Power supply control method and device for data center

Abstract

The application provides a power supply control method and device of a data center, the data center comprising a backup battery. The method comprises: predicting a power consumption trend of the data center in a future first time period, the power consumption trend comprising a plurality of power consumption peaks; analyzing an influence degree of performing a peak shaving function of the backup battery on a service at each power consumption peak in the plurality of power consumption peaks, the influence degree on the service comprising an influence degree on service performance and / or service revenue; and determining a peak shaving strategy for the first time period according to the influence degree on the service. Thus, when the power supply capacity of the data center cabinet is lower than the power supply demand, the influence on the service is effectively reduced.

Description

Technical Field

[0001] This application relates to the field of power supply technology, and more specifically, to a power supply control method and apparatus for a data center. Background Technology

[0002] With the rapid development of the internet, artificial intelligence, big data, and other fields, data centers are expanding in scale, and the number of power-consuming devices (such as servers, switches, and storage devices) is growing rapidly. To ensure power reliability, data centers need 24 / 7 uninterrupted power supply capability. Therefore, data centers are typically equipped with backup power devices such as large UPS systems, diesel generators, or distributed batteries. In the event of an abnormal AC power outage, backup power devices can provide short-term power while waiting for AC power to be restored. Among various backup power devices, distributed batteries are increasingly being adopted by data centers due to their lower cost and maintenance expenses.

[0003] In addition to backup power, distributed batteries can also be used for peak shaving to improve the power efficiency of data centers. Existing peak shaving control strategies regulate battery discharge by setting a discharge threshold. Specifically, when the power consumption of a rack deploying multiple power-intensive devices in a data center exceeds a set threshold, the distributed batteries begin discharging based on their charge level. Discharging stops if the batteries are low; conversely, charging begins when the rack power consumption falls below the threshold. However, if rack power consumption remains consistently high, the batteries quickly deplete. If even higher power consumption occurs, power limiting is necessary to prevent rack overload. However, power limiting restricts the operation of power-intensive devices within the rack, potentially leading to widespread service disruptions with significant impact.

[0004] Therefore, how to effectively reduce the impact on business operations when the power supply capacity of data center racks is lower than the power supply demand is an urgent problem to be solved. Summary of the Invention

[0005] This application provides a power supply control method and apparatus for data centers, which solves the problem that large-scale business disruptions can easily occur when the power supply capacity of data center cabinets is lower than the power supply demand.

[0006] In a first aspect, a power supply control method for a data center is provided, the data center including a backup battery, the method comprising: predicting the power consumption change trend of the data center in a future first time period, the power consumption change trend including multiple power consumption peaks; analyzing the impact of performing peak shaving function of the backup battery at each of the multiple power consumption peaks on services, the impact on services including the impact on service performance and / or service revenue; and determining a peak shaving strategy for the first time period based on the impact on services.

[0007] It should be understood that the above-mentioned prediction of the power consumption trend of the data center in the first time period is the same as the prediction of the power consumption trend of the data center racks in the first time period. The above-mentioned services are the services carried by the power-consuming equipment and loads in the data center racks, which can also be referred to as power-consuming services.

[0008] It should be understood that the peak value of each of the multiple power consumption peaks is greater than a preset threshold, for example, greater than a set threshold value.

[0009] It should be understood that determining the peak scaling strategy for the first time period involves determining whether to perform peak scaling at each power consumption peak within that first time period. Optionally, the peak scaling strategy includes performing battery peak scaling or power capping at each power consumption peak. That is, for peaks where peak scaling is not performed, power capping can be performed to limit the power consumption of the cabinet.

[0010] Optionally, the analysis of the impact on services can be an analysis of the impact on service performance, the impact on service revenue, or the impact on both. Service performance refers to the performance of power-consuming services within the server rack. By analyzing the impact on service performance, a peak-shaving strategy can be determined based on this impact, thus reducing the impact on service performance. Service revenue refers to the economic benefits generated by the power-consuming services within the server rack. Service revenue is related to the nature, type, and volume of services performed at each peak. By analyzing the impact on service revenue, a peak-shaving strategy can be determined based on this impact, thus reducing the impact on the economic benefits generated by the services.

[0011] Optionally, the first time period can be in units of minutes, hours, days, etc.

[0012] In this embodiment, the power consumption trend of the data center racks is first predicted within a future first time period. The impact of executing peak-shaving at each power consumption peak during this period on the services at that peak is analyzed. Then, based on the degree of impact on the services, the peak-shaving strategy for that first time period is determined. This avoids the problem of using existing simple threshold values ​​to trigger battery peak-shaving, which may cause operational limitations of power-consuming equipment and widespread service disruption when the power supply capacity of the data center racks is lower than the power demand.

[0013] In conjunction with the first aspect, in some implementations of the first aspect, the plurality of power consumption peaks include at least one first power consumption peak and at least one second power consumption peak. The peak shaving strategy for determining the first time period based on the degree of impact on services includes: preferentially determining to perform peak shaving at the at least one first power consumption peak, wherein the degree of impact of performing peak shaving at the at least one first power consumption peak on services is greater than the degree of impact of performing peak shaving at the at least one second power consumption peak on services.

[0014] In conjunction with the first aspect, in some implementations of the first aspect, the determination of the peak shaving strategy for the first time period based on the degree of impact on the business further includes: determining to perform peak shaving or power capping at the at least one second power consumption peak.

[0015] Optionally, peak clipping or power capping can be selectively applied to the second power consumption peak. For example, when the battery is sufficiently charged, peak clipping can be applied to the second power consumption peak, while when the battery is insufficient, power capping can be applied to the second power consumption peak.

[0016] It should be understood that limiting smaller power consumption peaks has a smaller impact on the scope and severity of service disruption; however, limiting larger power consumption peaks may cause widespread service disruption to a large number of users, resulting in a significant impact. Therefore, in this embodiment, peak shaving is prioritized for the first power consumption peak, which has a greater impact on service, while peak shaving or power capping is selectively applied to the second power consumption peak, which has a smaller impact on service, based on actual power consumption conditions. This effectively reduces the scope and severity of the impact on service while ensuring power supply safety.

[0017] In conjunction with the first aspect, in some implementations of the first aspect, the larger the peak value of the power consumption peak, the greater the impact of the peak shaving function of the backup battery on the business.

[0018] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: after predicting the power consumption change trend, and before determining the peak shaving strategy for the first time period based on the degree of impact on the business, analyzing the impact of performing peak shaving at each power consumption peak on energy-saving benefits, the impact on energy-saving benefits including the impact on the lifespan of the backup battery; the determination of the peak shaving strategy for the first time period based on the degree of impact on the business includes: determining the peak shaving strategy for the first time period based on the degree of impact on the business and the impact on energy-saving benefits.

[0019] Optionally, the impact on energy-saving benefits may also include the impact on resource utilization. For example, the impact on rack utilization, the density of power-consuming equipment deployed in the rack, and power supply efficiency, thereby maximizing energy-saving benefits.

[0020] It should be understood that the number of charge-discharge cycles within the lifespan of a backup battery is limited, and this number refers to the number of full charge-discharge cycles. Taking charging as an example, during use, when the amount of charge accumulated in one or more charges reaches a certain value, such as the battery capacity value, it is recorded as a full charge. Therefore, minimizing the number of peak-shaving cycles and minimizing the amount of battery discharge can extend the lifespan of the backup battery, allowing it to maximize its value within its lifespan and realize the energy-saving benefits of the battery.

[0021] In this embodiment, whether the battery is used for peak shaving considers both the impact on service and the impact on battery life. This extends the battery's lifespan while ensuring service continuity, maximizing its value and achieving energy savings. For example, when the battery is fully charged, peak shaving is performed on power consumption peaks that significantly impact service, while power consumption peaks with less impact on service can be capped, rather than performing peak shaving only when the battery has remaining charge. This extends battery life without disrupting service.

[0022] Optionally, the discharge current can be selected during discharge, i.e., fast or slow discharge, to prevent a decrease in battery performance.

[0023] In conjunction with the first aspect, in some implementations of the first aspect, the power consumption trend also includes at least one power consumption trough, and the method further includes: determining whether charging is performed at each power consumption trough in the at least one power consumption trough.

[0024] In conjunction with the first aspect, in some implementations of the first aspect, when determining to perform charging, the method further includes: determining the magnitude of the charging current.

[0025] In the embodiments of this application, charging can be performed at each power consumption trough based on the actual situation. During charging, the charging current can be selected, i.e., fast charging or slow charging can be selected, instead of charging when the battery has capacity. This can extend the battery's lifespan, prevent the battery's performance from deteriorating, and enable the battery to play its maximum value during its life cycle, thereby realizing the battery's energy-saving benefits.

[0026] Secondly, a power supply control device for a data center is provided, the data center including a backup battery, the device including: a controller, the controller being configured to: predict the power consumption change trend of the data center in a future first time period, the power consumption change trend including multiple power consumption peaks; analyze the impact of performing peak shaving function of the backup battery at each of the multiple power consumption peaks on services, the impact on services including the impact on service performance and / or service revenue; and determine the peak shaving strategy for the first time period based on the impact on services.

[0027] In conjunction with the second aspect, in some implementations of the second aspect, the plurality of power consumption peaks include at least one first power consumption peak and at least one second power consumption peak, and the controller is further configured to: preferentially determine to perform peak clipping at at least one first power consumption peak, wherein the impact of performing peak clipping at at least one first power consumption peak on the service is greater than the impact of performing peak clipping at at least one second power consumption peak on the service.

[0028] In conjunction with the second aspect, in some implementations of the second aspect, the controller is also used to: determine whether to perform peak clipping or power capping at at least one second power peak.

[0029] In conjunction with the second aspect, in some implementations of the second aspect, the larger the peak value of the power consumption peak, the greater the impact of the peak shaving function of the backup battery on the business.

[0030] In conjunction with the second aspect, in some implementations of the second aspect, the controller is further configured to: after predicting the power consumption change trend, and before determining the peak shaving strategy for the first time period based on the degree of impact on the business, analyze the impact of performing peak shaving at each power consumption peak on energy-saving benefits, the impact on energy-saving benefits including the impact on backup battery life; the determination of the peak shaving strategy for the first time period based on the degree of impact on the business includes: determining the peak shaving strategy for the first time period based on the degree of impact on the business and the impact on energy-saving benefits.

[0031] In conjunction with the second aspect, in some implementations of the second aspect, the power consumption trend also includes at least one power consumption trough, and the controller is further configured to: determine whether charging is performed at each power consumption trough in the at least one power consumption trough.

[0032] In conjunction with the second aspect, in some implementations of the second aspect, the controller is also used to: determine the magnitude of the charging current.

[0033] Thirdly, a control device is provided, comprising: a processor coupled to a memory; the memory for storing instructions; and the processor for executing the instructions stored in the memory to cause the device to perform the control method of the first aspect described above.

[0034] Fourthly, a power supply device is provided, comprising: a power supply management component, a distributed battery, a management board control component, an IT device, a power supply bus, a communication network, a power supply input, and a power supply output of the distributed battery, wherein the power supply input and the power supply output of the distributed battery are connected to the power supply bus, and the power supply management component, the management board control component, and the IT device communicate through the communication network to implement the method of the first aspect above.

[0035] Fifthly, a computer-readable medium is provided, including instructions that, when executed on a processor, cause the processor to perform the method described in the first aspect.

[0036] In a sixth aspect, a chip is provided, the chip including a processor and a data interface, the processor reading instructions stored in a memory through the data interface and executing the method in the first aspect above.

[0037] Optionally, as one implementation, the chip may further include a memory storing instructions, and the processor is used to execute the instructions stored in the memory. When the instructions are executed, the processor is used to perform the method in the first aspect described above.

[0038] The aforementioned chip can be a Field Programmable Gate Array (FPGA) or an Application-Specific Integrated Circuit (ASIC).

[0039] It should be understood that, in this application, the method of the first aspect can specifically refer to the first aspect and the method in any of the various implementations of the first aspect. Attached Figure Description

[0040] Figure 1 This is an example diagram of a power supply control system architecture for a data center provided in an embodiment of this application;

[0041] Figure 2 This is an exemplary block diagram of a control subsystem provided in an embodiment of this application;

[0042] Figure 3 This is an example diagram of a power supply control method for a data center provided in an embodiment of this application;

[0043] Figures 4-6 This application provides three scenario example diagrams in its embodiments;

[0044] Figure 7 This is an example diagram of a power supply control device for a data center provided in an embodiment of this application;

[0045] Figure 8 This is an exemplary block diagram of the hardware structure of a control device provided in an embodiment of this application. Detailed Implementation

[0046] To facilitate understanding, some technical terms involved in the various embodiments of this application will be introduced first.

[0047] Data center: A globally collaborative network of specific devices used to transmit, accelerate, display, compute, and store data information on the Internet network infrastructure.

[0048] An uninterruptible power supply (UPS) is a constant voltage and frequency power supply containing energy storage devices and with an inverter as its main component. It is primarily used to provide uninterrupted power to single computers, computer network systems, or other power electronic equipment.

[0049] Distributed batteries: These are small, modular, environmentally compatible, independent power sources with power ranging from several kilowatts to 50 MW. These power sources are owned by power companies, power users, or third parties to meet specific requirements of the power system and users.

[0050] Peak shaving and valley filling is a measure to adjust electricity load. Since power plants generate electricity continuously around the clock, if the generated electricity is not used, the energy used for power generation is wasted. A power plant's generating capacity is usually fixed and not easily changed, but peak electricity demand typically occurs during the day, resulting in insufficient electricity during the day and low demand at night, with excess generated electricity being wasted. To address this, the power system can shift some peak load to the off-peak period at night, thus utilizing the excess electricity at night and achieving energy conservation. In this application, energy storage devices (such as distributed batteries) are used to store excess electricity at night and use it during the day, thereby achieving peak shaving and valley filling. Specifically, during peak electricity demand periods during the day, the energy storage device is controlled to supply power to power-consuming equipment. Thus, during peak electricity demand periods, the addition of power from the energy storage device alleviates the pressure on electricity supply, achieving "peak shaving." During off-peak electricity demand periods at night, the energy storage device is controlled to charge, storing the excess electricity at night for use during the day, thereby achieving "valley filling."

[0051] Power consumption capping limits the overall power consumption of a server. For example, if a server's maximum power is 1000W, and the data center can provide a maximum power of 40000W, typically only 40 servers can be deployed. However, practical testing shows that a server's actual power consumption under its current configuration is generally less than 800W. Therefore, setting the server's maximum power consumption to 800W allows up to 50 servers to be deployed in the same data center, significantly improving power utilization.

[0052] To facilitate understanding, the background technology involved in the embodiments of this application will be described in detail.

[0053] With the rapid development of the Internet, artificial intelligence, big data, and other fields, data centers are constantly expanding in scale. The number of power-consuming devices (such as servers, switches, storage devices, and other IT equipment) in data center racks is growing rapidly, leading to an increase in the number of racks deploying these devices. To ensure uninterrupted operation of the power-consuming devices in the racks, traditional data centers are equipped with large centralized UPS systems, diesel generators, or distributed battery backup systems. These backup systems can provide short-term power supply in the event of an AC power outage, pending the restoration of AC power.

[0054] Among various backup power devices, centralized UPS and distributed battery systems are widely used. Centralized UPS systems treat the data center as a whole, using batteries for backup power. Centralized UPS systems typically require a dedicated room to house the battery packs, resulting in higher purchase and maintenance costs. Distributed batteries, on the other hand, are placed within server racks, making full use of rack space to provide backup power for one or more IT racks. Furthermore, distributed batteries are dispersed throughout the IT racks, leading to relatively lower purchase and maintenance costs and easier maintenance. Therefore, the proportion of data centers using distributed backup power has been increasing in recent years.

[0055] In practical applications, distributed batteries, besides serving as backup power, can also be used for peak shaving to improve the power efficiency of data centers. Specifically, because the power supply capacity of server racks in a data center has an upper limit, if the power consumption of power-consuming devices exceeds this limit, it may cause the circuit breaker in the rack to trip, resulting in the downtime of all power-consuming devices in the rack. Typically, to avoid circuit breaker tripping, data centers need to reduce the number of power-consuming devices or limit their power consumption to lower rack power consumption. Reducing the number of power-consuming devices and lowering deployment density leads to insufficient rack space utilization and wasted resources. Limiting the power consumption of power-consuming devices reduces their performance, wastes computing resources, and negatively impacts user experience and satisfaction for cloud computing services. When distributed batteries provide peak shaving functionality, they can effectively avoid these situations, improving the power efficiency of the data center while increasing the number of power-consuming devices in a single rack and the rack's utilization rate.

[0056] Current peak-shaving power supply control strategies primarily control battery discharge by setting discharge thresholds. This means that when rack power consumption exceeds a set threshold, the battery begins to discharge based on its charge level; if the battery is low, discharge stops; and if rack power consumption falls below the threshold, the battery begins to charge. However, this threshold-based battery discharge control method cannot identify service activity or the magnitude of power consumption peaks. If the rack's power consumption remains consistently high, the battery will quickly run out of power. If a higher service peak then occurs, the only solution is to limit the power consumption of the equipment to prevent rack overload and circuit breaker tripping. Typically, limiting small power peaks has a limited impact on the service scope and extent, while limiting large power peaks can lead to widespread service disruptions and significant impacts. Therefore, existing solutions have a very negative impact on cloud computing services.

[0057] In summary, existing peak-shaving power supply control strategies cannot identify business conditions and cannot solve the problem that may lead to severe damage to large-scale services when the power supply capacity of data center cabinets is lower than the power consumption requirements.

[0058] To address the problems existing in the prior art, this application provides a power supply control method for data centers. By analyzing the impact of peak shaving on services at each power consumption peak in the power consumption change trend over a future period, and determining the peak shaving strategy for that period based on the impact, the method can effectively reduce the impact on services when the power supply capacity of the data center cabinet is lower than the power supply demand.

[0059] To better understand the solutions of the embodiments of this application, before describing the methods, we will first refer to the appendix. Figure 1 and 2 The system architecture of the embodiments of this application will be briefly described.

[0060] Figure 1 This is an example diagram of a power supply control system architecture for a data center, provided in an embodiment of this application. This system architecture 100 can be applied to the power supply control of data center cabinets. Figure 1 As shown, the system architecture 100 includes a power management component 101, a distributed battery 102, a management board control component 103, IT equipment 104, a power supply bus 105, a communication network 106, a power supply input for the server rack 107, and a power supply output for the distributed battery 108. The IT equipment refers to the power-consuming devices in the data center. For ease of description, the power-consuming devices in the data center will be uniformly referred to as IT equipment in the following text, which may include servers, storage devices, switches, etc. It should be understood that the above components can also be called units, modules, parts, etc., and this application does not limit them in this way. For ease of description, they will be collectively referred to as components in the following text.

[0061] like Figure 1 As shown, both power input 107 and power output 108 of the distributed battery are connected to power bus 105, and IT device 104 draws power from power bus 105. It should be understood that IT device 104 may include one or more power inputs, and at least one of them is connected to distributed battery 102. Power input 107 also has at least one connection to distributed battery 102, providing power to the entire rack while simultaneously charging distributed battery 102.

[0062] For example, in practical applications, the mains power is primarily connected to the cabinet via power input 107. The mains power can be connected via a single or multiple lines. With a single line connection, the mains power is connected to the distributed battery 102, which then supplies power to the IT device 104. With dual mains power connections, both lines can be connected to the distributed battery 102, and then connected to the IT device 104 via the cabinet power bus 105. Alternatively, one line can be connected to the distributed battery 102, which is then connected to the IT device 104, while the other line is directly connected to the IT device 104. In general, at least one line of mains power is connected to the distributed battery 102, and at least one power input to the IT device 104 originates from the distributed battery 102.

[0063] The power supply management component 101, the management board control component 103, and the IT equipment 104 are connected via a communication network 106 to form a control subsystem 110. It should be understood that the above subsystem may also be referred to as a unit or module, etc., and this application does not limit it in this way; for ease of description, it will be collectively referred to as a subsystem below.

[0064] During system operation, the control subsystem 110 can query the data of battery 102 in real time, and can also interact with battery 102 through various communication modes such as network and serial port to obtain battery data and control battery charging and discharging. Simultaneously, the control subsystem 110 queries the data of IT device 104 in real time to obtain various data such as power consumption and temperature of IT device 104 for data analysis. The control subsystem 110 can also control IT device 104, for example, by controlling it to implement power consumption capping to limit cabinet power consumption and prevent tripping. It should be understood that the control subsystem may also include a controller, which can be a control chip, to control and implement the above functions. The following section will combine... Figure 2 The communication relationships of the control subsystem 110 are described in detail.

[0065] Figure 2 This is an exemplary block diagram of a control subsystem provided in an embodiment of this application. For example... Figure 2 As shown, the control subsystem 110 includes three parts: a power management component 101, a management board control component 103, and an IT device 104. These three parts are all connected to the communication network 106 and can communicate with each other. Under their joint action, they realize the power supply control function of the distributed battery 102.

[0066] The power management component 101 is used to collect data from the distributed batteries 102 and control their charging and discharging, and may include a battery management system (BMS). The management board control component 103 is used for energy-saving sensing, decision-making, control, and the collection of various parameters of the cabinet. The IT equipment is used for power consumption reporting and adjustment. It should be understood that the power management component 101, the management board control component 103, and the IT equipment 104 in the control subsystem 110 may include sensors, detection components, storage chips, hard drives, etc., to implement related functions, and may each include a controller (chip) or share a controller (chip) to work together to achieve the above functions.

[0067] Specifically, the management board control component 103 of the server rack communicates with the power management component 101 to collect the real-time parameter status of the distributed battery 102 and the real-time power of the server rack. For any abnormalities in the distributed battery 102, it immediately takes action, uses the collected data to formulate a power supply strategy, and implements the charging and discharging strategy of the distributed battery 102. The management board control component 103 also communicates with the IT device 104, collecting the real-time data status of the IT device 104 and determining if there are any abnormalities. It uses the collected power consumption data of the load on the IT device 104, along with the data from the distributed battery 102, to formulate a power supply strategy. The IT device 104 receives data acquisition and power consumption adjustment control commands from the management board control component 103 and executes the actions of the management board control component 103.

[0068] It should be understood that, in this embodiment of the application, the number of distributed batteries 102, management board control components 103, and IT devices 104 in system architecture 100 is not limited. And for ease of description, in this embodiment of the application, the distributed batteries are described as batteries.

[0069] Figure 3 This is an example diagram of a power supply control method for a data center provided in an embodiment of this application. It should be understood that the data center includes a backup battery, and this method 300 can be applied to system architecture 100. Method 300 includes steps S310-330, which are described in detail below.

[0070] S310 predicts the power consumption trend of a data center in the first time period in the future.

[0071] It should be understood that the first time period refers to a certain period of time in the future. Optionally, the first time period can be a few minutes, a few hours, a few days, etc. in the future.

[0072] It should be understood that the power consumption trend may include multiple power consumption peaks. It should also be understood that the peak value of each of the multiple power consumption peaks is greater than a preset threshold, for example, greater than a set threshold value.

[0073] Optionally, the predicted power consumption trend can be a series of data or a power consumption change curve. This application does not limit this; for ease of description, the power consumption trend in this application is exemplified by a curve. Optionally, the power consumption trend may or may not be presented to the user; this application does not limit this.

[0074] It should be understood that, since a data center has multiple server racks, each rack houses multiple IT devices and batteries. The power input of the batteries and racks is used to power the IT devices within the racks, and these IT devices generate power consumption during operation. Therefore, in this embodiment, the aforementioned prediction of the power consumption trend of the data center in the first future time period is equivalent to the prediction of the power consumption trend of the data center racks in the first future time period.

[0075] Optionally, before predicting the power consumption trend, the method further includes acquiring historical power consumption data, which represents the power consumption generated by the IT equipment in the rack over a historical period, referred to in this application as the rack's historical power consumption. It should be understood that this rack historical power consumption data can be collected in real time. After acquiring the historical power consumption data, the power consumption trend for the next time period can be predicted based on this data.

[0076] S320 analyzes the impact of performing peak shaving on services at each of the multiple power consumption peaks.

[0077] It should be understood that the above-mentioned services are those carried by power-consuming equipment and loads in data center cabinets, and can also be called power-consuming services, such as cloud computing services, data processing services, storage services, etc. This application does not limit them.

[0078] The degree of impact on business operations includes the impact on business performance and / or business revenue. That is, analyzing the impact on business operations can be done by analyzing the impact on business performance, the impact on business revenue, or the impact on both. Business performance refers to the performance of power-consuming services within the server rack, such as the computing performance of cloud computing services, the processing capacity of data processing services, and the storage capacity of storage services. This business performance can be expressed as numerical indicators; for example, computing performance can be expressed as the number of instructions processed per unit of time. By analyzing the degree of impact on business performance, peak shaving strategies are determined to reduce the impact on business performance. For example, the analysis process can be as follows: At a certain power consumption peak, if battery peak shaving is not performed, the number of instructions processed by the computing service per unit time is A; if battery peak shaving is performed, the number of instructions processed per unit time is B. If the absolute value of the difference between A and B is less than a preset threshold, it can be considered that the impact of peak shaving on service performance is small; if the absolute value of the difference between A and B is greater than or equal to the preset threshold, it can be considered that the impact of peak shaving on service performance is large. The specific preset threshold is determined according to the actual situation and is not limited here. Service revenue refers to the economic benefits brought by the power-consuming services in the cabinet itself. The service revenue is related to the nature, type, and volume of the services performed at each peak. For example, service revenue can be the revenue brought by cloud computing services, data processing services, storage services, etc. By analyzing the degree of impact on service revenue and determining the peak shaving strategy based on the degree of impact on service revenue, the impact on the economic benefits generated by the service can be reduced. It should be understood that the above method for judging and determining the degree of impact on service performance also applies to determining the degree of impact on service revenue, which will not be elaborated here.

[0079] Optionally, generally speaking, the larger the peak value of the power consumption spike, the greater the impact of peak shaving by the backup battery on services. A larger peak value indicates higher power consumption from power-consuming services, meaning these services are more numerous or widespread. Limiting the power consumption spike at that point could lead to widespread service disruptions for many users, significantly impacting user experience and satisfaction. Therefore, performing peak shaving at that point can meet the needs of normal service operation, with a significant impact on services. Conversely, a smaller peak value indicates lower power consumption from power-consuming services, meaning fewer or smaller power-consuming services are involved. Limiting the power consumption spike at that point has a smaller impact on the scope and extent of services affected. Therefore, performing peak shaving at that point has a smaller impact on services compared to limiting power consumption.

[0080] Alternatively, the degree of impact can also be determined based on the numerical range in which the peak value of the power consumption peak is located. In other words, power consumption peaks falling within the same numerical range can be considered to have the same degree of impact on the service. The division of the range can be determined according to the actual situation, and no limitation is made here.

[0081] Optionally, the degree of impact can also be determined based on the specific service information carried at each power consumption peak. In short, the degree of impact on services can be determined from multiple aspects, and this application does not limit this.

[0082] For ease of description, when analyzing the impact of performing the peak-shaving function of the backup battery on services at each of the multiple power consumption peaks, it is understood as analyzing the peak value of the power consumption peak.

[0083] S330 determines the peak shaving strategy for the first time period based on the degree of impact on business.

[0084] Optionally, the above peak clipping strategy includes determining whether to perform peak clipping at each power consumption peak.

[0085] Optionally, the above peak-shaving strategy can be to selectively perform battery peak-shaving or power capping at each power consumption peak based on the degree of impact on business. That is, power capping can be performed on power consumption peaks where peak-shaving is not performed to limit the power consumption of the cabinet.

[0086] In this embodiment, the peak clipping function is determined at each power consumption peak based on the degree of impact on the business, instead of using the existing simple threshold value to trigger battery peak clipping. This effectively reduces the impact on the business when the power supply capacity of the data center rack is lower than the power supply demand.

[0087] Optionally, the multiple power consumption peaks include at least one first power consumption peak and at least one second power consumption peak. Determining the peak reduction strategy for the first time period based on the degree of impact on the service includes: preferentially determining to perform peak reduction at at least one first power consumption peak, wherein the impact of performing peak reduction at at least one first power consumption peak on the service is greater than the impact of performing peak reduction at at least one second power consumption peak on the service.

[0088] Optionally, determining the peak shaving strategy for the first time period based on the degree of impact on business also includes: determining to perform peak shaving or power capping at at least one second power peak.

[0089] It should be understood that the impact of performing the peak-shaving function at the first power consumption peak on the service is greater than the impact of performing the peak-shaving function at the second power consumption peak. In other words, the power consumption peak of the first power consumption peak is greater than the power consumption peak of the second power consumption peak. It should also be understood that limiting smaller power consumption peaks results in a smaller scope and degree of service disruption; while limiting larger power consumption peaks may lead to widespread service disruption for a large number of users, with a significant impact.

[0090] In this application, when high power consumption is predicted to occur in the near future, the power supply control method prioritizes using battery peak clipping for higher power consumption peaks. For other lower power consumption peaks, peak clipping or power capping can be selectively performed based on the location and timing of the power consumption peaks, as well as battery level and charging / discharging opportunities. This effectively reduces the scope and severity of the impact on services while ensuring power supply safety.

[0091] Optionally, the method 300 further includes obtaining the power level of the backup battery to determine a peak shaving strategy for the first time period based on the battery's power level and its impact on services. For example, when the battery is sufficiently charged, it can be determined that peak shaving will be performed on both the first and second power consumption peaks. When the battery is insufficient, it can be selected that peak shaving will be performed on the first power consumption peak, while power consumption will be capped on the second power consumption peak.

[0092] Optionally, method 300 further includes: after predicting the power consumption change trend, and before determining the peak shaving strategy for the first time period based on the degree of impact on the business, analyzing the impact of performing peak shaving at each power consumption peak on energy saving benefits, the impact on energy saving benefits including the impact on backup battery life; determining the peak shaving strategy for the first time period based on the degree of impact on the business includes: determining the peak shaving strategy for the first time period based on the degree of impact on the business and the impact on energy saving benefits.

[0093] It should be understood that the number of charge-discharge cycles within the lifespan of a backup battery is limited, and this number refers to the number of full charge-discharge cycles. Taking charging as an example, during use, when the amount of charge accumulated in one or more charges reaches a certain value, such as the battery capacity value, it is recorded as a full charge. Therefore, minimizing the number of peak-shaving cycles and minimizing the amount of battery discharge can extend the lifespan of the backup battery, allowing it to maximize its value within its lifespan and realize the energy-saving benefits of the battery.

[0094] Therefore, optionally, based on the degree of impact on business and the impact on energy-saving benefits, the peak shaving strategy for the first time period can be determined as follows: if performing peak shaving at a certain power consumption peak has a small impact on business, but not performing peak shaving can bring certain energy-saving benefits, then even if the battery has sufficient power, power capping can be performed at that power consumption peak to realize the energy-saving benefits of the battery.

[0095] It should be understood that, generally speaking, the smaller the peak value of the power consumption peak, the less impact it has on the service. Therefore, in practice, it can be set that the peak value of the power consumption peak is less than a predetermined threshold, and the peak value will not be executed.

[0096] Optionally, the plurality of power consumption peaks includes a third power consumption peak, and determining the peak shaving strategy for the first time period based on the degree of impact on services and the impact on energy-saving benefits includes: if the peak value of the third power consumption peak is less than or equal to a preset threshold, determining to perform power consumption capping at the third power consumption peak.

[0097] In this embodiment, whether the battery is used for peak shaving considers both the impact on service and the impact on battery life. This extends the battery's lifespan while ensuring service continuity, maximizing its value and realizing energy savings. For example, when the battery is fully charged, peak shaving is performed on power consumption peaks that significantly impact service, while power consumption peaks with less impact on service are capped, rather than performing peak shaving whenever the battery has remaining charge. This extends battery life while ensuring service continuity and further guarantees energy savings.

[0098] Optionally, the impact on energy-saving benefits may also include the impact on resource utilization. For example, the impact on rack utilization, the density of power-consuming equipment deployed in the rack, and power supply efficiency, thereby maximizing energy-saving benefits.

[0099] Optionally, the discharge current can be selected during discharge, i.e., fast or slow discharge, to prevent a decrease in battery performance.

[0100] Optionally, the power consumption trend may also include at least one power consumption trough, and the method 300 may further include: determining whether charging is performed at each power consumption trough in the at least one power consumption trough.

[0101] Optionally, when determining to perform charging, method 300 may further include: determining the magnitude of the charging current.

[0102] In the embodiments of this application, charging can be performed at each power consumption trough based on the actual situation. During charging, the charging current can be selected, i.e., fast charging or slow charging can be selected, instead of charging when the battery has capacity. This can extend the battery's lifespan, prevent the battery's performance from deteriorating, and enable the battery to play its maximum value during its life cycle, thereby realizing the battery's energy-saving benefits.

[0103] For example, Figures 4-6 These are example diagrams illustrating three scenarios provided in the embodiments of this application. The following is in conjunction with… Figures 4-6 The power supply control method in the embodiments of this application is illustrated by example.

[0104] like Figure 4 As shown in Scenario 1, in the future, the power consumption at points 1 and 2 both exceed the threshold, but the peak value at point 1 is smaller than that at point 2, and there is no valley between points 1 and 2, meaning the battery has no opportunity to charge. Point 2, with its larger peak value, needs sufficient power for peak shaving to avoid reaching a power cap. In this case, a decision needs to be made based on the battery's current charge level.

[0105] For example, if the battery has sufficient power to clip the peak power consumption at points 1 and 2, then the battery peak clipping function will be triggered at points 1 and 2 according to the threshold values ​​respectively. It should be understood that triggering peak clipping according to the threshold value means that when the power consumption of the service exceeds the threshold value, peak clipping is performed.

[0106] For example, if the battery power is insufficient, meaning the battery power is not enough to provide peak shaving for both points 1 and 2 simultaneously, then power consumption capping is applied to point 1 to reduce the power consumption of the rack. In other words, the power consumption of point 1 is forcibly limited to a threshold value, thereby reserving power for peak shaving at point 2.

[0107] like Figure 5 As shown, in scenario two, the power consumption at points 1 and 3 will exceed the threshold value in the future. The peak value at point 1 is less than that at point 3. There are two points between points 1 and 3. The power consumption at point 2 is significantly lower than the threshold value, which means that the battery power can be replenished at point 2.

[0108] In this scenario, decisions also need to be made by considering multiple factors, such as battery power, battery capacity, the amount of electricity consumed in peak shaving at points 1 and 3, and the amount of electricity supplemented when filling the valley at point 2.

[0109] For example, if after peak clipping of power consumption at point 1, there is enough remaining battery capacity to clip peak consumption at point 3 through charging at point 2, then the battery peak clipping function is triggered at points 1 and 3 according to the threshold values, and the battery charging function is triggered at point 2.

[0110] For example, if peak scaling is performed at point 1, and power is replenished at point 2, but there isn't enough power to scale up the power at point 3, a decision can be made based on the actual situation regarding whether to implement power capping at point 1. Specifically, if the lack of sufficient power at point 3 for peak scaling is due to power consumption during peak scaling at point 1, then power capping can be applied to point 1 to reduce the rack's power consumption, essentially forcing the power consumption at point 1 to be limited to a threshold value, thus reserving power for peak scaling at point 3. However, if the lack of sufficient power at point 3 for peak scaling is not due to peak scaling at point 1, but rather due to various factors such as battery charge, battery capacity, the power consumed during peak scaling at points 1 and 3, and the power replenished during valley filling at point 2, for example, if the battery is currently fully charged, a fully charged battery itself cannot guarantee sufficient power for peak scaling at point 3. Furthermore, the amount of electricity consumed by peak shaving at point 1 is less than or equal to the amount of electricity replenished during valley filling at point 2. This means that after peak shaving at point 1 and valley filling at point 2, the battery remains fully charged. Therefore, peak shaving at point 1 has no impact on point 3. In this case, the battery peak shaving function can be triggered at points 1 and 3 according to threshold values, while the battery charging function can be triggered at point 2.

[0111] For example, if the battery has enough power to clip the peak power consumption at points 1 and 3 even without charging at point 2, the battery peak clipping function can be triggered at points 1 and 3 according to the threshold values ​​respectively. At point 2, charging is optional.

[0112] It should be understood that the above examples are not intended to limit this application. In practice, decisions should be made based on specific circumstances to minimize the impact on business and ensure maximum benefits.

[0113] like Figure 6 As shown in Scenario 3, in the future, the power consumption at points 1 and 2 will both exceed the threshold, but point 2, with higher power consumption, has higher priority, meaning it comes first in terms of time. In this case, the battery power will be prioritized for peak shaving, which can be triggered by the threshold. Whether point 2 will be shaving depends on the battery power level.

[0114] It should be understood that the above three scenarios are merely examples and are not intended to limit this application. In many actual application scenarios, the methods described in the embodiments of this application are generally followed to prioritize the use of battery peak clipping for higher power consumption peaks, thereby reducing the impact on the cabinet's services.

[0115] In summary, in this embodiment, on the one hand, when the power supply capacity of the server rack is limited, battery peak shaving can improve the power supply capacity of the rack, increase equipment density, and improve power utilization and space utilization, thereby saving data center resources. On the other hand, by using historical power consumption data of batteries and IT equipment, the power consumption trend of the rack over a future period can be predicted. Based on the future power consumption trend and battery power changes, the impact on business and energy-saving benefits are analyzed, and decisions are made on whether to activate battery peak shaving for each power consumption peak and whether to initiate battery charging, fast charging or slow charging, for each valley. This is not a simple case of using a threshold-triggered battery peak shaving technique. Therefore, it is possible to avoid the problem of business disruption when the power supply capacity of the data center rack is lower than the power consumption requirements of the IT equipment, achieving better energy saving and benefits.

[0116] Figure 7 This is an example diagram of a power supply control device for a data center provided in an embodiment of this application. The data center includes a backup battery. Figure 7 As shown, the control device 700 includes a controller 710.

[0117] Specifically, the controller 710 is used to predict the power consumption trend of the data center in the first time period in the future, the power consumption trend including multiple power consumption peaks; analyze the impact of performing the peak shaving function of the backup battery at each of the multiple power consumption peaks on the business, the impact on the business including the impact on business performance and / or business revenue; and determine the peak shaving strategy for the first time period based on the impact on the business.

[0118] Optionally, the multiple power consumption peaks include at least one first power consumption peak and at least one second power consumption peak. The controller 710 can also be used to: preferentially determine to perform peak clipping at at least one first power consumption peak, wherein the impact of performing peak clipping at at least one first power consumption peak on the service is greater than the impact of performing peak clipping at at least one second power consumption peak on the service.

[0119] Optionally, the controller 710 can also be used to: determine whether to perform peak clipping or power capping at at least one second power peak.

[0120] Optionally, the larger the peak value of the power consumption peak, the greater the impact of the peak shaving function of the backup battery on the business.

[0121] Optionally, the controller 710 can also be used to: after predicting the power consumption change trend, and before determining the peak shaving strategy for the first time period based on the degree of impact on the service, analyze the impact of performing peak shaving at each power consumption peak on energy saving benefits, the impact on energy saving benefits including the impact on backup battery life; determining the peak shaving strategy for the first time period based on the degree of impact on the service includes: determining the peak shaving strategy for the first time period based on the degree of impact on the service and the impact on energy saving benefits.

[0122] Optionally, the power consumption trend also includes at least one power consumption trough, and the controller 710 can also be used to: determine whether charging is performed at each power consumption trough in the at least one power consumption trough.

[0123] Optionally, when determining to perform charging, the controller 710 can also be used to: determine the magnitude of the charging current.

[0124] Optionally, the controller 710 can also be used to obtain the power level of the backup battery in order to determine a peak shaving strategy based on the battery power level and the degree of impact on the business.

[0125] Optionally, the controller 710 can also be used to acquire the historical power consumption of the IT device in order to predict the power consumption for a future first time period based on the historical power consumption.

[0126] Figure 8 This is an exemplary block diagram of the hardware structure of a control device provided in an embodiment of this application. The device 800 (specifically, it can be a computer device) includes a memory 810, a processor 820, a communication interface 830, and a bus 840. The memory 810, processor 820, and communication interface 830 are interconnected via the bus 840.

[0127] The memory 810 may be a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 810 may store a program, and when the program stored in the memory 810 is executed by the processor 820, the processor 820 is used to execute the various steps of the planning method of the embodiments of this application.

[0128] The processor 820 may be a general-purpose central processing unit (CPU), microprocessor, application-specific integrated circuit (ASIC), graphics processing unit (GPU), or one or more integrated circuits, used to execute relevant programs to implement the control method of the embodiments of this application.

[0129] The processor 820 can also be an integrated circuit chip with signal processing capabilities. In implementation, each step of the control method of this application can be accomplished through integrated logic circuits in the processor 820 or through software instructions.

[0130] The processor 820 described above can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory 810. The processor 820 reads the information in memory 810 and, in conjunction with its hardware, completes the functions required by the modules included in the planning device of the embodiments of this application, or executes the control method of the method embodiments of this application.

[0131] The communication interface 830 uses transceiver devices, such as, but not limited to, transceivers, to enable communication between the device 800 and other devices or communication networks.

[0132] Bus 840 may include a pathway for transmitting information between various components of device 800 (e.g., memory 810, processor 820, communication interface 830).

[0133] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0134] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0135] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0136] 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 network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0137] In addition, the functional units in the various embodiments of this application 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.

[0138] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion 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 this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0139] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A power supply control method of a data center, characterized by, The data center comprises a backup battery, and the method comprises: predicting a power consumption change trend of the data center in a first time period in the future, the power consumption change trend comprising a plurality of power consumption peaks; analyzing an influence degree on business of performing a peak shaving function of the backup battery at each power consumption peak in the plurality of power consumption peaks, the influence degree on business comprising an influence degree on business performance and / or business revenue; determining a peak shaving strategy for the first time period according to the influence degree on business, the peak shaving strategy comprising whether to perform a peak shaving function at each power consumption peak in the plurality of power consumption peaks.

2. The method of claim 1, wherein, The plurality of power consumption peaks comprises at least one first power consumption peak and at least one second power consumption peak, and the determining of the peak shaving strategy for the first time period according to the influence degree on business comprises: preferentially determining to perform peak shaving at the at least one first power consumption peak, wherein the influence degree on business of performing peak shaving at the at least one first power consumption peak is greater than the influence degree on business of performing peak shaving at the at least one second power consumption peak.

3. The method of claim 2, wherein, The determining of the peak shaving strategy for the first time period according to the influence degree on business further comprises: determining to perform peak shaving or power consumption capping at the at least one second power consumption peak.

4. The method of claim 1, wherein, The greater the peak value of the power consumption peak, the greater the influence degree on business of performing the peak shaving function of the backup battery.

5. The method of claim 1, wherein, The method further comprises: after predicting the power consumption change trend, before determining the peak shaving strategy for the first time period according to the influence degree on business, analyzing an influence on energy saving revenue of performing the peak shaving function at each power consumption peak, the influence on energy saving revenue comprising an influence on the service life of the backup battery; The determining of the peak shaving strategy for the first time period according to the influence degree on business comprises: determining the peak shaving strategy for the first time period according to the influence degree on business and the influence on energy saving revenue.

6. The method according to any one of claims 1-5, characterized in that, The power consumption change trend further comprises at least one power consumption valley, and the method further comprises: determining whether to perform charging at each power consumption valley in the at least one power consumption valley.

7. The method of claim 6, wherein, When it is determined to perform charging, the method further comprises: determining a size of a charging current.

8. A power supply control device of a data center, characterized by, The data center comprises a backup battery, and the apparatus comprises: a controller configured to: predict a power consumption change trend of the data center in a first time period in the future, the power consumption change trend comprising a plurality of power consumption peaks; analyze an influence degree on business of performing a peak shaving function of the backup battery at each power consumption peak in the plurality of power consumption peaks, the influence degree on business comprising an influence degree on business performance and / or business revenue; determine a peak shaving strategy for the first time period according to the influence degree on business, the peak shaving strategy comprising whether to perform a peak shaving function at each power consumption peak in the plurality of power consumption peaks.

9. The apparatus of claim 8, wherein, The plurality of power consumption peaks comprises at least one first power consumption peak and at least one second power consumption peak, and the controller is further configured to: The priority is determined to perform peak clipping at the at least one first power consumption peak, wherein an impact on services of performing peak clipping at the at least one first power consumption peak is greater than an impact on services of performing peak clipping at the at least one second power consumption peak.

10. The apparatus of claim 9, wherein, The controller is further configured to: determine to perform peak clipping or power consumption capping at the at least one second power consumption peak.

11. The apparatus of claim 8, wherein, The greater the peak value of the power consumption peak, the greater the impact on services of performing the peak clipping function of the backup battery.

12. The apparatus of claim 8, wherein, The controller is further configured to: after predicting the power consumption trend, determine a peak clipping strategy for the first time period according to the impact on services before analyzing an impact on energy saving benefits of performing the peak clipping function at each power consumption peak, the impact on energy saving benefits including an impact on the backup battery life; the determination of the peak clipping strategy for the first time period according to the impact on services comprises: determining the peak clipping strategy for the first time period according to the impact on services and the impact on energy saving benefits.

13. The apparatus of any one of claims 8-12, wherein, The power consumption trend further includes at least one power consumption valley, and the controller is further configured to: determine whether to perform charging at each power consumption valley in the at least one power consumption valley.

14. The apparatus of claim 13, wherein, When it is determined to perform charging, the controller is further configured to determine a size of a charging current.

15. A control device characterized by comprising: comprising: a processor coupled with a memory; the memory is configured to store instructions; the processor is configured to execute the instructions stored in the memory to cause the apparatus to perform the method of any one of claims 1-7.

16. A power supply device, comprising: comprising: a power supply management component, a distributed battery, a management board control component, IT equipment, a power supply bus, a communication network, a power supply input, and a power supply output of the distributed battery, wherein the power supply input and the power supply output of the distributed battery are connected with the power supply bus, the power supply management component, the management board control component, and the IT equipment communicate through the communication network to implement the method of any one of claims 1-7.

17. A computer readable medium characterized by instructions that, when executed on a processor, cause the processor to perform the method of any one of claims 1-7.

Images