Optimal design and scheduling method of multi-energy complementary distributed energy system for data center

Abstract

The application provides a data center multi-energy complementary distributed energy system and an optimization design and scheduling method thereof, and the method comprises the following steps: obtaining data required for energy system design and operation optimization; constructing a multi-energy complementary distributed energy system mathematical model; establishing a multi-energy complementary distributed energy system life cycle objective function; formulating a data center multi-energy complementary distributed energy system operation strategy; determining constraint conditions according to the energy balance principle and actual operation; establishing a multi-objective optimization model according to the equipment model, the operation strategy, the constraint conditions and the objective function; improving the traditional multi-objective locust optimization algorithm based on the idea of population mutation, and solving the model to obtain a non-inferior solution set; obtaining an optimal solution in the non-inferior solution set by using the optimal and inferior solution distance method, and obtaining an optimal design and scheduling. The application can not only meet the energy demand of the data center, but also ensure the economic, environmental and energy benefits in the system life cycle.

Description

Technical Field

[0001] This invention belongs to the field of energy system technology, specifically relating to an optimization design and scheduling method for a multi-energy complementary distributed energy system for data centers. Background Technology

[0002] Data centers, as the infrastructure for data processing and information management, play a vital role in modern society. In recent years, with the rapid development of big data, the Internet of Things, and artificial intelligence, the energy consumption and carbon emissions of data centers have been growing at an alarming rate. Statistics show that data centers currently account for 2% of global electricity consumption, and this growth rate is expected to reach 15%–20% in the coming years, with carbon emissions projected to reach 8% of global emissions by 2030. Besides technological upgrades to the data centers themselves, efficient, environmentally friendly, and economical energy supply systems are another effective way to address the energy consumption and carbon emission problems of data centers. Multi-energy complementary distributed energy systems (DMES) are energy supply systems deployed near the user's location, combining combined cooling, heating, and power (CCHP) systems, renewable energy systems, and energy storage systems to simultaneously provide users with cooling, heating, and electricity. The cleanliness, efficiency, and reliability of DMES perfectly meet the energy supply system requirements of data centers. The optimized design and operation of data center DMES will affect the overall system performance. However, currently, there are few methods for optimizing the design and scheduling of data center DMES, and methods for comprehensive optimization of data center DMES across its entire lifecycle, considering economic, environmental, and energy objectives, are also rare. Therefore, it is necessary to propose a method for multi-objective optimization design and scheduling of the entire lifecycle of data center DMES. Summary of the Invention

[0003] In order to overcome the problems existing in the prior art, the purpose of this invention is to provide an optimization design and scheduling method for a multi-energy complementary distributed energy system for data centers, so as to make up for the deficiencies in optimization objectives, operation strategies and solution methods, and to achieve clean and efficient energy supply for data centers.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] An optimization design and scheduling method for a multi-energy complementary distributed energy system in a data center, the optimization design and scheduling method comprising the following steps:

[0006] Step 1: Obtain the data required for the optimized design and operation of the energy station, mainly including local meteorological data such as temperature, irradiance, and wind speed, as well as the cold, heat, and electrical load data of the data center;

[0007] Step 2: Construct a multi-energy complementary distributed energy system and establish mathematical models for each device;

[0008] Step 3: Based on the mathematical model of the device, establish the life cycle objective function of the data center multi-energy complementary distributed energy system;

[0009] Step 4: Analyze the distribution characteristics of the data center's cold, heat, and electricity load data, and formulate an operation strategy suitable for the data center's multi-energy complementary distributed energy system;

[0010] Step 5: Determine the constraints based on the principle of energy balance and actual operation;

[0011] Step 6: Establish a multi-objective optimization model based on the equipment model, operating strategy, constraints, and objective function;

[0012] Step 7: Input the relevant data required for solving the multi-objective optimization model, including meteorological parameters, load data, and equipment parameters. Based on the idea of ​​population variation, improve the traditional multi-objective locust optimization algorithm, and solve the established multi-objective optimization model to obtain the optimal solution set.

[0013] Step 8: Select the optimal solution from the set of non-dominated solutions using the superior-inferior solution distance method, and then determine the optimal design and scheduling scheme of the data center multi-energy complementary distributed energy system.

[0014] Based on the constructed multi-energy complementary distributed energy system configuration of the data center, the system's equipment mainly includes photovoltaics, wind turbines, ground source heat pumps, gas internal combustion engines, waste heat recovery equipment, batteries, an ice storage system (ice storage tank + electric chiller unit), a thermal storage tank, an absorption chiller unit, and a gas boiler. The generators of the photovoltaics, wind turbines, and gas internal combustion engines are connected to the power grid via power lines, which are respectively connected to the electric chiller unit, the battery, and the data center. Natural gas inlet pipes are connected to the steam inlets of the gas internal combustion engine and the gas boiler. The flue gas outlet of the gas internal combustion engine is connected to the waste heat recovery equipment. The hot water outlet of the waste heat recovery equipment is connected to the hot water outlet of the gas boiler via a pipeline, and then sequentially connected to the absorption chiller unit, the thermal storage tank, and the data center along the pipeline. The cold water outlet of the electric chiller unit is connected to the ice storage tank and its outlet, and then connected to the cold water outlet of the ground source heat pump and the outlet of the absorption chiller unit via a pipeline to the data center.

[0015] Based on the mathematical model of the aforementioned equipment, the lifecycle cost, carbon emissions, and primary energy consumption of the data center multi-energy complementary distributed energy system are established as objective functions, specifically including:

[0016] Based on the mathematical model of the device, the lifecycle cost objective function LCC for the data center multi-energy complementary distributed energy system is established:

[0017]

[0018] Among them, PRVini It is the initial investment; PRV main It's the maintenance cost; PRV run It is operating cost; PRV tax It's a carbon tax; C i It refers to the capacity of device i, in kW; UP i The unit capacity cost of the i-th device is yuan / kW; ε i It is the percentage of maintenance costs to capacity costs; E grid Electricity purchased from the grid, kW·h; UP grid Electricity price, yuan / kW·h; V ng For natural gas consumption, Nm 3 ;UP ng The price of natural gas is in yuan / Nm³. 3 ACE represents annual carbon emissions in tons (t); UP tax It represents the carbon tax, expressed in yuan / ton; R is the present value factor for an ordinary annuity; i0 is the discount rate, expressed as a percentage; n is the life cycle of the equipment; r sal For the economic cycle rate;

[0019] Based on the mathematical model of the device, the lifecycle carbon emission objective function (LCCE) of the data center multi-energy complementary distributed energy system is established:

[0020]

[0021] Where, μ ng Carbon emissions per unit volume of natural gas, kg / Nm³ 3 μ grid This refers to the carbon emissions of the power grid, expressed in kg / kW·h; rec is the equipment's recycling rate; CE i Carbon emissions per unit capacity of equipment during production, transportation, and installation, expressed in kg / kW.

[0022] Based on the mathematical model of the device, the lifecycle primary energy consumption objective function LCEC of the data center multi-energy complementary distributed energy system is established:

[0023]

[0024] Where HVng is the lower heating value of natural gas, kJ / m3; ξ prd The power generation efficiency of the power grid; ξ trd For power grid transmission efficiency; EC i The primary energy consumption per unit capacity of the equipment during production, transportation, and installation, expressed in kW·h / kW.

[0025] Based on the analysis of the cold, heat, and electricity load data of the data center, its distribution characteristics are analyzed, and an operation strategy suitable for the multi-energy complementary distributed energy system of the data center is formulated. This strategy prioritizes meeting the cold load, then the electricity demand, and finally the heat demand. The operation process for meeting the cold load is as follows: first, the ground source heat pump and absorption chiller units are turned on for cooling. When the cooling capacity exceeds the cold load, the electric chiller units are turned on to make ice until the ice storage tank reaches its maximum capacity. When the cooling capacity is less than the cold load, the ice storage tank melts ice for cooling. If this still cannot meet the cold load demand, the electric chiller units are turned on for cooling. The operation process for meeting the electricity demand is as follows: the system electricity demand is determined by the electricity load, the power consumption of the electric chiller units, and the ground source heat pump... The total power consumption of the system is the sum of the power consumption of the photovoltaic, wind turbine, and gas internal combustion engine. When the power generation of the photovoltaic, wind turbine, and gas internal combustion engine exceeds the power demand, the battery stores electricity until it is fully charged. At this time, the power generation of the gas internal combustion engine is reduced first to balance supply and demand. When the power generation of the photovoltaic, wind turbine, and gas internal combustion engine is less than the power demand, the battery discharges. If the demand is still not met, electricity is purchased from the grid. The operation process of the system to meet the heat demand is as follows: the system heat demand is determined as the sum of the heat consumption power of the absorption chiller and the heat load. When the recovered heat cannot meet the heat load demand, the gas boiler is turned on for supplementary heating. When the recovered heat is greater than the heat load demand, the heat storage tank stores heat and the excess heat is discharged.

[0026] Based on the energy balance principle and actual operation, the constraints are determined, mainly including:

[0027] The cold balance constraint is determined based on the principle of energy balance:

[0028]

[0029] Among them, C load (t) represents the cooling load at time t, C gshp (t) represents the cooling power of the ground source heat pump at time t, C abs (t) represents the cooling capacity of the absorption chiller at time t. Let C be the power of the ice storage tank at time t. ec (t) represents the cooling power of the electric chiller unit at time t. Let t be the ice storage power of the ice storage tank.

[0030] The electrical balance constraints are determined based on the principle of energy balance.

[0031]

[0032] Among them, E load (t) represents the electrical load at time t, E pv (t) represents the photovoltaic power generation at time t, E wt (t) represents the wind turbine's power generation at time t, Eice (t) represents the power generation capacity of the internal combustion engine at time t. Let E be the battery discharge power at time t. grid (t) represents the power purchased from the grid at time t, E ec (t) represents the power consumption of the electric chiller unit at time t, E gshp (t) represents the power consumption of the ground source heat pump at time t. Let t be the battery charging power at time t.

[0033] The thermal balance constraint is determined based on the principle of energy balance:

[0034]

[0035] Among them, Q load (t) represents the heat load at time t, Q re (t) represents the power of the waste heat recovery system at time t, Q. abs (t) represents the heat consumption power of the absorption chiller at time t. Let t be the heat release power of the thermal storage tank. Let Q be the thermal storage power of the thermal storage tank at time t. gb (t) represents the thermal power of the gas-fired boiler at time t.

[0036] Based on actual operation, capacity limits and operational limits are determined. The operational limits are defined by the operational strategy, and the capacity limits are as follows:

[0037]

[0038] Among them, V device,i For the capacity of device i, The maximum capacity of device i.

[0039] Based on the idea of ​​population variation, the traditional multi-objective locust optimization algorithm is improved, resulting in the improved multi-objective locust optimization algorithm IMOGOA. The operators of the multi-objective locust optimization algorithm are as follows:

[0040]

[0041]

[0042] Among them, P i d Let c be the location of the locust, and c be the reduction factor. max Let c be the maximum value. min ub is the minimum value of c. d and lb d These represent the upper and lower bounds in the d-th dimension, respectively. ij For individual P j With P i distance, is the target value in the d-th dimension, l is the current iteration number, and L is the maximum iteration number.

[0043] The calculation process of IMOGOA includes: initializing the population and the number of iterations, population size, and c. min c min , L, calculate the fitness function of the initial population, find the non-dominated Pareto optimal solution and initialize the external archive, and perform the following operations within the number of iterations: update parameter c, standardize the distance between locusts to make it within the range of [1,4], set the mutation probability to perform population mutation operation, update the position of the locusts and calculate the fitness function at the current position, determine the optimal individual, update the archive members in the archive set, and the final archive set is the optimal solution set.

[0044] Compared with existing technologies, the advantages of this technology are:

[0045] (1) Objective functions for life cycle cost, carbon emissions, and primary energy consumption were established, and carbon tax was incorporated into the cost function. Compared with traditional objective functions for cost, carbon emissions, and primary energy consumption, this study considers more factors and is more comprehensive, resulting in more reliable design results.

[0046] (2) The present invention fully considers the characteristics of cold, heat and electricity loads of data centers, and the operating strategy formulated makes the supply and demand of cold, heat and electricity loads just balanced, thus ensuring the energy supply reliability of data centers.

[0047] (3) An improved multi-objective locust optimization algorithm (IMOGOA) using a population mutation mechanism provides a new method for multi-objective optimization of multi-energy complementary systems. Compared with the traditional multi-objective locust optimization algorithm (MOGOA) and the non-dominated sorting multi-objective genetic algorithm II (IMOGOA), the solution set of IMOGOA has better convergence, more uniform distribution, and wider coverage. Attached Figure Description

[0048] Figure 1 A flowchart illustrating the multi-objective optimization design and scheduling method for a data center multi-energy complementary distributed energy system provided by the present invention;

[0049] Figure 2 A schematic diagram of a data center multi-energy complementary distributed energy system provided by the present invention;

[0050] Figure 3 A flowchart illustrating the operation strategy of a data center multi-energy complementary distributed energy system provided by the present invention.

[0051] Figure 4 The flowchart of the multi-objective locust optimization algorithm based on population variation mechanism provided by the present invention is shown. Detailed Implementation

[0052] The purpose of this invention is to provide a configuration of a multi-energy complementary distributed energy system for data centers and its optimization design and scheduling method, so as to make up for the deficiencies in optimization objectives, operation strategies and solution methods, and realize clean and efficient energy supply for data centers.

[0053] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0054] like Figure 1 and Figure 2 As shown, this invention discloses a multi-energy complementary distributed energy system for data centers and its optimized design and scheduling method. Through the coordinated optimization of the design and scheduling of the multi-energy complementary distributed energy system for data centers, it achieves the overall optimal economic, environmental, and energy benefits of the system throughout its life cycle while meeting the energy demands of the data center. The main steps include:

[0055] Step 1: Obtain the data required for the optimized design and operation of the energy station, mainly including local meteorological data such as temperature, irradiance, and wind speed, as well as the cold, heat, and electrical load data of the data center.

[0056] Step 2: Construct a multi-energy complementary distributed energy system for the data center and establish mathematical models for each device. The main devices in the multi-energy complementary distributed energy system for the data center include photovoltaics, wind turbines, ground source heat pumps, gas internal combustion engines, waste heat recovery equipment, batteries, an ice storage system (ice storage tank + electric chiller unit), a thermal storage tank, an absorption chiller unit, and a gas boiler. The generators of the photovoltaics, wind turbines, and gas internal combustion engines are connected to the power grid via electrical wires, which are then connected to the electric chiller unit, batteries, and the data center. The natural gas inlet pipes are connected to the inlets of the gas internal combustion engine and the gas boiler. The flue gas outlet of the gas internal combustion engine is connected to the waste heat recovery equipment. The hot water outlet of the waste heat recovery equipment is connected to the hot water outlet of the gas boiler via a pipeline, which is then connected sequentially to the absorption chiller unit, the thermal storage tank, and the data center. The cold water outlet of the electric chiller unit is connected to the outlet of the ice storage tank, and then connected to the cold water outlet of the ground source heat pump and the outlet of the absorption chiller unit via a pipeline to the data center.

[0057] Step 3: Based on the mathematical model of the equipment, establish the life cycle cost, carbon emissions, and primary energy consumption of the data center multi-energy complementary distributed energy system as objective functions, specifically including:

[0058] Based on the mathematical model of the device, the lifecycle cost objective function LCC for the data center multi-energy complementary distributed energy system is established:

[0059]

[0060] Among them, PRV ini It is the initial investment; PRV main It's the maintenance cost; PRV run It is operating cost; PRV tax It's a carbon tax; C i It refers to the capacity of device i, in kW; UP i The unit capacity cost of the i-th device is yuan / kW; ε i It is the percentage of maintenance costs to capacity costs; E grid Electricity purchased from the grid, kW·h; UP grid Electricity price, yuan / kW·h; V ng For natural gas consumption, Nm 3 ;UP ng The price of natural gas is in yuan / Nm³. 3 ACE represents annual carbon emissions in tons (t); UP tax It represents the carbon tax, expressed in yuan / ton; R is the present value factor for an ordinary annuity; i0 is the discount rate, expressed as a percentage; n is the life cycle of the equipment; r sal For the economic cycle rate.

[0061] Based on the mathematical model of the device, the lifecycle carbon emission objective function (LCCE) of the data center multi-energy complementary distributed energy system is established:

[0062]

[0063] Where, μ ng Carbon emissions per unit volume of natural gas, kg / Nm³ 3 μ grid This refers to the carbon emissions of the power grid, expressed in kg / kW·h; rec is the equipment's recycling rate; CE i Carbon emissions per unit capacity of equipment during production, transportation, and installation, expressed in kg / kW.

[0064] Based on the mathematical model of the device, the lifecycle primary energy consumption objective function LCEC of the data center multi-energy complementary distributed energy system is established:

[0065]

[0066] Where HVng is the lower heating value of natural gas, kJ / m3; ξ prd The power generation efficiency of the power grid; ξ trd For power grid transmission efficiency; EC i The primary energy consumption per unit capacity of the equipment during production, transportation, and installation, expressed in kW·h / kW.

[0067] Step 4: Analyze the distribution characteristics of the data center load data and formulate an operation strategy suitable for the data center's multi-energy complementary distributed energy system, such as... Figure 3 As shown, the system prioritizes meeting cooling load, then electricity demand, and finally heat demand. The process for meeting cooling load is as follows: first, the ground source heat pump and absorption chiller units are activated for cooling. When the cooling capacity exceeds the cooling load, the electric chiller units are activated to make ice until the ice storage tank reaches its maximum capacity. When the cooling capacity is less than the cooling load, the ice storage tank melts ice for cooling. If this still cannot meet the cooling load demand, the electric chiller units are activated for cooling. The process for meeting electricity demand is as follows: the system electricity demand is determined as the sum of electrical load, electric chiller unit power consumption, ground source heat pump power consumption, and auxiliary equipment power consumption. When photovoltaic, wind turbine, and... When the power generation of the gas internal combustion engine exceeds the electricity demand, the battery stores electricity until it is fully charged. At this point, the power generation of the gas internal combustion engine is reduced first to balance supply and demand. When the power generation of photovoltaic, wind turbine and gas internal combustion engine is less than the electricity demand, the battery discharges. If the demand is still not met, electricity is purchased from the grid. The operation process of the system to meet the heat demand is as follows: the system heat demand is determined as the sum of the heat consumption power of the absorption chiller and the heat load. When the recovered heat cannot meet the heat load demand, the gas boiler is turned on for supplementary heating. When the recovered heat is greater than the heat load demand, the heat storage tank stores heat and the excess heat is discharged.

[0068] Step 5: Determine the constraints based on the energy balance principle and actual operation, mainly including:

[0069] Based on the principle of energy balance, the cold balance constraint is determined as follows:

[0070]

[0071] Among them, C load (t) represents the cooling load at time t, C gshp (t) represents the cooling power of the ground source heat pump at time t, C abs (t) represents the cooling capacity of the absorption chiller at time t. Let C be the power of the ice storage tank at time t. ec (t) represents the cooling power of the electric chiller unit at time t. Let t be the ice storage power of the ice storage tank.

[0072] Based on the principle of energy balance, the electrical balance constraints are determined as follows:

[0073]

[0074] Among them, E load (t) represents the electrical load at time t, E pv (t) represents the photovoltaic power generation at time t, E wt(t) represents the wind turbine's power generation at time t, E ice (t) represents the power generation capacity of the internal combustion engine at time t. Let E be the battery discharge power at time t. grid (t) represents the power purchased from the grid at time t, E ec (t) represents the power consumption of the electric chiller unit at time t, E gshp (t) represents the power consumption of the ground source heat pump at time t. Let t be the battery charging power at time t.

[0075] Based on the principle of energy balance, the thermal balance constraints are determined as follows:

[0076]

[0077] Among them, Q load (t) represents the heat load at time t, Q re (t) represents the power of the waste heat recovery system at time t, Q. abs (t) represents the heat consumption power of the absorption chiller at time t. Let t be the heat release power of the thermal storage tank. Let Q be the thermal storage power of the thermal storage tank at time t. gb (t) represents the thermal power of the gas-fired boiler at time t.

[0078] Based on actual operation, capacity limits and operational limits are determined. The operational limits are defined by the operational strategy, and the capacity limits are as follows:

[0079]

[0080] Among them, V device,i For the capacity of device i, The maximum capacity of device i.

[0081] Step 6: Establish a multi-objective optimization model based on the equipment model, operating strategy, constraints, and objective function.

[0082] Step 7: Improve the traditional multi-objective locust optimization algorithm based on the idea of ​​population variation, solve the established multi-objective optimization model, and obtain the optimal solution set.

[0083] Based on the idea of ​​population variation, the traditional multi-objective locust optimization algorithm is improved, resulting in the improved multi-objective locust optimization algorithm IMOGOA. The operators of the multi-objective locust optimization algorithm are as follows:

[0084]

[0085]

[0086] Among them, P id Let c be the location of the locust, and c be the reduction factor. max Let c be the maximum value. min ub is the minimum value of c. d and lb d These represent the upper and lower bounds in the d-th dimension, respectively. ij For individual P j With P i distance, is the target value in the d-th dimension, l is the current iteration number, and L is the maximum iteration number.

[0087] The calculation process of IMOGOA is as follows: Figure 4 As shown, this mainly includes: initializing the population and the number of iterations, population size, and c. min c max , L, calculate the fitness function of the initial population, find non-dominated Pareto optimal solutions and use them to initialize the external archive, and perform the following operations within the number of iterations: update c, standardize the distance between locusts to make it within the range of [1,4], set the mutation probability to perform population mutation operation, update the position of the locusts and calculate the fitness function at the current position, determine the optimal individual, update the archive members in the archive set, and the final archive set is the optimal solution set.

[0088] Step 8: Use the superior-inferior solution distance method to select the optimal solution from the set of non-inferior solutions, and then determine the optimal design and scheduling scheme of the data center DMES.

[0089] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0090] Taking a data center in Qinghai as an example, this data center consists of 4000m 2 Data center and 5000m 2 The facility comprises office and living areas, including a data center with 450 server racks. It acquires 8760 hours of annual cooling, heating, and electrical load data for the data center. The power grid's generation efficiency and transmission efficiency are 0.35 and 0.95, respectively. The photovoltaic panel installation area will not exceed 2000m². 2 The project is designed for a 20-year lifespan. Electricity prices are shown in Table 1. Natural gas prices and lower calorific value are 2.3 yuan / Nm³. 3 and 38931kJ / m 3 The IMOGOA population size is 200, the number of iterations is 200, the mutation rate is 0.1, and the number of archived members is 30.

[0091] Table 1 Time-of-use electricity prices

[0092]

[0093]

[0094] The multi-energy complementary distributed energy system for data centers proposed in this invention achieves a perfect balance between the supply and demand of cooling and electrical loads, meeting the energy needs of the data center. Renewable energy generation accounts for over 30% of the total power generation, demonstrating a high utilization rate of renewable energy. Furthermore, the optimization results of three algorithms—IMOGOA, MOGOA, and NSGA-II—were quantitatively compared, as shown in Table 2. SP represents the uniformity of the solution set (a smaller value is better), while MS represents the breadth of the solution set (a larger value is better). IMOGOA demonstrates superior uniformity and breadth of the solution set compared to the other two algorithms.

[0095] Table 2. Comparison of logarithmic results of different algorithms

[0096] algorithm NSGA-II MOGOA IMOGOA <![CDATA[SP(10 4 )]]> 6.71 3.27 3.11 <![CDATA[MS(10 6 )]]> 4.54 4.13 5.59 .

Claims

1. An optimized design and scheduling method for a multi-energy complementary distributed energy system in a data center, characterized in that, The optimization design and scheduling method includes the following steps: Step 1: Obtain the data required for the optimized design and operation of the energy station, including meteorological data and the cold, heat, and electricity load data of the data center; Step 2: Construct a multi-energy complementary distributed energy system for the data center and establish mathematical models for each device; Step 3: Based on the mathematical model of the device, establish the life cycle objective function of the data center multi-energy complementary distributed energy system; Step 4: Analyze the distribution characteristics of the data center's cold, heat, and electricity load data, and formulate... Operational strategies applicable to multi-energy complementary distributed energy systems in data centers; Step 5: Determine the constraints based on the principle of energy balance and actual operation; Step 6: Establish a multi-objective optimization model based on the mathematical model, operating strategy, constraints, and objective function of the equipment; Step 7: Input the relevant data required for solving the multi-objective optimization model, including meteorological parameters, load data, and equipment parameters. Improve the multi-objective locust optimization algorithm based on the idea of ​​population variation, and solve the established multi-objective optimization model to obtain the optimal solution set. Step 8: Select the optimal solution from the set of non-dominated solutions using the superior-inferior solution distance method, and then determine the optimal design and scheduling scheme of the data center multi-energy complementary distributed energy system. Step 3, based on the mathematical model of the equipment, establishes the lifecycle cost, carbon emissions, and primary energy consumption of the data center multi-energy complementary distributed energy system as objective functions, specifically including: Based on the mathematical model of the device, a lifecycle cost objective function for the data center multi-energy complementary distributed energy system is established. LCC : in, It is the initial investment; It's the maintenance cost; It is the operating cost; It's a carbon tax; It is equipment i Capacity, kW; For the first i Unit capacity cost of each device, in yuan / kW; It is the percentage of maintenance costs to capacity costs; To purchase electricity from the grid, kW h; Electricity price, yuan / kW h; For natural gas consumption, Nm 3 ; The price of natural gas is in yuan / Nm³. 3 ; Carbon emissions per year, in tons; It's a carbon tax, in yuan / ton; R It is the present value factor of an ordinary annuity; It is the discount rate, % n For the lifecycle of the equipment; For the economic cycle rate; Based on the mathematical model of the device, a lifecycle carbon emission objective function for the data center multi-energy complementary distributed energy system is established. LCCE : in, Carbon emissions per unit volume of natural gas, kg / Nm³ 3 ; This refers to the carbon emissions of the power grid, in kg / kW. h; It refers to the equipment's recycling rate; Carbon emissions per unit capacity of equipment during production, transportation, and installation, expressed in kg / kW; Based on the mathematical model of the device, a lifecycle primary energy consumption objective function for the data center multi-energy complementary distributed energy system is established. LCEC : in, The lower heating value of natural gas is kJ / m3; The power generation efficiency of the power grid; For power grid transmission efficiency; Primary energy consumption per unit capacity of equipment during production, transportation, and installation (kW) h / kW.

2. The optimization design and scheduling method for a data center multi-energy complementary distributed energy system according to claim 1, characterized in that, The equipment in the multi-energy complementary distributed energy system constructed in step 2 includes photovoltaics, wind turbines, ground source heat pumps, gas internal combustion engines, waste heat recovery equipment, batteries, ice storage tanks, and an ice storage system composed of electric chillers, a heat storage tank, an absorption chiller, and a gas boiler. The generators of the photovoltaics, wind turbines, and gas internal combustion engines are connected to the power grid via wires, which are then connected to the electric chiller, batteries, and data center, respectively. The natural gas inlet pipes are connected to the inlets of the gas internal combustion engine and the gas boiler, respectively. The flue gas outlet of the gas internal combustion engine is connected to the waste heat recovery equipment. The hot water outlet of the waste heat recovery equipment is connected to the hot water outlet of the gas boiler via a pipeline, and then sequentially connected to the absorption chiller, the heat storage tank, and the data center along the pipeline. The cold water outlet of the electric chiller is connected to the ice storage tank and its outlet, respectively, and then connected to the cold water outlet of the ground source heat pump and the outlet of the absorption chiller via a pipeline to the data center.

3. The optimization design and scheduling method for a data center multi-energy complementary distributed energy system according to claim 1, characterized in that, In step 4, the distribution characteristics of the data center's cold, heat, and electricity load data are analyzed to formulate... The operational strategy for a multi-energy complementary distributed energy system in a data center prioritizes meeting cooling load, then electricity demand, and finally heat demand. The process for meeting cooling load is as follows: first, the ground source heat pump and absorption chiller units are activated for cooling. When the cooling capacity exceeds the cooling load, electric chiller units are activated to produce ice until the ice storage tank reaches its maximum capacity. When the cooling capacity is less than the cooling load, the ice storage tank melts ice for cooling. If this still cannot meet the cooling load demand, electric chiller units are activated for cooling. The process for meeting electricity demand is as follows: the system's electricity demand is determined by the electrical load, the power consumption of the electric chiller units, the power consumption of the ground source heat pump, and the power consumption of auxiliary equipment. In summary, when the power generation of photovoltaic, wind turbine, and gas internal combustion engine exceeds the electricity demand, the battery stores electricity until it is fully charged. At this point, the power generation of the gas internal combustion engine is reduced first to balance supply and demand. When the power generation of photovoltaic, wind turbine, and gas internal combustion engine is less than the electricity demand, the battery discharges. If the demand is still not met, electricity is purchased from the grid. The system's operation process to meet heat demand is as follows: the system heat demand is determined as the sum of the heat consumption power of the absorption chiller and the heat load. When the recovered heat cannot meet the heat load demand, the gas boiler is turned on for supplementary heating. When the recovered heat is greater than the heat load demand, the heat storage tank stores heat, and the excess heat is discharged.

4. The optimization design and scheduling method for a data center multi-energy complementary distributed energy system according to claim 1, characterized in that, Step 5, which involves determining the constraints based on the energy balance principle and actual operation, includes: Based on the principle of energy balance, the cold balance constraint is determined as follows: in, for t The cooling load at all times, for t Real-time ground source heat pump cooling capacity, for t The cooling capacity of the absorption chiller at any given time. for t The power of the ice storage tank at all times. for t The cooling capacity of the refrigeration unit is constantly changing. for t Ice storage tank capacity Based on the principle of energy balance, the electrical balance constraints are determined as follows: in, for t Constant electrical load, for t Photovoltaic power generation at all times for t Wind turbine power generation capacity at all times for t The power generation capacity of the internal combustion engine at any given time. for t Battery discharge power at all times for t Power purchased from the grid at all times for t Power consumption of the refrigeration unit at all times. for t The power consumption of a ground source heat pump at all times. for t Battery charging power at all times; Based on the principle of energy balance, the thermal balance constraints are determined as follows: in, for t Constant heat load, for t The power of the waste heat recovery system at all times. for t The heat consumption power of a constant absorption chiller for t The heat release power of the thermal storage tank at all times. for t The thermal storage capacity of the thermal storage tank at all times. for t The thermal power of the gas-fired boiler at all times; Based on actual operation, capacity limits and operational limits are determined. The operational limits are defined by the operational strategy, and the capacity limits are as follows: in, For equipment i capacity, For equipment i The maximum capacity.

5. The optimization design and scheduling method for a data center multi-energy complementary distributed energy system according to claim 1, characterized in that, In step 7, the multi-objective locust optimization algorithm is improved based on the idea of ​​population mutation. Specifically, the mutation probability of individuals in the population is set during computation to generate new individuals, thereby improving the population's search capability. This results in the improved multi-objective locust optimization algorithm IMOGOA. The operators of the multi-objective locust optimization algorithm are as follows: in, The location of the locusts. c The reduction factor is... c max for c The maximum value, c min for c The minimum value, and Representing respectively in the d Upper and lower limits in dimension For individuals and distance, It is the first d Dimensional target value, l This represents the current iteration number. L This represents the maximum number of iterations. The computation process of the multi-objective locust optimization algorithm IMOGOA includes: initializing the population and determining the number of iterations, population size, etc. c min , c max , L Calculate the fitness function of the initial population, find the non-dominated Pareto optimal solution and initialize the external archive, and perform the following operations within the number of iterations: update the reduction coefficient. c Standardize the distance between locusts to make them within the range of [1,4], set the mutation probability to perform population mutation operation, update the position of the locusts, calculate the fitness function at the current position, determine the optimal individual, update the archive members in the archive set, and the final archive set is the optimal solution set.

6. The optimization design and scheduling method for a data center multi-energy complementary distributed energy system according to claim 1, characterized in that, The meteorological data mentioned in step 1 includes local temperature, irradiance, and wind speed.

Images