CN-121529946-B - Centralized standby electricity energy storage optimal configuration method of data center based on cluster sharing architecture

CN121529946BCN 121529946 BCN121529946 BCN 121529946BCN-121529946-B

Abstract

The invention discloses a centralized standby power energy storage optimal configuration method of a data center based on a cluster sharing architecture. The method comprises the steps of collecting historical operation data, constructing a reliability evaluation module, constructing a centralized standby electricity energy storage configuration optimization model according to the reliability evaluation module, taking energy capacity and power capacity as energy storage configuration parameters, taking reliability constraint and energy storage configuration constraint as constraint conditions, taking minimized energy storage cost as an objective function, taking a combination function of a reliability punishment function and the objective function as an adaptability function, and solving the centralized standby electricity energy storage configuration optimization model to obtain optimal energy storage configuration parameters. The invention can solve the problem of the shared centralized standby power energy storage configuration of the data center cluster under the reliability constraint.

Inventors

YE CHENGJIN
LI BINGLE
DING YI
TANG XUEYONG
BAO MINGLEI

Assignees

浙江大学

Dates

Publication Date: 20260508
Application Date: 20260113

Claims (10)

1. The centralized standby power energy storage optimizing configuration method of the data center based on the cluster sharing architecture is characterized by comprising the following steps of: Step S1, collecting historical operation data or planning operation data of a target data center cluster system; S2, constructing a fault management module; Step S3, a reliability evaluation module is constructed according to the fault management module, wherein the reliability evaluation module is used for performing fault time sequence simulation, and when faults occur, the fault management module is used for determining the power supply paths and the power supply power of each data center, and finally the reliability value of each data center is obtained; The reliability is a combination of one or more power distribution network reliability indexes; S4, constructing a centralized standby energy storage configuration optimization model according to a reliability evaluation module, wherein the centralized standby energy storage configuration optimization model takes the energy capacity and the power capacity of a centralized standby energy storage power station as energy storage configuration parameters, takes reliability constraint and energy storage configuration constraint as constraint conditions, and takes the minimum energy storage cost as an objective function; The reliability constraint is expressed as: r i (E SESS ,C SESS )≤r i req r∈R Wherein R is a reliability index set, R epsilon R is a specific reliability index, R i (E SESS ,C SESS ) is a reliability value under corresponding energy storage configuration, and R i req is a preset reliability requirement value of the data center i; the reliability penalty function P (r) is expressed by the following formula: wherein R epsilon R is a specific reliability index, R req is a preset reliability requirement value of a data center i, and lambda is a penalty coefficient; And S5, solving the centralized standby power energy storage configuration optimization model through the fitness function according to the historical operation data to obtain optimal energy storage configuration parameters.
2. The data center centralized standby energy storage optimal configuration method based on the cluster sharing architecture is characterized in that the target data center cluster system comprises a centralized standby energy storage power station, at least one power supply and at least one data center, the target data center cluster system adopts the cluster sharing architecture, the centralized standby energy storage power station comprises an energy storage battery and an energy storage bus, each data center comprises a power supply and distribution system, a machine room and a direct current bus, the power supply and distribution system of each data center supplies power for the machine room by adopting direct current, the direct current buses of each data center are connected with the energy storage bus of the centralized standby energy storage power station through direct current cables, and the energy storage bus of the centralized standby energy storage power station is connected with the energy storage battery through the direct current cables.
3. The method for optimizing configuration of centralized standby energy storage of a data center based on a cluster sharing architecture as set forth in claim 1, wherein the fault management module comprises a power supply path confirmation unit and a multi-main power distribution unit; the power supply path confirming unit is used for acquiring power supply paths of the data centers and is configured to confirm the power supply paths of the data centers according to a method of firstly utilizing a power supply to supply power, secondly utilizing a centralized energy storage to supply power and finally utilizing the centralized energy storage to prepare electric energy; The multi-main-body power distribution unit is used for obtaining power supply power of the data centers and is configured to judge that the output is sufficient if the total maximum power supply capacity is larger than or equal to the total power consumption requirement of the data centers for a target data center cluster system or an electric isolated network, judge that the power supply power of each data center is equal to the power consumption requirement if the total maximum power supply capacity is smaller than the total power consumption requirement of the data centers, judge that the output is insufficient and divide the total maximum power supply capacity according to the power consumption requirement ratio of each data center to obtain the power supply power of each data center.
4. The method for centralized standby energy storage optimization configuration of a data center based on a cluster sharing architecture as set forth in claim 1, wherein the energy storage configuration constraints include an upper limit and a lower limit of energy capacity, an upper limit and a lower limit of power capacity, and an upper limit and a lower limit of a ratio of energy capacity to power capacity.
5. The method for centralized standby power energy storage optimization configuration of the data center based on the cluster sharing architecture of claim 1, wherein in the step S5, iterative solution is carried out by utilizing a heuristic algorithm, fault time sequence simulation is carried out on each group of energy storage configuration parameters by utilizing the reliability evaluation module in the iterative solution process, so as to obtain the reliability value of each data center, and the reliability value of each data center and the energy storage cost corresponding to the energy storage configuration parameters are input into a fitness function so as to obtain the fitness value.
6. The method for centralized standby power energy storage optimization configuration of the data centers based on the cluster sharing architecture of claim 5, wherein M times of fault time sequence simulation are conducted by utilizing the reliability evaluation module, and M original reliability values of each data center are averaged to obtain a reliability value; The fault time sequence simulation comprises the steps of randomly generating a system composite fault time sequence by using a sequential Monte Carlo method according to the fault rate and the repair rate of each element, and then carrying out fault time sequence simulation by combining a typical power load curve of each data center and a typical output curve of each power supply under the centralized standby power energy storage configuration corresponding to the energy storage configuration parameters according to the system composite fault time sequence, and obtaining the original reliability value of each data center according to time sequence simulation data.
7. The method for optimizing configuration of centralized standby power storage of data centers based on a cluster sharing architecture as set forth in claim 5, wherein the method is characterized in that at each simulation time step of fault timing simulation, if a fault exists in an element or the standby capacity of the centralized standby power storage power station is smaller than a preset standby capacity threshold, a fault management module is used, a power supply path and power supply power of each data center in a target data center cluster system are obtained through a two-stage calculation method, and the system is operated to the next simulation time step according to the power supply path and the power supply power, otherwise, the system keeps a normal power supply path and the power supply power and is operated to the next simulation time step.
8. The method for optimizing configuration of centralized standby power storage of a data center based on a cluster sharing architecture as set forth in claim 7, wherein the two-stage computing method is specifically: The first stage comprises taking the minimum electricity load of each data center as electricity demand, using a power supply path confirming unit to obtain temporary power supply paths and power supply power of each data center according to the electricity demand of each data center and the power supply capacity of each power supply, judging whether the power supply power of each data center can meet the electricity demand, if the electricity demand of the data center is met, taking the normal electricity load of the data center as new electricity demand, otherwise, taking the data center as power failure; and in the second stage, a power supply path confirming unit is used for obtaining the power supply path and the power supply of each data center according to the new power consumption requirement of each data center and the power supply capacity of each power supply.
9. The method for optimizing configuration of centralized standby power storage of a data center based on a cluster sharing architecture as set forth in claim 8, wherein the power supply path confirmation unit obtains the power supply path and the power supply power of each data center by: Firstly, taking a power supply as a power supply source, supplying power to each data center according to a first power supply path, judging whether an electric isolated network exists in a target data center cluster system of a centralized standby power energy storage facility or not by using a connectivity algorithm, and calculating first power supply power of each data center by using a multi-main-body power distribution unit according to power consumption requirements of each data center aiming at the target data center cluster system or each electric isolated network to obtain residual power consumption requirements of each data center; Secondly, taking a power supply and a centralized standby power energy storage power station as power supply sources, respectively supplying power to the data centers which do not meet the power consumption requirement according to a second power supply path and a third power supply path, calculating the second power supply of each data center which does not meet the power consumption requirement according to the residual power consumption requirement of each data center by using the multi-main-body power distribution unit, and calculating the cut load of each data center; And finally, taking the power supply as a power supply, and charging the centralized standby energy storage power station according to a fourth power supply path to obtain the charging power of the energy storage battery.
10. The data center centralized standby energy storage optimal configuration method based on the cluster sharing architecture, which is disclosed in claim 9, is characterized in that the first power supply path is a power supply-data center power supply and distribution system-machine room, the second power supply path is a power supply-redundant power supply and distribution system-energy storage bus-machine room, the third power supply path is an energy storage battery-energy storage bus-machine room, and the fourth power supply path is a power supply-redundant power supply and distribution system-energy storage bus-energy storage battery.

Description

Centralized standby electricity energy storage optimal configuration method of data center based on cluster sharing architecture Technical Field The invention belongs to the field of optimization of power supply and distribution systems of a data center and the field of planning and optimization of power systems, and particularly relates to a centralized standby energy storage optimization configuration method of a data center based on a cluster sharing architecture. Background The data center is taken as a basic stone of digital economy in China and is an important infrastructure for supporting the vigorous development of new generation digital technologies such as cloud computing, artificial intelligence and the like. Under the promotion of AI wave, the intelligent computing power of China will show a rapid increasing trend, and the intelligent computing center bearing the intelligent computing power shows the remarkable characteristics of high computing power and high power density. In this context, data centers need to continuously increase UPS or backup energy storage capacity to ensure their power reliability. However, the probability of directly generating power failure of the accessed commercial power is relatively low, so that the utilization rate of the uninterruptible power supply and the energy storage of the reliability configuration is greatly reduced, and the contradiction between the energy storage resource and the reliability utility is formed. In order to pursue lower network delay and marginal cost of infrastructure construction, future data centers will exhibit a new trend of clustering development, which puts higher demands on the construction of the power distribution network of the data center clusters. In recent years, with the rising of a novel energy storage mode of shared energy storage, a new thought is provided for solving the contradiction between energy storage resources and reliability utility for a data center. In the data center cluster, the centralized standby power storage can be configured based on the sharing mode, and electric energy support is provided for the operation of the data center under the fault condition, so that the utilization rate of the standby power storage is effectively improved. However, at present, a shared energy storage configuration method which simultaneously ensures reliability, low cost and efficiently utilizes energy storage resources is not available. Disclosure of Invention The invention provides a data center centralized standby electricity energy storage optimization configuration method based on a cluster sharing architecture for solving the problem of data center centralized standby electricity energy storage under reliability constraint. The invention firstly provides a configuration form of the standby power storage of the data center cluster, then constructs a fault management module for the condition of power supply priority deficiency and existence of a plurality of power supply paths, realizes quantitative evaluation of the power supply and distribution reliability of the data center by using a sequential Monte Carlo method, finally constructs a centralized standby power storage configuration optimization model, improves an objective function, solves optimal capacity configuration by using a heuristic algorithm, provides an optimization configuration method for the standby power storage configuration system of the data center cluster under the reliability constraint, and is suitable for guaranteeing the power supply reliability of the data center and efficiently utilizing energy storage resources under the condition of fast increase of calculation power requirements. The technical scheme adopted by the invention is as follows: 1. centralized standby power energy storage optimal configuration method for data center based on cluster sharing architecture The centralized standby electricity energy storage optimizing configuration method of the data center comprises the following steps: Step S1, determining a data center cluster based on a shared architecture and a design scheme of a centralized standby power energy storage power distribution network, and collecting historical operation data or planning operation data of a target data center cluster system. The target data center cluster system comprises a centralized standby power energy storage power station, at least one power supply and at least one data center, wherein the target data center cluster system adopts a cluster sharing architecture, each data center comprises a power supply and distribution system, a machine room and a direct current bus, the power supply and distribution system of each data center adopts direct current to supply power for the machine room, the direct current buses of each data center are connected with the energy storage buses of the centralized standby power energy storage power station through direct current cables, and the energy storage buses of t