Collaborative planning method for multi-microgrid system integrating cold-electricity hybrid shared energy storage and data center

Abstract

The invention discloses a multi-microgrid system collaborative planning method integrating cold-electricity hybrid shared energy storage and a data center. A commercial operation and income model of a cold-electricity hybrid shared energy storage station covering investment, operation and maintenance, energy transaction and service charge is established; cold energy released in the LNG gasification process is coupled with the cooling requirement of the data center, and an LNG-data center integrated refrigeration model is constructed; constructing a multi-microgrid system collaborative operation model containing the data center; and constructing a double-layer planning model taking the lowest cost of the shared energy storage station as an upper-layer optimization target and the lowest collaborative operation cost of the multi-microgrid system and the data center as a lower-layer optimization target, and solving the double-layer planning model to obtain the optimal configuration capacity of the energy storage station and a collaborative operation strategy of the system. According to the method, through integration of multi-dimensional transaction and cooperative scheduling of electric energy and cold energy, efficient consumption of renewable energy sources, flexible adaptation of load characteristics and comprehensive improvement of system operation economy are realized.

Description

Technical Field

[0001] This invention relates to the field of shared energy storage planning, operation and optimization technology for regional multi-energy systems, and specifically to a collaborative planning method for multi-microgrid systems that integrates cooling-electricity hybrid shared energy storage and data centers. Background Technology

[0002] The current energy and power system faces two major challenges: First, data centers, as the cornerstone of the digital economy, are experiencing a continuous increase in total energy consumption, with cooling systems accounting for as much as 30%-40% of this energy consumption. Furthermore, ensuring power supply reliability requires independent energy storage facilities, which face the bottleneck of high initial investment costs. Second, the large-scale integration of renewable energy sources such as wind and solar power has brought severe volatility and intermittency, leading to "wind and solar curtailment" in microgrid systems and insufficient renewable energy absorption capacity. To address these challenges, existing technological solutions mainly focus on two directions: one is researching energy-saving technologies for data centers themselves, such as optimizing server layout and adopting efficient cooling solutions like liquid cooling; the other is researching energy management within microgrid systems, such as using combined cooling, heating, and power (CCHP) to improve overall energy efficiency or configuring independent energy storage to mitigate renewable energy fluctuations. However, existing technological solutions have significant limitations: First, most studies treat data centers, shared energy storage, and multi-microgrid systems as independent optimization entities, or only perform pairwise coordination, such as data center-microgrid or energy storage-microgrid, failing to deeply reveal the coupling mechanism and overall benefits of their coordinated operation at the system architecture level in addressing the spatiotemporal uncertainties of new energy sources and mitigating overall energy storage configuration costs. Second, given the continuously expanding scale of my country's liquefied natural gas (LNG) imports, existing research has also failed to effectively explore the recovery of the large amount of high-grade cold energy released during its gasification process and its coupling with the cooling needs of data centers, thus missing a potential synergistic energy-saving path that could significantly reduce cooling power consumption and carbon emissions at the source. Furthermore, traditional single-layer optimization models struggle to coordinate the coupling relationship between the long-term capacity planning of shared energy storage power stations and the short-term operational strategies of microgrid systems containing data centers, leading to a disconnect between planning and operation, and the overall system economics failing to reach optimal levels. Summary of the Invention

[0003] To address issues such as reducing the cost of independently configuring energy storage in data centers, lowering the high energy consumption ratio of internal cooling systems, and addressing the spatiotemporal mismatch in the absorption of renewable energy by microgrids, this invention proposes a collaborative planning method for multi-microgrid systems that integrates hybrid shared energy storage (cooling-electricity) with data centers. This method achieves efficient absorption of renewable energy, flexible adaptation to load characteristics, and a comprehensive improvement in the economic efficiency of system operation by integrating multi-dimensional trading and collaborative scheduling of electrical and cooling energy.

[0004] The technical solution adopted in this invention is as follows: A collaborative planning method for multi-microgrid systems integrating hybrid energy storage and data centers includes the following steps: Step 1: Analyze the business operation model of the cold-electric hybrid shared energy storage station, and establish a business operation and revenue model for the cold-electric hybrid shared energy storage station that covers investment, operation and maintenance, energy trading and service fees; Step 2: Couple the cold energy released during LNG gasification with the cooling requirements of the data center to construct an integrated LNG-data center refrigeration model, thereby significantly reducing system cooling energy consumption and carbon emissions; Step 3: Based on the LNG-data center integrated cooling model constructed in Step 2, construct a collaborative operation model of a multi-microgrid system including a data center; Step 4: Combining the commercial operation and revenue model of the hybrid shared energy storage station established in Step 1, the integrated LNG-data center cooling model constructed in Step 2, and the collaborative operation model of the multi-microgrid system with data center established in Step 3, a two-layer programming model is constructed with the lowest cost of the shared energy storage station as the upper-level optimization objective and the lowest cost of collaborative operation of the multi-microgrid system and data center as the lower-level optimization objective. The two-layer programming model is then solved to obtain the optimal configuration capacity of the energy storage station and the collaborative operation strategy of the system.

[0005] In step 1, a business operation model for a shared energy storage station invested and operated by a third-party operator is designed, and a comprehensive revenue model is established that includes average daily investment and maintenance costs, typical daily energy purchase costs, energy sales revenue and service fee income, to provide storage and release services for electrical and cooling energy for multi-microgrid user groups, including data centers. 1) The average daily investment and maintenance costs are: (1); In formula (1): This represents the average daily investment and maintenance costs; two energy types were considered. K =1 represents electrical energy. K =2 represents cold energy; and The figures represent the power cost and capacity cost of different energy sources for the energy storage station, respectively, in yuan / kW and yuan / (kWh); and These represent the maximum charging and discharging power and maximum capacity of the energy storage station for different energy sources; The expected number of days of use for the energy storage station; Daily maintenance costs for different energy storage devices.

[0006] 2) The energy purchase cost for each typical daily energy storage station is: (2); In formula (2): The cost of purchasing energy for a typical daily energy storage station; N The number of microgrids; This represents the number of time periods in the scheduling cycle. for t The price at which microgrids sell energy to energy storage stations during specific time periods; For scheduling periods; For each typical day i micro-network t Energy sold to energy storage stations during specific time periods; i This indicates the number of the microgrid, where... i A value of 1 indicates data from microgrid 1; t It means 24 hours in a day. t A value of 1 represents a data point of 1. 3) The revenue from energy sales from each typical daily energy storage station to the microgrid is: (3); In formula (3): This represents the revenue generated by the energy storage station selling energy to the microgrid on a typical day; for t The price at which the microgrid purchases energy from the energy storage station during specific time periods; For each typical day i micro-network t Energy purchased from the energy storage station during specific time periods; 4) The service fee charged by the microgrid for each typical daily energy storage station is: (4); In equation (4): This represents the service fee charged by the energy storage station to the microgrid for each typical day; for t The microgrid pays a service fee per unit price to the energy storage station.

[0007] In step 2, considering the large proportion of energy consumption in traditional data center cooling, resulting in excessive electricity consumption, an integrated coupled cooling model of LNG cold energy and data center cooling needs is established. Utilizing the cold energy released during the liquefied natural gas vaporization process, a two-stage heat exchange is performed between the data center's chilled water system and a circulating loop using ethylene glycol aqueous solution as the refrigerant. This constructs a data center cooling system based on LNG cold energy recovery, achieving cascaded utilization and direct supply of cold energy to reduce the electricity consumption and carbon emissions of conventional data center cooling. 2.1: Theoretical Model for Cold Energy Recovery (5); In formula (5): This indicates the cold energy released during the LNG vaporization process; This indicates that the LNG is used for actual cooling in data centers during the vaporization process. and These represent LNG at 25 and -163 Enthalpy below; Indicates the mass of LNG; , and These represent primary heat exchange efficiency, secondary heat exchange efficiency, and actual efficiency, respectively. This represents pipe cooling loss; 2.2: The main operating costs of a data center include server operating energy consumption and server cooling energy consumption, expressed as follows: (6); In formula (6): Indicates that the data center is in t Total operating energy consumption at any given time; and They represent t Real-time server operating energy consumption and server cooling energy consumption; 2.3: The total server cost can be expressed as: (7); In equation (7): and These represent the server's power consumption and cooling costs, respectively. express t The electricity price on the power grid at any given time; express t The data center purchases electricity from the power grid at any given time; and express t The electricity price and service fees that the Time Data Center purchases from shared energy storage stations; , and express The data center purchases cooling from shared energy storage stations at a price that includes both the purchase price and the selling price, as well as service fees. express t The data center purchases electricity from shared energy storage stations. express t The real-time data center sells cooling capacity to the shared energy storage station; express t The amount of cooling space purchased for the data center at any given time. express t The real-time data center sells cooling capacity to the shared energy storage station; This indicates the scheduling period, which is 24 hours in this case.

[0008] 2.4: In addition to using the cooling energy released during the LNG vaporization process, the data center also uses electric chiller units for auxiliary cooling. (8); In equation (8): Indicates that the electric refrigeration unit is in t The electrical energy consumed constantly; This indicates that the server is in t The cooling energy generated by the constant-time electric refrigeration unit; This represents the ratio of electrical energy consumed by the refrigeration device to the cooling power generated.

[0009] Step 2: The integrated coupled cooling model of LNG cold energy and data center cooling needs includes: The theoretical model for cold energy recovery is shown in equation (5); The operating cost of the data center is shown in Equation (6); The total server cost is shown in equation (7); The electric refrigeration unit provides auxiliary cooling, as shown in equation (8).

[0010] In step 3: Based on the LNG-data center integrated cooling model constructed in step 2, a multi-microgrid system collaborative operation model containing a data center is constructed; It includes power balance, as shown in equation (9); The cold power balance formula is shown in equation (10); To address the differentiated load and structural characteristics of combined cooling, heating, and power (CCHP) microgrids and data center microgrids, internal power balance constraints for electricity and cooling are established: 3.1: The power balance of the data center is shown below: (9); In equation (9): express t The server's energy consumption at any given time; Indicates that the electric refrigeration unit is in t The electrical energy consumed constantly; For each typical day i Individual micro-networks t Power purchased from the microgrid during specific time periods; express t The data center purchases electricity from shared energy storage stations.

[0011] 3.2: The cold power balance of the data center is shown below: (10); In formula (10): This indicates that the server is in t The cooling energy generated by the constant-time electric refrigeration unit; express t The cooling energy used for actual data center cooling during the LNG vaporization process; expresst The real-time data center sells cooling capacity to the shared energy storage station; express t The amount of cooling space purchased by the data center at any given time.

[0012] In step 4, considering that existing research generally treats shared energy storage, multi-microgrid and data center as independent entities or only coordinates them in pairs when planning multi-microgrid systems with data centers, it fails to reveal the coupling mechanism of deep collaboration among the three in smoothing the spatiotemporal fluctuations of new energy and sharing the investment cost of energy storage. Therefore, this step proposes a technical solution to establish and solve a two-layer optimization configuration model for the system.

[0013] In step 4, the KKT conditions and the Big-M method are used to transform the bilevel programming model into a mixed-integer linear programming problem for solution, thereby obtaining the optimal configuration capacity of the energy storage station and the coordinated operation strategy of the system. Specifically: (1) Objective function: The upper-level optimization objective is to minimize the annual operating cost of the energy storage station, namely: (11); In equation (11): This represents the annual operating cost of a shared energy storage station. W The number of typical days; For the first w The number of days corresponding to a typical day; This represents the average daily investment and maintenance cost of the energy storage station. The cost of purchasing electricity from the microgrid for each typical day's energy storage station; The cost of selling electricity to the microgrid from each typical daily energy storage station; Service fee for each typical day of energy storage station.

[0014] The lower-level optimization objective is to minimize the annual operating cost of combined cooling, heating, and power (CCHP) multi-microgrid systems and data centers based on energy storage stations, namely: (12); (13); (14); In the above formula: This indicates the annual operating cost of multi-micronet systems and data centers; This represents the total cost of electricity purchased by the data center from the shared energy storage station; The cost of purchasing electricity from the grid for each typical day; for t Electricity purchase price for a given period of time; For each typical day i Individual micro-networks t Power purchased from the grid during the specified time period; Total gas cost for each typical day; for t The price per unit volume of natural gas at any given time; For each typical day i Individual micro-networks t Output power of the gas turbine during a given period; The power generation efficiency of micro gas turbines; The calorific value of the gas is taken as 9.7 kWh / m³. 3 ; For each typical day i Individual micro-networks t The output thermal power of the gas-fired boiler during a specific time period; The efficiency of the gas-fired boiler.

[0015] (2) Constraints: ①. Energy storage station rate constraint: The capacity and rated power of an energy storage station are directly proportional, specifically: (15); In equation (15): Indicates the capacity of the shared energy storage station; Energy ratio of the energy storage station; Indicates the rated power of the shared energy storage station; ②. Constraints on the state of charge and charging / discharging power of energy storage stations: (16); (17); In the above formula: Indicates shared energy storage station t The storage capacity of the Kth energy source during a given time period; Indicates shared energy storage station t The storage capacity of the Kth energy source during time period -1; and The charging and discharging efficiency of energy storage devices; and for t Real-time charging and discharging power of the energy storage station; To initially store energy for the energy storage station; The energy stored in the energy storage station; and These are the charging and discharging status bits of the energy storage station, respectively, and are 0-1 variables.

[0016] ③. Electrical balance constraints: (18); In formula (18): Indicates the first i Individual micro-networks t Wind power output during certain periods; This means the first i Individual micro-networks t Solar power output during specific time periods; Indicates the number of typical days i Individual micro-networks t Power purchased from the grid during the specified time period; This represents the energy storage station's capacity per typical day. t The time period starts from the i The electricity purchase capacity of each microgrid; For each typical day i Individual micro-networks t Power consumption of the time-limited electric chiller; For the first i Individual micro-networks t Electrical load power during a given time period.

[0017] ④. Cold balance constraint: (19); In equation (19): Indicates the first i Microgrid electric chillers in t Power consumption during a given time period; The energy efficiency ratio of an electric chiller; For each typical day i Individual micro-networks t Output cooling power of a time-period absorption chiller; For each typical day, the i-th micronet in t Cooling load power during specific time periods.

[0018] ⑤. Hydrogen balance power: (20); In equation (20): Indicates that the electrolytic cell equipment is in t Hydrogen gas produced during the period; Indicates the first i Individual micro-networks t Hydrogen load power during the time period.

[0019] ⑥. Thermal equilibrium power: (twenty one); In equation (21): For each typical day i Individual micro-networks t The output heating power of the time-period heat exchanger; For each typical day i Individual micro-networkst Heat load power during a given time period.

[0020] ⑦ Waste heat balance of waste heat boiler: (twenty two); In equation (22): The heat exchange efficiency of the heat exchange device; The energy efficiency ratio of an absorption chiller; The heat-to-electric ratio of a gas turbine; This refers to the efficiency of the waste heat boiler.

[0021] ⑧. Power balance constraints for charging and discharging of shared energy storage stations: The energy purchase and sale power of each combined cooling, heating and power microgrid and shared energy storage station, as well as the sum of the energy purchase and sale power of the data center and shared energy storage station, constitute the charging and discharging power of the energy storage station.

[0022] (twenty three); 9. Output limits of microgrid devices: (twenty four); In equation (24): and These represent the minimum and maximum output of the gas turbine; and These represent the minimum and maximum output values ​​of the absorption chiller. and These represent the minimum and maximum output of the electric chiller. and These represent the minimum and maximum output values ​​of the gas-fired boiler. and These represent the minimum and maximum output values ​​of the heat exchanger. and These represent the minimum and maximum output power of the electrolytic cell.

[0023] ⑩. Power grid purchase constraints: (25); In equation (25): This represents the maximum power that the microgrid purchases from the grid during that period.

[0024] Transmission constraints of shared energy storage stations: (26); In equation (26): This represents the maximum interactive power between the microgrid and the energy storage station. and Each typical day i The status of energy purchase and sale between microgrids and energy storage stations.

[0025] In step 4, the bi-level programming model includes the formula for the upper-level objective function, as shown in equation (11); Upper-level constraints include: The energy storage station rate constraint is shown in Equation (15); The state of charge and charging / discharging power constraints of the energy storage station are shown in Equations (16) and (17), respectively. The bi-level programming model includes the formula for the lower-level objective function, as shown in equation (12); Lower-level constraints include: The electrical balance constraint is shown in equation (18); The cold balance constraint is shown in equation (19); The hydrogen balance constraint is shown in equation (20); Thermal equilibrium constraints are shown in equation (21); The waste heat balance constraint of the waste heat boiler is shown in equation (22); The power balance constraint for charging and discharging of shared energy storage stations is shown in Equation (23); The upper and lower limits of the output of microgrid devices are constrained as shown in equation (24); The power grid purchase constraints are shown in equation (25); The transmission constraints of the shared energy storage station are shown in Equation (26); The bilevel programming model is equivalently transformed into a mixed-integer linear programming problem by applying the KKT conditions and the Big-M method for solution. Specifically: First, the KKT method is used to transform the lower-level model into additional constraints for the upper-level model. The first step is to construct the Lagrangian function of the lower-level model. Each constraint in the entire bi-level model corresponds to a different Lagrange multiplier; one equality constraint corresponds to one Lagrange multiplier, and one inequality constraint corresponds to two Lagrange multipliers. Therefore, the equality constraints—electric power balance, cold power balance, thermal power balance, waste heat boiler waste heat balance, energy storage power station charge / discharge power balance, and hydrogen balance—correspond to Lagrange multipliers. , , , , , Inequality constraints. The upper and lower limits of microgrid device output in the inequality constraints correspond to Lagrange multipliers from top to bottom. , , , , , , , , , , and The grid purchase constraint in the inequality constraints corresponds to the Lagrange multiplier. and The transmission constraint of the shared energy storage station in the inequality constraint, when K =1 corresponds to the Lagrange multiplier. , , , and In the inequality constraints, the transmission constraints of shared energy storage stations K =2 corresponds to Lagrange multipliers , , , , . Then, the KKT conditions consist of stationarity conditions, primal feasibility, dual feasibility, and complementary relaxation conditions.

[0026] The stationarity condition states that the gradient of the objective function at the optimal solution can be expressed as a linear combination of the gradients of the constraint functions. Simply put, it involves differentiating the Lagrangian function, primarily with respect to the decision variables, such as... , , , , and wait; For example, to Taking the derivative, the stationarity condition we obtain is: ; Next, the same process is repeated for the other decision variables.

[0027] The subsequent conditions for primal feasibility, dual feasibility, and complementary relaxation are as follows: and Let me illustrate with an example; The original feasibility conditions require that the original inequality conditions be met; that is: ; The duality feasibility condition requires that the Lagrange multipliers of the inequality be nonnegative; that is: ; Complementary relaxation conditions require that at least one of the multipliers and the constraint function values ​​be zero. That is; .

[0028] After these steps, the lower-level model is transformed into the upper-level model with additional constraints; that is, the two-level model is transformed into a single-level model. The transformed model also has nonlinear constraints, such as the three types of constraints derived from the inequality constraints mentioned earlier; as follows: ; The constraint described above cannot be solved by a solver, therefore we need to use the Big-M method for transformation; that is... ; Finally, through KKT and Big-M, the bilevel programming model described in the model was equivalently transformed into a mixed-integer linear programming problem for solution.

[0029] In step 4, the optimal configuration capacity of the energy storage station and the coordinated operation strategy of the system are obtained. See Table 1 for details.

[0030] As shown in Table 1, the energy storage capacity obtained in Scenario 4 is based on the goal of minimizing the cost of the shared energy storage station at the upper level and minimizing the cost of the microgrid system at the lower level. Therefore, this capacity represents the lowest-cost configuration capacity while ensuring the needs of microgrid users, and is thus called the optimal configuration capacity. The system's collaborative operation strategy involves the energy flow between the microgrid and the shared energy storage station, the output of each microgrid device, and the interaction with the upper-level power grid under this configuration capacity; for example, ... Figure 10 As shown, this illustrates the 24-hour energy flow of the microgrid system and shared energy storage station under this configuration capacity. In other words, it ultimately provides the power output of various facilities over a 24-hour period and the final configured energy storage capacity.

[0031] This invention discloses a collaborative planning method for multi-microgrid systems that integrates hybrid energy storage and data centers, with the following technical advantages: 1) In step 1, this invention addresses the high cost of traditional independent energy storage configuration by adopting the currently popular sharing concept to establish a shared energy storage station, which greatly reduces energy storage costs. As shown in Table 2, the configuration capacity of scenario 2 under independent energy storage configuration is greatly increased, leading to a significant increase in its cost.

[0032]

[0033] Secondly, in response to the fact that most research on shared energy storage focuses on the single issue of electricity sharing, this paper incorporates cold energy as a tradable and storable energy commodity into the commercial and dispatch framework, and creatively establishes a commercial sharing model of "cold-electric hybrid" energy, realizing multi-energy synergistic optimization. As shown in the table above, under multi-energy synergy, the energy storage configuration can be further reduced, and the cost can be reduced.

[0034] 2) This invention keenly recognizes three long-standing and unresolved technical challenges: high energy consumption for data center cooling, large carbon emissions from traditional cooling methods, and significant waste of LNG cooling energy. However, the supply of LNG cooling energy alone is constrained by the continuity of the gasification process, and its supply curve may not perfectly match the dynamically changing cooling load curve of the data center, posing a risk of instantaneous supply-demand mismatch. Therefore, the integrated coupled cooling model of LNG cooling energy and data center cooling demand established in step 2 of this invention organically complements the shared energy storage system established in step 1, jointly shaping a flexible new paradigm of collaborative cooling. In summary, step 2 not only proposes the concept of "using LNG cooling energy for data center cooling," but also, for the first time, constructs a complete, quantifiable, optimizable technical implementation model and system architecture that is deeply integrated with market mechanisms and shared energy storage.

[0035] 3) such as Figure 11 As shown, in step 3, this invention addresses the problem of difficult renewable energy consumption in traditional microgrid systems by creatively linking the microgrid system and the data center model established in step 2 through the shared energy storage model established in step 1. In the context of my country's east-west computing, this invention utilizes the high energy consumption characteristics of data centers to achieve renewable energy consumption in microgrid systems. Figure 11 The devices on the right are all in the microgrid. Through shared energy storage stations, the microgrid and data center are connected in the form of energy.

[0036] 4) Step 4 of this invention mainly combines the models established in steps 1, 2 and 3, and then creatively constructs a two-layer optimization decision model that is precisely matched to an unprecedented complex energy system "cold-electric hybrid shared energy storage + LNG data center + multi-microgrid", and overcomes its solution problem. Attached Figure Description

[0037] The present invention will be further described below with reference to the accompanying drawings and examples; Figure 1 Power curves for various data centers.

[0038] Figure 2 The power curves for each microgrid on a typical day in spring are shown.

[0039] Figure 3 This is a schematic diagram of a microgrid structure.

[0040] Figure 4 A comparison chart of the power output of various microgrids under multiple loads.

[0041] Figure 5 The result of capacity configuration for shared energy storage power stations.

[0042] Figure 6 The results of the charging and discharging optimization for shared energy storage power stations.

[0043] Figure 7 The results of the optimized charging and discharging of the electric-cooling hybrid shared energy storage system.

[0044] Figure 8 The energy purchase and sale results for each microgrid with hybrid electric-cooling energy storage.

[0045] Figure 9 This is a structural diagram of a data center multi-microgrid and shared energy storage system for cooling and power generation.

[0046] Figure 10 A 24-hour energy flow diagram for microgrid systems and shared energy storage stations.

[0047] Figure 11 A schematic diagram showing the connection between the microgrid system and the data center model established in step 2, and the shared energy storage model established in step 1. Detailed Implementation

[0048] This invention aims to address issues such as the high cost of independently configuring energy storage in data centers, the high energy consumption of internal cooling systems, and the spatiotemporal mismatch in the absorption of renewable energy by microgrids. A collaborative planning method for multi-microgrid systems integrating a hybrid cooling-electricity shared energy storage station with the data center is proposed, and its significant advantages in reducing energy storage configuration and system operating costs, and improving renewable energy absorption capacity are verified.

[0049] A collaborative planning method for multi-microgrid systems integrating hybrid energy storage and data centers includes the following steps: Step S1: Construct a business operation and revenue model for a hybrid cold-electricity shared energy storage station. Design a business model for a shared energy storage station invested by a third-party operator, and establish a comprehensive revenue model that includes average daily investment and maintenance costs, energy purchase costs, energy sales revenue, and service fee income, providing energy storage leasing and energy services to multi-microgrid user groups.

[0050] Step S2: Establish an integrated coupled refrigeration model for LNG cold energy and data center cooling needs. Utilize the cold energy released during the liquefied natural gas vaporization process, and conduct two-stage heat exchange with the data center cooling system through a refrigerant circulation loop to construct a cold energy recovery and cascade utilization model, thereby reducing the power consumption and carbon emissions of conventional data center cooling.

[0051] Step S3: Combining the integrated coupled cooling model from Step S2, construct a collaborative operation model for a multi-microgrid system including a data center. Establish internal power and cooling balance constraints for microgrids with combined cooling, heating, and power (CCHP) and data center microgrids, taking into account their differentiated load and structural characteristics.

[0052] Step S4: Combining the commercial operation and revenue model of the hybrid shared energy storage station in Step S1, the integrated coupled cooling model in Step S2, and the collaborative operation model of the multi-microgrid system with data center in Step S3, construct a two-layer programming model with the lowest cost of the shared energy storage station as the upper-layer optimization objective and the lowest cost of collaborative operation of the multi-microgrid and data center as the lower-layer optimization objective; and use KKT conditions and Big-M method to convert the two-layer model into a mixed integer linear programming problem for solution, to obtain the optimal configuration capacity of the energy storage station and the collaborative operation strategy of the system.

[0053] Example: (a) Example parameters: Based on step S1, a business operation and revenue model for the hybrid cold-electricity shared energy storage station is established. Energy storage station data is shown in Table 3; service prices are shown in Table 4.

[0054]

[0055] Based on step S2, an integrated coupled cooling model of LNG and data center is established, including the data center's electrical load, cooling load, and LNG-released cooling energy as follows: Figure 1 As shown.

[0056] Based on step S3, establish a collaborative operation model for a multi-microgrid system including a data center. The electrical load, cooling load, heating load, hydrogen load, and wind / solar output of the multi-microgrid system diagram are as follows: Figure 2 As shown in Table 5, the parameters of each device in the microgrid are shown in Table 5.

[0057]

[0058] A two-layer optimization configuration model is established and solved according to step S4. This embodiment uses actual energy data from a region in Jiangsu Province, China as a foundation to construct a test system comprising three combined cooling, heating, and power (CCHP) microgrids (MG1, MG2, MG3) and one microgrid (MG4) containing a data center. The system structure and... Figure 1 The results are consistent with those shown. The scheduling year is divided into four typical seasons: spring, summer, autumn, and winter. This paper selects a typical day in spring for simulation, with a scheduling cycle of 24 hours and a time interval of 1 hour.

[0059] (II) Optimization Result Configuration Analysis: To investigate the impact of constructing shared energy storage power stations on the scheduling results and economic efficiency of multi-microgrid systems and data centers, this paper sets up three different scenarios for comparative analysis.

[0060] The simulation scenario is set up as follows: 1) Scenario 1: In a multi-microgrid system containing a data center, no energy storage is configured. It operates independently, and excess power is directly discarded, while insufficient power is purchased from the grid.

[0061] 2) Scenario 2: Multi-microgrid systems with data centers select to configure independent energy storage devices, and the energy storage charging and discharging efficiency and other parameters of the independent energy storage devices are the same as those of the shared energy storage power station.

[0062] 3) Scenario 3: Multi-microgrid systems with data centers participate in shared energy storage power station services and use the energy storage charging and discharging services of the energy storage power station.

[0063]

[0064] The collaborative planning method for multi-microgrid systems integrating a hybrid cold-electricity shared energy storage station and a data center proposed in this invention demonstrates significant advantages in system economy and operational efficiency in both Scenario 3 and Scenario 4. Compared to Scenario 1 without energy storage configuration and Scenario 2 where each microgrid has its own independent energy storage configuration, Scenario 3 and Scenario 4, which adopt a shared energy storage model, effectively reduce the system's total lifecycle operating cost while achieving full renewable energy consumption (100% consumption rate) and intensive energy storage capacity configuration. Specifically, the annual operating cost of Scenario 3 is reduced by approximately 24.9% compared to Scenario 1, and the energy storage configuration capacity is reduced by approximately 52.1% compared to Scenario 2; Scenario 4 further optimizes this, achieving the lowest annual operating cost and a more economical energy storage configuration scale. In addition, the shared energy storage power station, as an independent operating entity, achieved revenues of RMB 8.7876 million and RMB 7.7882 million in Scenario 3 and Scenario 4, respectively, verifying its feasibility for commercial operation.

[0065] In summary, this invention, through the deep integration of multi-microgrid, shared energy storage, and data centers in terms of energy supply and cold energy utilization, systematically solves the problems of high cost of independently configuring energy storage in data centers, large energy consumption of cooling systems, and difficulties in the spatiotemporal absorption of new energy sources. It provides an effective technical path and methodological support for achieving low-carbon, efficient, and economical coordinated operation of regional multi-energy systems.

[0066] Comparison of multi-load power of each microgrid and the capacity configuration results of shared energy storage power stations are as follows: Figure 4 and Figure 5 As shown, from Figure 4 and Figure 5 It can be seen that each microgrid has different load characteristics. For example, microgrid 2 has a high heat load demand, while the data center has relatively high electrical and cooling loads. Figure 4 and Figure 5Data centers have significant electrical and cooling load demands. With shared energy storage, the renewable energy generated by microgrids can be sold to data centers through shared power stations, which is crucial for alleviating grid burden during peak electricity demand and promoting energy upgrades. Furthermore, compared to Scenario 2, the operating costs of microgrids and the construction costs of energy storage have decreased by 59.08% and 73.60% respectively. While microgrid costs have increased compared to Scenario 1, it's practically impossible to operate without energy storage. Large electricity consumers like data centers have high and uninterrupted power needs, necessitating energy storage. Moreover, as the service fees and unit prices of energy storage power stations decrease later, microgrid costs will further decline. Therefore, Scenario 3 connects microgrids through shared energy storage power stations. The differences and complementarities in the electricity consumption behaviors of microgrids and data centers create a degree of complementarity in their energy storage needs. This allows shared energy storage power stations to be configured with smaller storage capacities and maximum transmission power, improving the utilization efficiency of energy storage resources and reducing energy storage costs. Furthermore, theoretically, the cost can be recovered in just 5.81 years, and there is a large profit margin for energy storage power stations that have been operating for 8 years. Investors in shared energy storage power stations can also obtain considerable income.

[0067] To further analyze the impact of introducing multiple energy sources on the performance of the shared energy storage system, scenarios 3 and 4 are compared. The optimization results for charging and discharging of the shared energy storage power station and the optimization results for charging and discharging of the electric-cooled hybrid shared energy storage are as follows: Figure 6 and Figure 7 As shown. By Figure 6 It can be seen that the maximum charging power of shared energy storage power stations typically occurs in the early morning hours during typical days in spring, autumn, and winter, while it shifts to the afternoon during typical summer days. The maximum discharging power is generally concentrated in the peak load range of 19:00–20:00 at night, indicating that the system needs to complete a full charge-discharge cycle on each typical day, demonstrating good diurnal cycle regulation capability.

[0068] Further analysis Figure 7 The operational characteristics of the hybrid energy storage system (electricity-cooling) exhibit a more complex spatiotemporal distribution of electrical power compared to energy storage systems using only electrical energy. Specifically, the peak time of maximum charging power shows significant seasonal differences: it occurs in the early morning during autumn and winter, while shifting to the afternoon during spring and summer; while the maximum discharging power remains concentrated during peak nighttime load periods, exhibiting a response pattern similar to that of a single system. This difference is mainly due to the coupling effect of cold and electrical energy flows within the system and the operational mode shift brought about by the participation of LNG cold energy. Furthermore, the introduction of cold energy storage and LNG refrigeration units reduces the amplitude of single charge and discharge power in the hybrid system, reflecting the smoothing effect of energy diversification on system operating power.

[0069] The seasonal fluctuations in cooling power are significantly greater than those in electrical power, primarily because cooling load is more sensitive to changes in ambient temperature. Nevertheless, the hybrid electric-cooling system still exhibits a complete daily charge-discharge cycle on a typical spring day, demonstrating the system's stable operation and good dispatch adaptability under multi-energy coupling conditions.

[0070] Electricity-cooling hybrid energy storage for microgrids, such as energy purchase and sale Figure 8 As shown. In Figure 8 As shown in Figure a, the electricity trading activities of each unit exhibit significant differences in operational characteristics. Microgrid 1, benefiting from its superior renewable energy endowment and low self-consumption load, maintains a continuous surplus of electricity throughout the 24 hours of the day, thus serving as the primary energy seller to the shared energy storage station. Microgrid 2, with photovoltaic units as its main power generation unit, exhibits typical intermittent and periodic characteristics due to its output being constrained by sunlight conditions: it only sells energy to the energy storage station during the daytime when sunlight is abundant, while becoming an energy buyer during the night and early morning when there is no photovoltaic output, demonstrating its adaptability to renewable energy fluctuations. Microgrid 3's electricity trading behavior exhibits typical "peak shaving and valley filling" characteristics, purchasing energy from the energy storage station during peak system demand periods to alleviate power supply pressure, and selling energy to the energy storage station during off-peak periods, effectively participating in system balancing and regulation. Data centers, due to their continuous and stable high-energy-consumption operation, are electricity buyers throughout the day, relying on shared energy storage stations as crucial power support to ensure their reliable operation.

[0071] Figure 8 Figure b further reveals the collaborative operation mechanism of the system in terms of cold energy flow. On the one hand, the cold energy released during LNG vaporization is effectively collected and prioritized for meeting the basic cooling needs of the data center. The surplus is then transported to the shared energy storage station, realizing cross-temporal cooling and tiered utilization of cold energy, significantly improving the overall energy efficiency of the system. On the other hand, other microgrids flexibly formulate cold energy management strategies based on time-of-use electricity price signals: during off-peak hours when electricity prices are low, they actively operate electric chillers to produce cold energy and store any excess beyond immediate demand in the shared energy storage station; during peak hours when electricity prices are high, they utilize the stored cold energy to meet their own load demands. This strategy not only reduces system operating costs but also forms a price-response-based "cold energy anti-peak shaving" operation mode, further enhancing the system's coordination capability and economy in handling diverse energy flows.

[0072] In summary, the hybrid electric-cooling energy storage system, by integrating multi-dimensional trading and coordinated scheduling of electric and cooling energy, achieves efficient absorption of renewable energy, flexible adaptation to load characteristics, and comprehensive improvement of system operation economy, providing an effective paradigm for the optimized operation of multi-microgrid systems in complex energy scenarios.

Claims

1. A collaborative planning method for a multi-microgrid system integrating hybrid energy storage and data center, characterized in that... Includes the following steps: Step 1: Analyze the business operation model of the cold-electric hybrid shared energy storage station, and establish a business operation and revenue model for the cold-electric hybrid shared energy storage station that covers investment, operation and maintenance, energy trading and service fees; Step 2: Couple the cold energy released during LNG vaporization with the cooling requirements of the data center to construct an integrated LNG-data center cooling model; Step 3: Based on the LNG-data center integrated cooling model constructed in Step 2, construct a collaborative operation model of a multi-microgrid system including a data center; Step 4: Combining the commercial operation and revenue model of the hybrid shared energy storage station established in Step 1, the integrated LNG-data center cooling model constructed in Step 2, and the collaborative operation model of the multi-microgrid system with data center established in Step 3, a two-layer programming model is constructed with the lowest cost of the shared energy storage station as the upper-level optimization objective and the lowest cost of collaborative operation of the multi-microgrid system and data center as the lower-level optimization objective. The two-layer programming model is then solved to obtain the optimal configuration capacity of the energy storage station and the collaborative operation strategy of the system.

2. The collaborative planning method for a multi-microgrid system integrating hybrid energy storage and data center as described in claim 1, characterized in that: In step 1, a business operation model for a shared energy storage station invested and operated by a third-party operator is designed, and a comprehensive revenue model is established that includes average daily investment and maintenance costs, typical daily energy purchase costs, energy sales revenue and service fee income, to provide storage and release services for electrical and cooling energy for a multi-microgrid user group including data centers.

3. The collaborative planning method for a multi-microgrid system integrating hybrid energy storage and data center as described in claim 2, characterized in that: The average daily investment and maintenance cost is: （1）； In formula (1): This represents the average daily investment and maintenance costs; two energy types were considered. K =1 represents electrical energy. K =2 represents cold energy; and These are the power cost and capacity cost of different energy sources for energy storage stations; and These represent the maximum charging and discharging power and maximum capacity of the energy storage station for different energy sources; The expected number of days of use for the energy storage station; Daily maintenance costs for different energy storage devices.

4. The collaborative planning method for a multi-microgrid system integrating hybrid energy storage and data center as described in claim 3, characterized in that: The energy purchase cost for each typical daily energy storage station is: （2）； In formula (2): The cost of purchasing energy for a typical daily energy storage station; N The number of microgrids; This represents the number of time periods in the scheduling cycle. for t The price at which microgrids sell energy to energy storage stations during specific time periods; For scheduling periods; For each typical day i micro-network t Energy sold to energy storage stations during specific time periods; i This indicates the number of the microgrid, where... i A value of 1 indicates data from microgrid 1; t It means 24 hours in a day. t Taking 1 means that the data is 1 point.

5. The collaborative planning method for a multi-microgrid system integrating hybrid energy storage and data center as described in claim 4, characterized in that: The revenue generated by a typical daily energy storage station from selling energy to the microgrid is: （3）； In formula (3): This represents the revenue generated by the energy storage station selling energy to the microgrid on a typical day; for t The price at which the microgrid purchases energy from the energy storage station during specific time periods; For each typical day i micro-network t Energy purchased from the energy storage station during specific time periods; The service fee charged by the microgrid for each typical daily energy storage station is: （4）； In equation (4): This represents the service fee charged by the energy storage station to the microgrid for each typical day; for t The microgrid pays a service fee per unit price to the energy storage station.

6. The collaborative planning method for a multi-microgrid system integrating hybrid energy storage and data center as described in claim 5, characterized in that: In step 2, an integrated coupled refrigeration model of LNG cold energy and data center cooling demand is established; the cold energy released during the liquefied natural gas gasification process is used to conduct two-stage heat exchange with the data center's chilled water system through the refrigerant circulation loop, thereby constructing a data center cooling system based on LNG cold energy recovery, realizing the cascade utilization and direct supply of cold energy, so as to reduce the power consumption and carbon emissions of conventional data center cooling.

7. The collaborative planning method for a multi-microgrid system integrating hybrid energy storage and data center as described in claim 6, characterized in that: The integrated coupled cooling model for LNG cold energy and data center cooling needs includes: 2.1: Theoretical Model for Cold Energy Recovery （5）； In formula (5): This indicates the cold energy released during the LNG vaporization process; This indicates that the LNG is used for actual cooling in data centers during the vaporization process. and These represent LNG at 25 and -163 Enthalpy below; Indicates the mass of LNG; , and These represent primary heat exchange efficiency, secondary heat exchange efficiency, and actual efficiency, respectively. This represents pipe cooling loss; 2.2: Data center operating costs, including server operating energy consumption and server cooling energy consumption, are expressed as follows: （6）； In formula (6): Indicates that the data center is in t Total operating energy consumption at any given time; and They represent t Real-time server operating energy consumption and server cooling energy consumption; 2.3: The total server cost is expressed as follows: （7）； In equation (7): and These represent the server's power consumption and cooling costs, respectively. express t The electricity price on the power grid at any given time; express t The data center purchases electricity from the power grid at any given time; and express t The electricity price and service fees that the Time Data Center purchases from shared energy storage stations; , and express The data center purchases cooling from shared energy storage stations at a price that includes both the purchase price and the selling price, as well as service fees. express t The data center purchases electricity from shared energy storage stations. express t The real-time data center sells cooling capacity to the shared energy storage station; express t The amount of cooling space purchased for the data center at any given time. express t The real-time data center sells cooling capacity to the shared energy storage station; Indicates the scheduling period; 2.4: In addition to using the cooling energy released during the LNG vaporization process, the data center also uses electric chiller units for auxiliary cooling. （8）； In equation (8): Indicates that the electric refrigeration unit is in t The electrical energy consumed constantly; This indicates that the server is in t The cooling energy generated by the constant-time electric refrigeration unit; This represents the ratio of electrical energy consumed by the refrigeration device to the cooling power generated.

8. The collaborative planning method for a multi-microgrid system integrating hybrid energy storage and data center as described in claim 7, characterized in that: In step 3: Based on the LNG-data center integrated cooling model constructed in step 2, a multi-microgrid system collaborative operation model containing a data center is constructed; To address the differentiated load and structural characteristics of combined cooling, heating, and power (CCHP) microgrids and data center microgrids, internal power balance constraints for electricity and cooling are established: 3.1: The power balance of the data center is shown below: （9）； In equation (9): express t The server's energy consumption at any given time; Indicates that the electric refrigeration unit is in t The electrical energy consumed constantly; For each typical day i Individual micro-networks t Power purchased from the microgrid during specific time periods; express t The data center purchases electricity from shared energy storage stations. 3.2: The cold power balance of the data center is shown below: （10）； In formula (10): This indicates that the server is in t The cooling energy generated by the constant-time electric refrigeration unit; express t The cooling energy used for actual data center cooling during the LNG vaporization process; express t The real-time data center sells cooling capacity to the shared energy storage station; express t The amount of cooling space purchased by the data center at any given time.

9. The collaborative planning method for a multi-microgrid system integrating hybrid energy storage and data center as described in claim 8, characterized in that: In step 4, the KKT conditions and the Big-M method are used to transform the bilevel programming model into a mixed-integer linear programming problem for solution, thereby obtaining the optimal configuration capacity of the energy storage station and the coordinated operation strategy of the system; specifically as follows: The upper-level optimization objective is to minimize the annual operating cost of the energy storage station, namely: （11）； In equation (11): This represents the annual operating cost of a shared energy storage station. W The number of typical days; For the first w The number of days corresponding to a typical day; This represents the average daily investment and maintenance cost of the energy storage station. The cost of purchasing electricity from the microgrid for each typical day's energy storage station; The cost of selling electricity to the microgrid from each typical daily energy storage station; Service fee for each typical day of energy storage station; The lower-level optimization objective is to minimize the annual operating cost of combined cooling, heating, and power (CCHP) multi-microgrid systems and data centers based on energy storage stations, namely: （12）； （13）； （14）； In the above formula: This indicates the annual operating cost of multi-micronet systems and data centers; This represents the total cost of electricity purchased by the data center from the shared energy storage station; The cost of purchasing electricity from the grid for each typical day; for t Electricity purchase price for a given period of time; For each typical day i Individual micro-networks t Power purchased from the grid during the specified time period; Total gas cost for each typical day; for t The price per unit volume of natural gas at any given time; For each typical day i Individual micro-networks t Output power of the gas turbine during a given period; The power generation efficiency of micro gas turbines; This refers to the calorific value of the gas. For each typical day i Individual micro-networks t The output thermal power of the gas-fired boiler during a specific time period; The efficiency of the gas-fired boiler.

10. The collaborative planning method for a multi-microgrid system integrating hybrid energy storage and data center as described in claim 9, characterized in that: It also includes constraints: ①. Energy storage station rate constraint: The capacity and rated power of an energy storage station are directly proportional, specifically: （15）； In equation (15): Indicates the capacity of the shared energy storage station; Energy ratio of the energy storage station; Indicates the rated power of the shared energy storage station; ②. Constraints on the state of charge and charging / discharging power of energy storage stations: （16）； （17）； In the above formula: Indicates shared energy storage station t The storage capacity of the Kth energy source during a given time period; Indicates shared energy storage station t The storage capacity of the Kth energy source during time period -1; and The charging and discharging efficiency of energy storage devices; and for t Real-time charging and discharging power of the energy storage station; To initially store energy for the energy storage station; The energy stored in the energy storage station; and These represent the charging and discharging status bits of the energy storage station, and are 0-1 variables. ③. Electrical balance constraints: （18）； In formula (18): Indicates the first i Individual micro-networks t Wind power output during certain periods; This means the first i Individual micro-networks t Solar power output during specific time periods; Indicates the number of typical days i Individual micro-networks t Power purchased from the grid during the specified time period; This represents the energy storage station's capacity per typical day. t The time period starts from the i The electricity purchase capacity of each microgrid; For each typical day i Individual micro-networks t Power consumption of the time-limited electric chiller; For the first i Individual micro-networks t Electrical load power during a given time period; ④. Cold balance constraint: （19）； In equation (19): Indicates the first i Microgrid electric chillers in t Power consumption during a given time period; The energy efficiency ratio of an electric chiller; For each typical day i Individual micro-networks t Output cooling power of a time-period absorption chiller; For each typical day, the i-th micronet in t Cooling load power during certain periods; ⑤. Hydrogen balance power: （20）； In equation (20): Indicates that the electrolytic cell equipment is in t Hydrogen gas produced during the period; Indicates the first i Individual micro-networks t Hydrogen load power during the time period; ⑥. Thermal equilibrium power: （21）； In equation (21): For each typical day i Individual micro-networks t The output heating power of the time-period heat exchanger; For each typical day i Individual micro-networks t Heat load power during a given period; ⑦ Waste heat balance of waste heat boiler: （22）； In equation (22): The heat exchange efficiency of the heat exchange device; The energy efficiency ratio of an absorption chiller; The heat-to-electric ratio of a gas turbine; The efficiency of the waste heat boiler; ⑧. Power balance constraints for charging and discharging of shared energy storage stations: The energy purchase and sale power of each combined cooling, heating and power microgrid and shared energy storage station, as well as the sum of the energy purchase and sale power of the data center and shared energy storage station, constitute the charging and discharging power of the energy storage station. （23）； 9. Output limits of microgrid devices: （24）； In equation (24): and These represent the minimum and maximum output of the gas turbine; and These represent the minimum and maximum output values ​​of the absorption chiller. and These represent the minimum and maximum output of the electric chiller. and These represent the minimum and maximum output values ​​of the gas-fired boiler. and These represent the minimum and maximum output values ​​of the heat exchanger. and These represent the minimum and maximum output values ​​of the electrolytic cell; ⑩. Power grid purchase constraints: （25）； In equation (25): This represents the maximum power that the microgrid purchases from the grid during that period; Transmission constraints of shared energy storage stations: （26）； In equation (26): This represents the maximum interactive power between the microgrid and the energy storage station. and Each typical day i The status of energy purchase and sale between microgrids and energy storage stations.

Images