the i-th rigid computing task at time step k; The time interval between two power adjustments; For the i-th rigid computing task at the deadline The cumulative amount of calculations completed at any given time; This represents the total computational requirements / total workload for the i-th rigid computing task.

[0159] There are M adjustable computing power tasks, and the total computing power requirement of adjustable computing power task j is... The expected completion time is Define the computational power allocated to task j in time period k as follows: Its calculation progress satisfy:

[0160]

[0161] in, Let $\frac{j}{j}$ be the cumulative computational amount completed by the j-th adjustable computing power task at time step $k$, with an initial value of $\frac{j}{j}$. ; The computational energy efficiency factor (i.e., the computational energy efficiency parameter of the adjustable computing power task) is the computational power of the j-th adjustable computing power task, representing the amount of computation that can be completed per unit of electrical energy, calibrated through offline benchmark testing. The computing power allocated to the j-th adjustable computing power task at time step k; This represents the cumulative computational load of the j-th adjustable computing power task at the expected completion time and maximum delay time. The total computational requirement / total workload for the j-th adjustable computing power task; Let be the expected completion time of the j-th adjustable computing power task; Let j be the maximum allowed latency time for the j-th adjustable computing power task. .

[0162] Rigid computing tasks and adjustable computing tasks share server cluster resources. Let the maximum and minimum operating power of the server cluster be respectively... and .

[0163] The sum of the power consumption of all computing tasks on the server cluster must not exceed the cluster's physical limit, and should be higher than the minimum operating power to ensure equipment safety. The first safety constraint can be expressed as:

[0164]

[0165] in, The minimum operating power for the target computing task; The actual operating power of the i-th rigid computing task at time k; The actual operating power of the j-th adjustable computing power task at time k; The maximum operating power of the target computing task.

[0166] Total computing power of the computing system in time period k It is the sum of the power of rigid computing tasks and adjustable computing tasks, and the following conditions must be met:

[0167]

[0168] in, Let k be the total input power of the computing devices during time period k. Let k be the rigid computing load power corresponding to the rigid computing task in time period k. The adjustable computing power load power corresponds to the adjustable computing power task in time period k; Let be the running power of the i-th rigid computing task at time step k; The computing power allocated to the j-th adjustable computing power task at time step k.

[0169] Correspondingly, the second safety constraint is specifically a device operation constraint, which can be expressed as:

[0170]

[0171] in, It represents the minimum instantaneous feasible power at time k (i.e., time period k); Let be the interaction power at time k; The maximum instantaneous feasible power at time k; Let k be the minimum discharge power of the energy storage device during time period k. Let k be the discharge power of the energy storage device during time period k. Let k be the maximum discharge power of the energy storage device during time period k. Let k be the minimum charging power of the energy storage device during time period k. The charging power of the energy storage device during time period k; Let k be the maximum charging power of the energy storage device during time period k. Let k be the minimum power generation capacity of the photovoltaic equipment during time period k. Let k be the power generation capacity of the photovoltaic equipment during time period k; Let k be the maximum power generation capacity of the photovoltaic equipment during time period k.

[0172]

[0173] in, The minimum DC voltage during time period k; Let be the DC voltage during time period k; This represents the maximum DC voltage during time period k.

[0174] The energy storage constraints corresponding to energy storage devices can be expressed as:

[0175] Energy storage dynamic constraints are expressed as follows:

[0176]

[0177] in, The state of charge of the energy storage node (i.e., the energy storage device) during time period h; The charging power for energy storage nodes; The charging power of the energy storage node within the time period h; The discharge power of the energy storage node during the time period h; This represents the discharge power of the energy storage node. Configure the capacity of the energy storage nodes; This is the state-of-charge limit for energy storage nodes; This represents the upper limit of the state of charge of the energy storage node; The charging and discharging state of the energy storage node is a 0-1 variable (0 represents discharging, 1 represents charging). The maximum charging and discharging power of the energy storage node is related to the configured capacity of the energy storage node.

[0178] Cooling power constraint, also known as cold-electric coupling constraint, is used to establish a quantitative relationship between the power consumption of the refrigeration system, the heat generation of computing equipment, and the building's heat load. It can be expressed as:

[0179]

[0180] in, This is a function of the energy efficiency ratio of the refrigeration system (i.e., cooling equipment); The predicted ambient temperature for time period k; The power utilization efficiency of computing equipment; Let k be the ambient temperature during time period k; Let k be the total input power of the computing devices during time period k. For the refrigeration system (at the reference temperature) (Below) Rated operating condition energy efficiency ratio; This is a temperature correction factor; This represents the actual ambient temperature.

[0181] Carbon emission constraints (i.e., carbon emission intensity constraints) can be expressed as:

[0182]

[0183] in, The total load power during time period k; This represents the upper limit of carbon emission intensity corresponding to time period k; Let be the load power of the i-th time period k; The predicted carbon emission intensity per unit power of the i-th node in time period k can be estimated based on the current carbon flow status and future scheduling plan; For the set of load nodes; The hour index corresponding to time period k.

[0184] The power supply stability characteristics can be handled using chance constraints to address the uncertainty of the output of secondary equipment, such as photovoltaic output. Therefore, the constraints corresponding to the power supply stability characteristics can be expressed as follows:

[0185]

[0186] in, Probability symbols represent the probability of an event occurring. The actual photovoltaic power generation during time period k; For time period k, the predicted photovoltaic power generation is given. This represents the error boundary for photovoltaic power prediction. This is the minimum allowed probability threshold.

[0187] Furthermore, by solving the objective function, a sequence of DC voltage setpoints is obtained. Precise power setpoint sequence for each controllable device (such as modular multilevel converter, energy storage converter, photovoltaic converter). Power allocation scheme for each computing task .

[0188] Subsequently, the DC voltage setpoint sequence and the equipment power setpoint sequence are sent to the real-time control layer. Based on the latest optimization results and system status, the adjustable potential boundary of the system for the next few hours is further predicted, as follows:

[0189]

[0190] in, This represents the flexible power limit that the computing system can adjust upwards under the hour index h (i.e., the time period h); This represents the flexible lower limit of the computing power system that can be adjusted during the time period h. The energy storage discharge power during time period h; The maximum operating power of the server cluster during time period h; The energy storage charging power during time period h; The total computing power of the computing system during time period h; This represents the minimum operating power of the server cluster during time period h. The maximum dischargeable power of energy stored during time period h; The maximum rechargeable power of energy storage during time period h; The photovoltaic power that can be limited during the time period h is the difference between the predicted photovoltaic output and the minimum technical output (usually 0). This is to provide a power reserve margin for the system.

[0191] Will The corresponding strategy configuration layer is provided for reference in the next cycle.

[0192] The above optional implementation methods can achieve at least the following beneficial effects:

[0193] (1) Compared with related technologies, this invention can provide a basis for subsequent matching of computing power equipment and power supply correlation analysis by acquiring target computing power tasks and power operation data; it can accurately lock the core object of power regulation by determining the corresponding computing power equipment based on the target computing power tasks; it can clearly depict the power supply correlation between computing power equipment and power equipment by determining the power supply topology based on power operation data and computing power equipment, and provide a physical constraint framework for power configuration; it can quantify the power adjustment cost of different types of computing power tasks by determining the power deviation cost parameter of the first computing power task and the power change cost parameter of the second computing power task; and finally, it can determine the target power configuration based on multi-dimensional parameters, and achieve refined configuration of computing power equipment operation power under the premise of meeting the computing power task execution requirements and power system operation constraints, avoid unnecessary power loss, reduce the overall power configuration cost, and thus solve the technical problem of high power configuration cost when configuring the power system based on power and computing power collaboration in related technologies.

[0194] (2) Compared with related technologies, this invention determines the power interaction cost parameters of each computing device by combining power operation data and power supply topology. It can accurately quantify the power transmission cost between a single computing device and the corresponding power device. Then, it combines the power requirement parameters of computing tasks and the power cost parameters of the two types of computing tasks to formulate the target power configuration. This allows the power allocation scheme to take into account the actual power interaction loss of computing devices and the power adjustment cost of different computing tasks. It ensures that the power configuration of the computing system fits the actual transmission characteristics of cross-regional power supply, realizes the refinement and accuracy of power configuration, and further avoids unnecessary power loss, making the power configuration of the computing system more in line with the actual operation scenario.

[0195] (3) Compared with related technologies, this invention can quantify the correspondence between the power consumption of computing equipment and the progress of tasks by determining the computing power energy efficiency parameters. It can construct the progress execution constraint by combining the task execution period and the current task progress, and then combine the constraint with power demand, power operation, power supply topology and two types of power cost parameters to determine the target power configuration. Under the premise of ensuring the progress of computing power task execution, it can avoid power waste or progress delay, thereby achieving a precise match between power input and task completion efficiency, and improving the rationality and reliability of the power configuration of the computing power system.

[0196] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0197] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0198] Example 2

[0199] According to embodiments of the present invention, an apparatus for implementing the above-described power configuration method for coordinating power and computing power is also provided. Figure 5 This is a structural block diagram of a power configuration device for power and computing coordination according to an embodiment of the present invention, such as... Figure 5 As shown, the device includes: an acquisition module 502, a first determination module 504, a second determination module 506, a third determination module 508, and a fourth determination module 510. The device will be described in detail below.

[0200] The acquisition module 502 is used to acquire the target computing power task corresponding to the computing power system and the power operation data corresponding to the power system during the power configuration period.

[0201] The first determining module 504 is connected to the above-mentioned obtaining module 502 and is used to determine multiple computing devices corresponding to the computing power system based on the target computing power task.

[0202] The second determining module 506 is connected to the first determining module 504 and is used to determine the power supply topology corresponding to the computing system based on power operation data and multiple computing devices. The power supply topology is used to characterize the power supply association between the multiple computing devices and multiple power devices in the power system.

[0203] The third determining module 508 is connected to the second determining module 506 and is used to determine the power deviation cost parameter corresponding to the first computing power task and the power variation cost parameter corresponding to the second computing power task. The first computing power task is a computing power task executed in a fixed period of time, and the second computing power task is a computing power task executed in a non-fixed period of time. The target computing power task includes the first computing power task and the second computing power task.

[0204] The fourth determining module 510, connected to the third determining module 508, is used to determine the target power configuration corresponding to the computing power system based on the power demand parameters, power operation data, power supply topology, power deviation cost parameters, and power variation cost parameters corresponding to the target computing power task, so as to configure the operating power of multiple computing power devices.

[0205] It should be noted here that the above-mentioned acquisition module 502, first determination module 504, second determination module 506, third determination module 508, and fourth determination module 510 correspond to steps S102 to S110 in the power configuration method for implementing power and computing power collaboration. The multiple modules and the corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in the above embodiment 1.

[0206] Example 3

[0207] According to another aspect of the present invention, an electronic device is also provided, comprising: a processor; and a memory for storing processor-executable instructions, wherein the processor is configured to execute instructions to implement the power configuration method for power and computing power coordination as described above.

[0208] Example 4

[0209] According to another aspect of the present invention, a computer-readable storage medium is also provided, which, when the instructions in the computer-readable storage medium are executed by a processor of an electronic device, enables the electronic device to perform any of the above-described power and computing power coordinated power configuration methods.

[0210] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0211] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0212] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0213] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0214] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0215] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0216] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A power allocation method for the coordinated use of electricity and computing power, characterized in that, include: Acquire the target computing power tasks corresponding to the computing power system and the power operation data corresponding to the power system during the power configuration period; Based on the target computing power task, multiple computing power devices corresponding to the computing power system are identified; Based on the power operation data and the multiple computing devices, a power supply topology corresponding to the computing system is determined, wherein the power supply topology is used to characterize the power supply association between the multiple computing devices and multiple power devices in the power system. Determine the power deviation cost parameter corresponding to the first computing power task and the power variation cost parameter corresponding to the second computing power task, wherein the first computing power task is a computing power task executed in a fixed period of time, the second computing power task is a computing power task executed in a non-fixed period of time, and the target computing power task includes the first computing power task and the second computing power task. Based on the power requirement parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters, a target power configuration corresponding to the computing power system is determined for configuring the operating power of the multiple computing power devices.

2. The method according to claim 1, characterized in that, The determination of the target power configuration corresponding to the computing power system based on the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters includes: Based on the power operation data and the power supply topology, power interaction cost parameters corresponding to the multiple computing devices are determined respectively, wherein the corresponding power interaction cost parameters are the cost parameters generated when the corresponding computing device interacts with the corresponding power device. Based on the power interaction cost parameter, the power requirement parameter corresponding to the target computing power task, the power deviation cost parameter, and the power variation cost parameter, the target power configuration corresponding to the computing power system is determined.

3. The method according to claim 1, characterized in that, The determination of the target power configuration corresponding to the computing power system based on the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters includes: Determine the device operation data corresponding to each of the plurality of computing devices; Based on the device operation data corresponding to the multiple computing devices, the power operation data, and the power supply topology, a first power association constraint corresponding to the multiple target devices and a second power association constraint corresponding to the computing system are determined. The first power association constraint is the constraint corresponding to the power association between the input power and the output power of the corresponding target device, and the second power association constraint is the constraint corresponding to the power association between the power supply power of the power system and the power consumption power of the computing system. Based on the first power association constraint, the second power association constraint, the power requirement parameter corresponding to the target computing power task, the power deviation cost parameter, and the power variation cost parameter corresponding to the multiple target devices, the target power configuration corresponding to the computing power system is determined.

4. The method according to claim 1, characterized in that, The determination of the target power configuration corresponding to the computing power system based on the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters includes: Determine the computing power energy efficiency parameters corresponding to the target computing power task, wherein the computing power energy efficiency parameters characterize the task progress completed by the corresponding computing power device per unit of power consumed; Determine the task execution period corresponding to the target computing power task; Based on the task execution period, the computing power efficiency parameters, and the current task progress corresponding to the target computing power task, determine the progress execution constraints corresponding to the target computing power task; Based on the schedule execution constraints, the power requirement parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters, the target power configuration corresponding to the computing power system is determined.

5. The method according to claim 1, characterized in that, The determination of the target power configuration corresponding to the computing power system based on the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters includes: Determine the cooling energy efficiency parameters and environmental parameters corresponding to the computing power system, wherein the cooling energy efficiency parameters represent the degree of cooling of the computing power system per unit of power consumed; Based on the cooling energy efficiency parameters and the environmental parameters, the cooling power constraint corresponding to the computing power system is determined; Based on the cooling power constraint, the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters, the target power configuration corresponding to the computing power system is determined.

6. The method according to claim 1, characterized in that, The determination of the target power configuration corresponding to the computing power system based on the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters includes: In the case where the plurality of power devices include a first device and a second device, a carbon emission constraint corresponding to the computing power system is determined based on the carbon emission index corresponding to the first device and the target carbon emission amount corresponding to the power system, wherein the first device is a power device with a carbon emission index greater than or equal to a carbon emission threshold, and the second device is a power device with a carbon emission index less than a carbon emission threshold. Determine the power supply stability characteristics corresponding to the second device; Based on the power supply stability characteristics, determine the power supply stability constraints corresponding to the computing power system; Based on the carbon emission constraints, the power supply stability constraints, the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters, the target power configuration corresponding to the computing power system is determined.

7. The method according to any one of claims 1 to 6, characterized in that, The determination of the target power configuration corresponding to the computing power system based on the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters includes: Determine the first security constraint corresponding to the computing power system and the second security constraint corresponding to the power system; Based on the first security constraint, the second security constraint, the power requirement parameter corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameter, and the power variation cost parameter, the target power configuration corresponding to the computing power system is determined.

8. A power configuration device that coordinates electricity and computing power, characterized in that, include: The acquisition module is used to acquire the target computing power tasks corresponding to the computing power system and the power operation data corresponding to the power system during the power configuration period. The first determining module is used to determine multiple computing devices corresponding to the computing system based on the target computing power task; The second determining module is used to determine the power supply topology corresponding to the computing system based on the power operation data and the plurality of computing devices, wherein the power supply topology is used to characterize the power supply association relationship between the plurality of computing devices and the plurality of power devices in the power system. The third determining module is used to determine the power deviation cost parameter corresponding to the first computing power task and the power variation cost parameter corresponding to the second computing power task, wherein the first computing power task is a computing power task executed in a fixed period of time, the second computing power task is a computing power task executed in a non-fixed period of time, and the target computing power task includes the first computing power task and the second computing power task. The fourth determining module is used to determine the target power configuration corresponding to the computing power system based on the power demand parameters corresponding to the target computing power task, the power operation data, the power supply topology, the power deviation cost parameters, and the power variation cost parameters, so as to configure the operating power of the multiple computing power devices.

9. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the instructions to implement the power configuration method for power and computing power coordination as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, When the instructions in the computer-readable storage medium are executed by the processor of the electronic device, the electronic device is able to perform the power configuration method of power and computing power coordination as described in any one of claims 1 to 7.

Images