A Low-Carbon Operation Method for a Multi-Station Integrated System

Abstract

This invention discloses a low-carbon operation method for a multi-station integrated system, overcoming the problem that existing methods for low-carbon operation of multi-station integrated systems do not fully consider the operational characteristics and operational entity positioning of the system. The method includes the following steps: S1: Classifying the various components of the multi-station integrated system; S2: Setting low-carbon operation goals for the multi-station integrated system; S3: Collecting and summarizing operational data and other relevant data of the multi-station integrated system, and continuously accumulating historical database data; S4: Setting low-carbon operation constraint functions for the multi-station integrated system, and digitizing the constraints using relevant data to form a constraint domain; S5: Obtaining corresponding load forecast curves and new energy power generation forecast curves based on data mining and analysis of the historical database; S6: Constructing a low-carbon operation strategy model for the multi-station integrated system; S7: Optimizing and solving the model to obtain an optimized low-carbon daytime operation strategy. This method enables the formulation of low-carbon operation strategies for multi-station integrated systems from a global perspective.

Description

Technical Field

[0001] This invention relates to the field of integrated energy system operation technology, and in particular to a low-carbon operation method for a multi-station integrated system. Background Technology

[0002] Low-carbon operation is the core construction and operation goal of multi-station integrated systems. In formulating operational strategies, it is necessary to comprehensively consider economic constraints, operational characteristic constraints, and safety constraints, taking a holistic view of grid carbon reduction and minimizing carbon emissions while ensuring system robustness. Because multi-station integrated systems involve numerous components and are deeply coupled with various energy sources and diverse factors such as climate, industrial park production activities, and local customs, the integrated energy system requires optimal scheduling of multiple energy sources. This involves maximizing the performance of each component within the integrated energy system while meeting the industrial park's production and living needs, thereby achieving optimal energy utilization efficiency. Therefore, how to formulate and optimize low-carbon operation strategies for multi-station integrated systems is fundamental to whether such systems can achieve optimal carbon reduction.

[0003] Currently, there is very little research specifically on low-carbon operation strategies for multi-station integrated systems. Related research mainly focuses on operation strategies for a single subsystem of the system or for traditional integrated energy systems. Most of these strategies are formulated from a single perspective, such as operational economy and system stability. The operational methods do not fully consider the operational characteristics of multi-station integrated systems and the positioning of the operating entities. Furthermore, some strategies and optimization methods that consider multiple indicators are difficult to implement due to the difficulty in obtaining coupled variables. In practice, the operation of multi-station integrated systems often treats each subsystem independently and operates them separately, which makes it difficult to achieve optimal results. For example, the Chinese Patent Office published an invention entitled "A Low-Carbon Operation Method for Integrated Energy Systems in Industrial Parks Based on Q-Learning" on March 21, 2021, with publication number CN 112580867A. This method includes: establishing a model of an integrated energy system in an industrial park (PIES) based on the concept of an energy hub; considering the treatment costs of PIES carbon dioxide emissions and aiming to minimize the daily operating costs of PIES, establishing a mathematical optimization model for a low-carbon economic operation strategy of PIES; establishing a Markov Decision Process (MDP) corresponding to the above optimization model, and defining the state space, action space, and reward function of the MDP problem accordingly; and solving the MDP problem using an improved Q-learning algorithm, with improvements made to initialize the Q-value table using the reward function. This invention can achieve low-carbon operation of integrated energy systems in industrial parks. However, this invention only applies to the operation strategy of traditional integrated energy systems and is not suitable for multi-station integrated systems. Summary of the Invention

[0004] The purpose of this invention is to overcome the problems existing in the prior art and provide a low-carbon operation method for a multi-station fusion system. This method formulates a low-carbon operation strategy for the multi-station fusion system from a global perspective, fully considers the characteristics of the multi-station fusion system, and enables the operation strategy to minimize the carbon emissions of the entire system while taking into account constraints.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a low-carbon operation method for a multi-station integrated system, characterized in that it includes the following steps:

[0006] S1: Classify the various components of the multi-station fusion system into two categories: carbon emission sources and carbon reduction measures;

[0007] S2: Set goals for low-carbon operation of multiple stations;

[0008] S3: Collect and summarize the operational data and other relevant data of the multi-station fusion system, and continuously accumulate the historical database;

[0009] S4: Define the low-carbon operation constraint function for the multi-station integrated system, and use relevant data to digitize the constraint data to form a constraint domain;

[0010] S5: Based on the analysis of historical databases, load and renewable energy power generation are predicted to obtain the corresponding load prediction curve and renewable energy power generation prediction curve;

[0011] S6: Construct a low-carbon operation strategy model for a multi-station integrated system;

[0012] S7: Based on the low-carbon operation target, relevant data, constraint functions and prediction curves of the multi-station fusion system obtained in steps S2, S3, S4 and S5, optimize and solve the model in step S6 to obtain the optimized low-carbon operation daytime strategy.

[0013] This invention, from the perspective of regional power grids, aims to meet the operational safety, stability, and economy requirements of multi-station integrated systems. It divides the multi-station integrated system into two parts: carbon emission sources and carbon reduction measures. It sets carbon reduction objective functions and constraint functions, and adopts a combination of prediction and robust optimization methods to formulate the final low-carbon operation strategy for the multi-station integrated system.

[0014] Preferably, in step S1, the carbon emission sources include load-type equipment such as data center stations, 5G base stations, station power, electric heating systems, electric cooling systems, and electric vehicle charging stations; and the carbon reduction methods include adjustable resources such as combined heat and power (CHP), battery energy storage stations, cold storage devices, thermal storage devices, and renewable energy power plants.

[0015] It can greatly leverage the carbon reduction capabilities of each component of the multi-station integrated system and formulate low-carbon operation strategies for the multi-station integrated system from a global perspective.

[0016] Preferably, in step S2, a multi-station integrated low-carbon operation target is set:

[0017] S2.1: Calculate the equivalent carbon emissions based on the energy consumption of the carbon emission sources described in step S1, and use it as the carbon emission benchmark value.

[0018] S2.2: Reduce the carbon emissions of the multi-station integrated system by using the carbon reduction measures in step S1, and set the ratio of carbon reduction to carbon emission benchmark value not lower than the set value as the low-carbon operation target of the multi-station integrated system.

[0019] The carbon emissions mentioned are calculated based on internationally accepted carbon emission calculation formulas and coefficients.

[0020] Preferably, in step S3, the collected and aggregated operating data of the multi-station fusion system includes, but is not limited to: energy consumption monitoring data of each node of the data center, power consumption data of 5G base stations, operating data of electric heating systems, operating data of electric cooling systems, energy consumption monitoring data of loads, operating data of gas combined heat and power systems, operating data of battery energy storage stations, operating data of cold storage devices, operating data of thermal storage devices, and power generation data of renewable energy power plants; other data includes, but is not limited to, wind power, temperature, and sunlight data.

[0021] The energy consumption monitoring data for each node of the data center includes various types of energy; the energy consumption monitoring data for the load includes energy consumption monitoring data for electric vehicle charging stations; the operation data for the gas-fired combined heat and power system includes gas monitoring data, power generation data, and cooling and heating data; the operation data for the battery energy storage station includes status monitoring data such as battery status and related operation data; the operation data for the cold storage device includes status monitoring data such as cold storage medium status and related operation data; and the operation data for the heat storage device includes status monitoring data such as heat storage medium status and related operation data.

[0022] Preferably, the constraints in step S4 include economic constraints, power constraints, carbon reduction capacity constraints, and system safety and stability constraints.

[0023] Economic constraints are the cost-benefit relationships of low-carbon operation set according to the established economic indicators of the multi-station integrated system; power balance constraints require that the low-carbon operation strategy of the multi-station integrated system must follow the law of conservation of energy; carbon reduction capacity constraints require that the low-carbon operation strategy of the multi-station integrated system must be formulated under its carbon reduction capacity constraints; system safety and stability constraints require that the low-carbon operation strategy of the multi-station integrated system must be carried out under the premise of safe and stable operation of the entire system.

[0024] Preferably, the economic constraint is expressed as follows:

[0025]

[0026] In the formula, PRF(x) is the operating revenue corresponding to variable x, PRFa-thr is the indicator threshold of the operating revenue of the multi-site fusion system, COS(x) is the operating cost corresponding to variable x, COSa-thr is the indicator threshold of the operating cost of the multi-site fusion system, and x is a variable;

[0027] The power constraint is expressed as:

[0028] ∑P in&p (x)(t)-∑P out&c (x)(t)-∑P l (x)(t)=0

[0029] In the formula, P in&p (x)(t) represents the sum of the input power and the renewable energy generation power at a certain moment under variable x, ∑P out&c (x)(t) represents the sum of output power and power consumption at a certain moment under variable x, P l (x)(t) represents the power loss at a certain moment under variable x;

[0030] The aforementioned carbon reduction capacity constraint is expressed as:

[0031]

[0032] In the formula, Let be the controllability coefficient and power of the i-th controllable system in time period t, respectively. These represent the upper and lower limits of the capacity of the i-th controllable system. These are the uphill and downhill ramp rates of the i-th controllable system, respectively.

[0033] The system security and stability constraints are expressed as follows:

[0034]

[0035] In the formula, This represents the minimum amount of backup battery power required by a data center UPS. T represents the usable energy at the minimum state of charge of the energy storage battery. idc T idc,B ΔT and ΔT represent the temperature, reference temperature, and allowable range of the data center server room at any given time, respectively. P(x) risk Let P be the probability value of a certain type of risk occurring. thr (x risk ) represents the probability threshold for the occurrence of a certain type of risk.

[0036] The requirements for low-carbon operation of multi-station integrated systems stipulate that the operating costs must be lower than predetermined threshold values, and the operating revenue must be higher than predetermined threshold values. The threshold values ​​for operating costs of multi-station integrated systems are set objectively according to the different objects. In the carbon reduction capacity constraints, for gas-fired combined cooling, heating, and power systems, the controllable capacity and ramp-up rate of electricity, cooling, and heating need to be described separately.

[0037] Preferably, in step S5, the load and renewable energy generation are predicted using an improved empirical mode decomposition method, specifically as follows:

[0038]

[0039] In the formula, Dats(t) is the historical data sequence for time period t, and Dats-pv(t) is the mean sequence of the upper and lower extreme points for time period t. Subtracting the mean envelope from the original data sequence yields a new data sequence Dats1(t). Typically, this process needs to be repeated multiple times until the obtained data sequence is an intrinsic mode function component (IMF). sk1 (t) up to;

[0040] Subtract the IMF from the original data sequence sk1 After (t), repeat the above interpolation process. After repeated iterations, until a difference data sequence that cannot be further decomposed is obtained, the sequence at this time can represent the mean or range of the prediction curve.

[0041] The forecasting work in this invention is based on a continuously accumulating historical database. Since both load and renewable energy generation are continuously changing data, an improved empirical mode decomposition method is used for forecasting. By using the improved empirical mode decomposition method for forecasting and scenario selection, the number of scenarios and computational difficulty can be greatly reduced without affecting the accuracy and robustness of the solution.

[0042] Preferably, in step S6, the constructed low-carbon operation strategy model for the multi-station fusion system is a mathematical model that combines constraints, operational objectives, and prediction results, and is made easy to solve through robust optimization methods, specifically:

[0043]

[0044] After robust optimization, it is as follows:

[0045]

[0046] In the formula, F er (δ,t), F it(δ,t) are the worst-case inverse cumulative distribution functions corresponding to renewable energy output and load forecasting errors, respectively, where δ is the acceptable probability.

[0047] A combination of prediction and robust optimization methods was adopted to formulate the final low-carbon operation strategy for the multi-station fusion system, which meets the requirements of the system's operational safety, stability, and economy.

[0048] Preferably, in step S7, the strategy solution method adopts the particle swarm optimization algorithm, specifically as follows:

[0049]

[0050] In the formula, v ij x ij Let c1 and c2 be the search velocity and position of the i-th variable in the j-th dimension, respectively; c1 and c2 are acceleration constants, which are non-negative; P ij P represents the local optimal position of the i-th variable in the j-th dimension. Gj The position of the global optimal solution in the j-th dimension; rand is a random number between [0,1].

[0051] The particle swarm optimization algorithm is used to solve the model. It has a fast solution speed and high solution accuracy. The resulting optimization results of the intraday low-carbon operation strategy are reasonable and have different robustness depending on the parameter settings.

[0052] Preferably, the specific solution process in step S7 is as follows:

[0053] S7.1: Input relevant parameters of the multi-station integrated system, including parameters related to the objective function and constraint function, system parameters, load and renewable energy output forecast curves;

[0054] S7.2: Set the model parameters σ and δ, and set the optimal solution f_0 based on the predicted values ​​of the random variables;

[0055] S7.3: Input constraints, initialize the parameters of the particle swarm including velocity, position, individual optimal value and swarm optimal value, set the current iteration number to start the particle swarm iteration, k=0, k represents the iteration number;

[0056] S7.4: Calculate the fitness values ​​of the policy variables to obtain the optimal values ​​and optimal positions of the policy variables and the population;

[0057] S7.5: Update the position and velocity of the policy variables, and update the number of iterations k = k + 1;

[0058] S7.6: Repeat S7.4 and S7.5, iterating until the iteration termination condition is met.

[0059] In S7.4, it's important to note that if the position of a policy variable exceeds the set value range in one dimension, that position must be removed. The fitness of this policy variable can be set to positive infinity. After this setting, the particle's position will never be the individual optimal position, and therefore, it is even less likely to become the swarm optimal position. In S7.5, the update of the policy variable's position and velocity can refer to the historical value position of the corresponding policy variable. In S7.6, the iteration termination condition is whether the target value is less than a set threshold and the number of iterations reaches a set value. If the termination condition is met, the iteration terminates and the result is output.

[0060] Therefore, this invention has the following beneficial effects: 1. It can maximize the carbon reduction effect of each component of the multi-station fusion system and formulate a low-carbon operation strategy for the multi-station fusion system from a global perspective; 2. It fully considers the characteristics of the multi-station fusion system itself, so that the operation strategy can minimize the carbon emissions of the entire system under the conditions of taking into account economic constraints, power constraints, carbon reduction capacity constraints, and safety and stability constraints; 3. By using the empirical mode decomposition method for prediction and scenario selection, the number of scenarios and computational difficulty can be greatly reduced without affecting the accuracy and robustness of the solution; 4. The particle swarm optimization algorithm is used to solve the model, which has a fast solution speed and high solution accuracy. The resulting intraday low-carbon operation strategy optimization results are reasonable and have different robustnesses depending on the parameter settings. Attached Figure Description

[0061] Figure 1 This is a flowchart illustrating the specific operation of the method of the present invention;

[0062] Figure 2 This is a flowchart of the solution process for the low-carbon strategy model of the present invention. Detailed Implementation

[0063] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0064] like Figure 1The illustrated embodiment demonstrates a low-carbon operation method for a multi-station integrated system. The operational flow is as follows: Step 1, classifying the various components of the multi-station integrated system into two categories: carbon emission sources and carbon reduction measures; Step 2, setting low-carbon operation targets for the multi-station integrated system; Step 3, collecting and summarizing the operational data and other relevant data of the multi-station integrated system, and continuously accumulating historical database data; Step 4, setting low-carbon operation constraint functions for the multi-station integrated system, and digitizing the constraint data using relevant data to form a constraint domain; Step 5, based on the analysis of the historical database, predicting load and renewable energy power generation to obtain corresponding load prediction curves and renewable energy power generation prediction curves; Step 6, constructing a low-carbon operation strategy model for the multi-station integrated system; Step 7, optimizing and solving the model from Step 6 to obtain an optimized low-carbon daytime operation strategy. This invention, from the perspective of regional power grids, aims to meet the operational safety, stability, and economic efficiency of the multi-station integrated system. It divides the multi-station integrated system into two parts: carbon emission sources and carbon reduction measures, sets carbon reduction target functions and constraint functions, and adopts a combination of prediction and robust optimization methods to formulate the final low-carbon operation strategy for the multi-station integrated system.

[0065] The following specific examples further illustrate the technical solution and effects of the present invention. The examples below are explanations of the present invention, but the present invention is not limited to the following examples.

[0066] Step 1: Classify the various components of the multi-site fusion system

[0067] The components of the multi-station integrated system are divided into two categories: carbon emission sources and carbon reduction measures. Based on the construction status of the multi-station integrated system, load-type equipment such as data center stations, 5G base station power consumption, electric heating systems, electric cooling systems, and electric vehicle charging stations are classified as carbon emission sources of the multi-station integrated system; and adjustable resources such as gas combined heat and power, battery energy storage stations, cold storage devices, thermal storage devices, and renewable energy power plants are classified as carbon reduction measures of the multi-station integrated system.

[0068] Step 2: Setting targets for integrated low-carbon operation of multiple stations

[0069] The equivalent carbon emissions are calculated based on the energy consumption of the carbon emission sources identified in the first step, and this is used as a benchmark value for carbon emissions. The carbon emissions are calculated using internationally accepted carbon emission calculation formulas and coefficients.

[0070] Adopt appropriate operational strategies and make full use of the various carbon reduction methods mentioned in the first step to reduce the carbon emissions of the multi-station integrated system. Set the low-carbon operation target of the multi-station integrated system as follows: the ratio of carbon reduction to the carbon emission benchmark value is not lower than the set value.

[0071] Step 3: Collect and summarize the operational data and other relevant data of the multi-site fusion system, and continuously accumulate the historical database.

[0072] The data collected and compiled in this application includes all operational data and other relevant data of the multi-station fusion system, including but not limited to: energy consumption monitoring data of each node of the data center, including various types of energy; power consumption data of 5G base stations, operation data of electric heating systems, operation data of electric cooling systems, and energy consumption monitoring data of loads such as electric vehicle charging stations; operation data of gas-fired combined cooling, heating, and power systems, including gas monitoring data, power generation data, and cooling and heating data; operation data of battery energy storage stations, including status monitoring data such as battery status and related operation data; operation data of cold storage devices, including status monitoring data such as cold storage medium status and related operation data; thermal storage devices, including status monitoring data such as thermal storage medium status and related operation data; power generation data of renewable energy power plants; and meteorological data such as wind, temperature, and sunlight.

[0073] Step 4: Define the low-carbon operation constraint function for the multi-station integrated system, and digitize the constraint data using relevant data to form the constraint domain.

[0074] This application sets economic constraints, power constraints, carbon reduction capacity constraints, and system safety and stability constraints.

[0075] Among them, the economic constraint is the cost-benefit relationship of low-carbon operation set according to the established economic indicators of the multi-station integrated system. It requires that the operating cost of low-carbon operation of the multi-station integrated system must be lower than the established indicator threshold, and the operating revenue must be higher than the established indicator threshold. The specifics are as follows:

[0076]

[0077] In the formula, PRF(x) represents the operating revenue corresponding to variable x. a-thr COS(x) represents the threshold for the operational revenue of a multi-site integrated system, where COS(x) is the operating cost corresponding to variable x. a-thr The threshold value represents the operating cost of a multi-site integrated system, where x is a variable. The threshold value is set objectively based on the specific circumstances of the system.

[0078] Among these, the power balance constraint requires that the low-carbon operation strategy of multi-station integration must adhere to the law of conservation of energy. Specifically:

[0079] ∑P in&p (x)(t)-∑P out&c (x)(t)-∑P l (x)(t)=0

[0080] In the formula, P in&p (x)(t) represents the sum of the input power and the renewable energy generation power at a certain moment under variable x, ∑P out&c(x)(t) represents the sum of output power and power consumption at a certain moment under variable x, P l (x)(t) represents the power loss at a certain moment under variable x.

[0081] Among these, the carbon reduction capacity constraint requires that the low-carbon operation strategy of the multi-station integrated system be formulated under its carbon reduction capacity constraint. Specifically:

[0082]

[0083] In the formula, Let be the controllability coefficient and power of the i-th controllable system in time period t, respectively. These represent the upper and lower limits of the capacity of the i-th controllable system. These represent the uphill and downhill ramp rates of the i-th controllable system, respectively. For a combined cooling, heating, and power (CCHP) system, the controllable capacity and ramp rate of the electricity, cooling, and heating systems need to be described separately.

[0084] Among these, the system safety and stability constraint requires that the low-carbon operation strategy of multi-station integration must be implemented under the premise of the safe and stable operation of the entire system. Specifically:

[0085]

[0086] In the formula, This represents the minimum amount of backup battery power required by a data center UPS. T represents the usable energy at the minimum state of charge of the energy storage battery. idc T idc,B ΔT and ΔT represent the temperature, reference temperature, and allowable range of the data center server room at any given time, respectively. P(x) risk Let P be the probability value of a certain type of risk occurring. thr (x risk ) represents the probability threshold for the occurrence of a certain type of risk.

[0087] Step 5: Based on the analysis of historical databases, predict load and renewable energy generation to obtain corresponding load forecast curves and renewable energy generation forecast curves.

[0088] The forecasting work in this application is based on a continuously accumulating historical database. Since both load and renewable energy generation are continuously changing data, an improved empirical mode decomposition method can be used for forecasting. Specifically:

[0089]

[0090] In the formula, Dats(t) is the historical data sequence for time period t, and Dats-pv(t) is the mean sequence of the upper and lower extreme points for time period t. Subtracting the mean envelope from the original data sequence yields a new data sequence Dats1(t). Typically, this process needs to be repeated multiple times until the obtained data sequence is an intrinsic mode function component (IMF). sk1 (t) up to;

[0091] Subtract the IMF from the original data sequence sk1 After (t), repeat the above interpolation process. After repeated iterations, until a difference data sequence that cannot be further decomposed is obtained, the sequence at this time can represent the mean or range of the prediction curve.

[0092] Step 6: Construct a low-carbon operation strategy model for a multi-station integrated system

[0093] The low-carbon operation strategy model for the multi-station integrated system constructed in this application is a mathematical model that combines constraints, operational objectives, and prediction results, and is made easy to solve through robust optimization methods. Specifically:

[0094]

[0095] After robust optimization, it is as follows:

[0096]

[0097] In the formula, F er (δ,t), F it (δ,t) are the worst-case inverse cumulative distribution functions corresponding to renewable energy output and load forecasting errors, respectively, where δ is the acceptable probability.

[0098] Step 7: Optimize and solve the model from Step 6 to obtain the optimized low-carbon daytime operation strategy.

[0099] Based on the low-carbon operation target of the multi-station fusion system obtained in the second step, the relevant data of the multi-station fusion system obtained in the third step, the constraint function obtained in the fourth step, and the prediction curve obtained in the fifth step, the model obtained in the sixth step is optimized and solved to obtain the optimized low-carbon operation daytime strategy.

[0100] The strategy solution method in this application adopts the particle swarm optimization algorithm, specifically as follows:

[0101]

[0102] In the formula, v ij x ij Let c1 and c2 be the search velocity and position of the i-th variable in the j-th dimension, respectively; c1 and c2 are acceleration constants, which are non-negative; P ijP represents the local optimal position of the i-th variable in the j-th dimension. Gj The position of the global optimal solution in the j-th dimension; rand is a random number between [0,1].

[0103] The specific solution process is as follows: Figure 2 As shown:

[0104] 1) Input relevant parameters of the multi-station integrated system, including parameters related to the objective function and constraint function, as well as system parameters, and input load and renewable energy output prediction curves.

[0105] 2) Set the model target tolerance coefficient σ and acceptable probability δ, and set the optimal solution f0 based on the predicted values ​​of random variables.

[0106] 3) Input constraints, initialize parameters such as velocity, position, individual optimal value and group optimal value of the particle swarm, set the current iteration number k = 0, and start particle swarm iteration, where k represents the iteration number.

[0107] 4) Determine if the number of iterations is less than the set value. If not, end the iteration and output the result. If yes, continue to execute the subsequent steps.

[0108] 5) Generate an optimized initial population, calculate the fitness values ​​of each policy variable, and obtain the optimal values ​​of the policy variables, the population, and the optimal position f. IESIMSF It is important to note here that if the position of a certain policy variable exceeds the set value range in a certain dimension, this position should be removed. The fitness of this policy variable can be set to positive infinity. After this setting, the position of this particle will definitely not become the individual optimal position, and therefore it is even less likely to become the group optimal position.

[0109] 6) Update the position and velocity of the policy variable, with an update iteration count of k = k + 1. The update of the position and velocity of the policy variable can be referenced from the historical value position of the corresponding policy variable.

[0110] 7) Determine the termination condition of the iteration, i.e., the objective value f IESIMSF Is it less than or equal to the set threshold f? IESIMSF If the result is max, the iteration terminates and the result is output; otherwise, proceed to step 4) and repeat steps 4)-7).

[0111] This invention can significantly leverage the carbon reduction capabilities of each component in a multi-station fusion system. It enables the formulation of low-carbon operation strategies for multi-station fusion systems from a global perspective, fully considering the unique characteristics of such systems. This allows the operation strategy to minimize overall system carbon emissions while balancing economic, power, carbon reduction capacity, and safety and stability constraints. By employing an improved empirical mode decomposition method for prediction and scenario selection, the number of scenarios and computational complexity are greatly reduced without compromising the accuracy and robustness of the solution. The particle swarm optimization algorithm is used to solve the model, resulting in fast and accurate solutions. The optimized intraday low-carbon operation strategy is reasonable and exhibits varying robustness depending on the parameter settings.

[0112] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any way. Other variations and modifications are possible without departing from the technical solutions described in the claims.

Claims

1. A low-carbon operation method for a multi-station integrated system, characterized in that, It includes the following steps: S1: Classify the various components of the multi-station fusion system into two categories: carbon emission sources and carbon reduction measures; S2: Set goals for low-carbon operation of multiple stations; S3: Collect and summarize the operational data and other relevant data of the multi-station fusion system, and continuously accumulate the historical database; S4: Define the low-carbon operation constraint function for the multi-station integrated system. The constraints include power constraints and carbon reduction capacity constraints. Use relevant data to digitize the constraints and form a constraint domain. Carbon reduction capacity constraints include: the controllable power of the i-th controllable system in time period t is between the product of the lower limit of the capacity of the i-th controllable system and the controllability coefficient of the system in time period t, and the product of the upper limit of the capacity of the i-th controllable system and the controllability coefficient of the system in time period t; the difference between the controllable power of the i-th controllable system in time period t and time period t-1 is between the uphill rate and downhill rate of the i-th controllable system; S5: Based on the analysis of historical databases, load and renewable energy power generation are predicted to obtain the corresponding load prediction curve and renewable energy power generation prediction curve; S6: Construct a low-carbon operation strategy model for a multi-station integrated system. The low-carbon operation strategy model for a multi-station integrated system is a mathematical model that combines constraints, operation objectives, and prediction results. S7: Based on the low-carbon operation target, relevant data, constraint functions and prediction curves of the multi-station fusion system obtained in steps S2, S3, S4 and S5, optimize and solve the model in step S6 to obtain the optimized low-carbon operation daytime strategy.

2. The low-carbon operation method for a multi-station integrated system according to claim 1, characterized in that, In step S1, the carbon emission sources include load-type equipment such as data center stations, 5G base stations, station power, electric heating systems, electric cooling systems, and electric vehicle charging stations; the carbon reduction methods include adjustable resources such as combined gas power, battery energy storage stations, cold storage devices, thermal storage devices, and renewable energy power plants.

3. A low-carbon operation method for a multi-station integrated system according to claim 1 or 2, characterized in that, In step S2, a multi-station integrated low-carbon operation target is set: S2.1: Calculate the equivalent carbon emissions based on the energy consumption of the carbon emission sources described in step S1, and use it as the carbon emission benchmark value. S2.2: Reduce the carbon emissions of the multi-station integrated system by using the carbon reduction measures in step S1, and set the ratio of carbon reduction to carbon emission benchmark value not lower than the set value as the low-carbon operation target of the multi-station integrated system.

4. The low-carbon operation method of a multi-station integrated system according to claim 1, characterized in that, In step S3, the collected and aggregated operational data of the multi-station fusion system includes: energy consumption monitoring data of each node in the data center, power consumption data of 5G base stations, operational data of electric heating systems, operational data of electric cooling systems, energy consumption monitoring data of loads, operational data of gas combined heat and power systems, operational data of battery energy storage stations, operational data of cold storage devices, operational data of thermal storage devices, and power generation data of renewable energy power plants; other data include, but are not limited to, wind power, temperature, and solar radiation data.

5. A low-carbon operation method for a multi-station integrated system according to claim 1 or 4, characterized in that, The constraints in step S4 include economic constraints and system security and stability constraints.

6. A low-carbon operation method for a multi-station integrated system according to claim 5, characterized in that, The aforementioned economic constraints are expressed as follows: In the formula, PRF(x) represents the operating revenue corresponding to variable x. a-thr COS(x) represents the threshold for the operational revenue of a multi-site integrated system, where COS(x) is the operating cost corresponding to variable x. a-thr , where x is the threshold value for the operating cost of a multi-site fusion system; The power constraint is expressed as: ΣP in&p (x)(t)-ΣP out&c (x)(t)-ΣP l (x)(t)=0 In the formula, P in&p (x)(t) represents the sum of the input power and the renewable energy generation power at a certain moment under variable x, ΣP out&c (x)(t) represents the sum of output power and power consumption at a certain moment under variable x, P l (x)(t) represents the power loss at a certain moment under variable x; The aforementioned carbon reduction capacity constraint is expressed as: In the formula, Let be the controllability coefficient and power of the i-th controllable system in time period t, respectively. These represent the upper and lower limits of the capacity of the i-th controllable system. These are the uphill and downhill ramp rates of the i-th controllable system, respectively. The system security and stability constraints are expressed as follows: In the formula, This represents the minimum amount of backup battery power required by a data center UPS. T represents the usable energy at the minimum state of charge of the energy storage battery. idc T idc,B ΔT and ΔT represent the temperature, reference temperature, and allowable range of the data center server room at any given time, respectively. P(x) risk Let P be the probability value of a certain type of risk occurring. thr (x risk ) represents the probability threshold for the occurrence of a certain type of risk.

7. A low-carbon operation method for a multi-station integrated system according to claim 1 or 2, characterized in that, In step S5, load and renewable energy generation are predicted: an improved empirical mode decomposition method is used to predict load and renewable energy generation, specifically as follows: In the formula, Dat s (t) represents the historical data sequence for time period t, where Dat s-pv (t) is the mean sequence of the upper and lower extreme points in time period t. Subtracting the mean envelope from the original data sequence yields the new data sequence Dat. s1 (t), the above process needs to be repeated multiple times until the obtained data sequence is an intrinsic modulus function component (IMF). sk1 (t) up to; Subtract the IMF from the original data sequence sk1 After (t), repeat the above difference process, iterating repeatedly until a difference data sequence that cannot be further decomposed is obtained. At this point, the sequence can represent the mean or range of the prediction curve.

8. A low-carbon operation method for a multi-station integrated system according to claim 1, characterized in that, In step S6, the constructed low-carbon operation strategy model for the multi-station integrated system is a mathematical model that combines constraints, operational objectives, and prediction results. This model is then made easier to solve using robust optimization methods. Specifically: After robust optimization, it is as follows: In the formula, F er (δ,t), F it (δ,t) are the worst-case inverse cumulative distribution functions corresponding to renewable energy output and load forecasting errors, respectively. δ is the acceptable probability, α is the acceptable range, and f_IESIMSF is the result of the optimization calculation.

9. A low-carbon operation method for a multi-station integrated system according to claim 1, characterized in that, In step S7, the strategy solution method adopts the particle swarm optimization algorithm, specifically as follows: In the formula, v ij x ij Let c1 and c2 be the search velocity and position of the i-th variable in the j-th dimension, respectively; c1 and c2 are acceleration constants, which are non-negative; P ij P represents the local optimal position of the i-th variable in the j-th dimension. Gj The position of the global optimal solution in the j-th dimension; rand is a random number between [0,1].

10. A low-carbon operation method for a multi-station integrated system according to claim 1 or 9, characterized in that, In step S7, the specific solution process is as follows: S7.1: Input relevant parameters of the multi-station integrated system, including parameters related to the objective function and constraint function, system parameters, load and renewable energy output forecast curves; S7.2: Set the model target tolerance coefficient σ and acceptable probability δ, and set the optimal solution f0 based on the predicted values ​​of random variables; S7.3: Input constraints, initialize the parameters of the particle swarm including velocity, position, individual optimal value and swarm optimal value, set the current iteration number to start the particle swarm iteration, k=0, k represents the iteration number; S7.4: Calculate the fitness values ​​of the policy variables to obtain the optimal values ​​and optimal positions of the policy variables, the population; S7.5: Update the position and velocity of the policy variables, and update the number of iterations k = k + 1; S7.6: Repeat S7.4 and S7.5, iterating until the iteration termination condition is met.

Images