A multi-scenario multi-energy dynamic optimization regulation method based on a complex energy system

Abstract

This invention relates to a multi-scenario, multi-energy dynamic optimization and control method for complex energy systems, comprising the following steps: S1, acquiring production plans and production data; S2, preprocessing the acquired production plans and production data; S3, inputting the preprocessed data into a production scenario identification model to determine the corresponding production scenario; S4, matching the determined production scenario with the corresponding multi-energy coupled optimization scheduling model; S5, formulating and implementing a scheduling scheme based on the corresponding multi-energy coupled optimization scheduling model. The control method provided by this invention not only improves the economic efficiency of enterprises, but its operational logic also better meets the actual production needs under multiple scenarios.

Description

Technical Field

[0001] This invention belongs to the field of energy optimization and scheduling technology, and in particular relates to a multi-scenario, multi-energy dynamic optimization and control method based on complex energy systems. Background Technology

[0002] In my country, the steel industry is a pillar industry of the national economy, a resource- and energy-intensive sector, and a typical high-energy-consuming and high-emission industry. The steel production process involves the generation and consumption of large amounts of energy, which is coupled together through various energy conversion devices and is difficult to manage and store. In actual production, situations such as gas release and steam release occur frequently, which not only wastes energy resources but also pollutes the environment. Therefore, multi-energy coupling and optimized scheduling is of great significance.

[0003] Steel companies utilize a wide variety of energy sources, generating substantial amounts of secondary energy, such as byproducts like coal gas, steam, and electricity, throughout their production processes. These energy sources are converted through energy conversion equipment within the company's energy system, such as boilers and steam turbines. Different types and quantities of energy flow through these conversion devices during the process, resulting in strong coupling between the energy sources. This makes systematic modeling of the energy and equipment within the system extremely difficult.

[0004] Chinese patent CN108490904B, "An Energy System Optimization Scheduling Method Based on Multi-Condition Equipment Operation," provides an energy system optimization scheduling method. This patent obtains key information such as the network topology of a steel enterprise's energy system, the operating parameters of key scheduling equipment, and energy production and consumption data. Combined with the operating constraints under current production conditions, a mixed-integer nonlinear programming (MINLP) model is established with the optimal economic operating cost as the objective function. By solving this model, an optimal allocation scheme for multiple energy media is obtained, overcoming the shortcomings of single-energy-media regulation and comprehensively improving the system's energy efficiency and economic benefits.

[0005] The optimization model established in Chinese patent "CN108490904B: An Energy System Optimization and Scheduling Method Based on Multi-Condition Equipment Operation" only considers equipment operating conditions and does not divide the various production scenarios in steel enterprises at the plant-wide level. In the actual production process of steel enterprises, there are multiple production scenarios, each containing various equipment operating conditions and different energy consumption rules. The working status of production or energy equipment in each scenario is always closely related to the enterprise's production or maintenance plan. Typical scenarios include normal production according to plan, blast furnace shutdown, blast furnace restart, boiler maintenance, generator unit shutdown, and sudden equipment failure. In normal production scenarios, the energy consumption of equipment is relatively stable. However, in abnormal scenarios, such as boiler maintenance or failure, the energy consumption of equipment will change significantly. These changes in production scenarios will have a significant impact on the energy system allocation strategy. Therefore, energy system optimization needs to consider factors from multiple production scenarios. Summary of the Invention

[0006] (a) Technical problems to be solved

[0007] To address the aforementioned problems in existing technologies, this invention provides a multi-scenario, multi-energy dynamic optimization and control method based on complex energy systems, which solves the technical problems of existing optimization scheduling methods, such as large limitations and low reliability.

[0008] (II) Technical Solution

[0009] To achieve the above objectives, the main technical solutions adopted by the present invention include:

[0010] A multi-scenario, multi-energy dynamic optimization and control method for complex energy systems includes the following steps:

[0011] S1. Obtain production plans and production data;

[0012] S2. Preprocess the obtained production plan and production data;

[0013] S3. Input the preprocessed data into the production scene recognition model to determine the corresponding production scene;

[0014] S4. Match the corresponding multi-energy coupling optimization scheduling model according to the determined production scenario;

[0015] S5. Specify and implement the scheduling scheme according to the corresponding multi-energy coupling optimization scheduling model.

[0016] Preferably, step S1 further includes: obtaining energy production and consumption data of the enterprise's energy topology network structure and energy system through the enterprise integrated data integration platform server, and determining the schedulable equipment;

[0017] Obtain design information and scheduling parameters for schedulable equipment, as well as production and maintenance plans within the scheduling cycle;

[0018] The historical energy production and consumption data include at least: boiler gas consumption, boiler steam production, generator steam intake and extraction, generator power generation, real-time gas holder position and gas release, and gas generation.

[0019] Preferably, S2 further includes:

[0020] The collected real-time data is normalized to linearly map the data to the range of 0-1. The transformation function is expressed as follows:

[0021]

[0022] Where x represents the collected real-time data, min represents the minimum value in the dataset, and max represents the maximum value in the dataset.

[0023] Preferably, S3 further includes:

[0024] The preprocessed data is imported into the scene recognition model to perform scene recognition based on the characteristics of the data.

[0025] Scene recognition models are divided into planned scene recognition models and abnormal scene recognition models:

[0026] When production is carried out according to plan, the production scenario is determined based on the planning model;

[0027] When an abnormal production scenario occurs, the scenario is identified based on the abnormal scenario identification model.

[0028] Preferably, S4 further includes:

[0029] The output results of the production scene recognition model are labeled accordingly and stored in the database;

[0030] By using pattern matching, tags and production scenarios are paired up, and a multi-energy coupling optimization scheduling model is selected for the corresponding production scenario.

[0031] Preferably, in step S401, the real-time energy medium production and consumption data of the enterprise's energy system are input into the scene recognition model, and the production scene recognition model outputs corresponding identification labels based on the recognition results;

[0032] S402. Input the production scenario identifier into the production scenario library for matching. If the matching is successful, select the multi-energy coupling optimization scheduling model under the corresponding production scenario. If the matching fails, update the production scenario library.

[0033] Preferably, the scene recognition model uses the CH coefficient to evaluate the abnormal scene recognition model. A higher Calinski-Harabaz score S indicates better clustering performance. The formula for calculating S is as follows:

[0034]

[0035] In formula (2), tr represents the trace of the matrix, and B k W is the covariance matrix between categories. k Let be the covariance matrix of the data within each category, m be the number of samples in the training set, and k be the number of target categories.

[0036] Preferably, before step S1, the method further includes: establishing a multi-energy coupling optimization scheduling model corresponding to each production scenario;

[0037] Includes the following steps:

[0038] Obtain the production and consumption data of various energy media in the enterprise's energy system, determine the input and output of each energy conversion device, as well as the thermodynamic properties of each energy medium, and calculate the energy conversion efficiency of the energy device through the positive balance thermal efficiency calculation method;

[0039] Based on the correspondence between the energy conversion efficiency and the operating load of each energy conversion device within a certain period, the operating load of the energy conversion device and the energy conversion efficiency under the corresponding load are characterized as quadratic functions by the method of quadratic curve fitting.

[0040] Obtain the design parameters of various energy equipment in different production scenarios, and establish a multi-energy medium coupling optimization scheduling model for each production scenario based on the energy usage price, energy supply and demand constraints, and energy equipment constraints of the multi-energy system during the scheduling cycle.

[0041] Preferably, the multi-energy coupling optimization control model includes an objective function and constraints;

[0042] The objective function of the model is expressed as:

[0043]

[0044] In formula (3), T is the number of time periods contained in a scheduling cycle, k represents the number of fuel types, and C k Let F represent the price of the k-th fuel, B represent the number of dispatchable boilers, and F represent the price of the k-th fuel. bi,k,t C represents the consumption of the k-th type of fuel in boiler bi during time period t. bp,t P represents the amount of electricity purchased during time period t. bp,t C represents the purchased electricity price during time period t. bi,k ΔF represents the penalty coefficient when boiler bi adjusts k types of gas.bi,k,t C represents the adjustment amount of boiler bi to k types of gas during time period t. gd,k Let F represent the penalty price for releasing the k-th type of gas. gd,k,t This represents the amount of gas released during time period t.

[0045] Preferably, the constraints of the models under different production scenarios are as follows:

[0046] Energy system energy demand constraints:

[0047]

[0048]

[0049]

[0050] Formula (4) represents where F represents the total amount of blast furnace gas or coke oven gas consumed by each boiler. release F represents the amount of blast furnace gas or coke oven gas emitted. sale F represents the amount of gas delivered. g This indicates the surplus amount of blast furnace gas or coke oven gas.

[0051] Formula (5) represents where This represents the total steam 's' produced by all boilers. F represents the total steam extracted by each steam turbine. s This indicates the demand for steam (s).

[0052] Formula (6) represents where P represents the total power generated by all generators. pur / sale This indicates purchased or transmitted electricity, and P represents the electricity demand of the energy system.

[0053] Boiler constraints:

[0054]

[0055]

[0056] F b,i,min ≤F b,i ≤F b,i,max (9)

[0057] Formula (7) represents the boiler's calorific value constraint on the mixed gas, where q k q represents the calorific value of gas k. bi,min and q bi,max These represent the upper and lower limits of the calorific value of the mixed gas consumed by each boiler bi;

[0058] Formula (8) represents the material balance constraint of a fuel boiler, indicating the relationship between boiler feedwater, steam production, and blowdown, where F b,fw F represents the boiler's feedwater volume. b,s F represents the amount of steam s produced by the boiler. b,sew Indicates the boiler's blowdown volume;

[0059] Formula (9) represents the upper and lower limits of boiler gas consumption, water supply, and steam generation;

[0060] Steam turbine constraints:

[0061] F st,in,min ≤F st,in ≤F st,in,max (10)

[0062] F st,ex,min ≤F st,ex ≤F st,ex,max (11)

[0063] P st,min ≤P st ≤P st,max (12)

[0064] Formulas (10)-(12) represent the upper and lower limits of the turbine inlet steam flow, turbine exhaust steam flow, and turbine extraction steam flow;

[0065] Gas holder storage constraints:

[0066]

[0067]

[0068] Formula (13) represents the maximum and minimum allowable capacity of gas stored in gas holder i;

[0069] Formula (14) represents the maximum allowable throughput of gas holder i per unit time, ΔF i gasmax This indicates the maximum allowable range of change for gas holder i within a unit of time.

[0070] Equipment operation penalty coefficient settings:

[0071] P i >P j >P k (15)

[0072] C i,g <C j,g <C k,g (16)

[0073] For high-priority equipment, a smaller penalty coefficient is set for fuel load fluctuations. When energy medium production and consumption fluctuate, these equipment are adjusted first, with a larger adjustment range. The priority and penalty coefficients for the three types of equipment are P, respectively. i P j P k and C i,g C j,g C k,g As shown in formulas (15) and (16).

[0074] (III) Beneficial Effects

[0075] The beneficial effects of this invention are:

[0076] This application takes into account the equipment operating characteristics and energy balance constraints under various production scenarios. By using this invention, in normal production scenarios, frequent equipment adjustments can be effectively avoided, ensuring stable system operation. In the scenario of blast furnace shutdown, the system raises the gas holder to the high position in advance before shutdown, while prioritizing the use of by-product gas by high-efficiency units and increasing the purchased electricity to ensure that the electricity demand is met. In the scenario of power generation equipment maintenance, the gas holder is lowered to the low position in advance to reduce the amount of gas released during the maintenance cycle.

[0077] Compared to traditional optimization and control methods, this method can not only improve the economic benefits of enterprises, but its operational logic is also more in line with the actual production needs in multiple scenarios. Attached Figure Description

[0078] Figure 1 A flowchart of a multi-scenario, multi-energy dynamic optimization and control method for complex energy systems provided by this invention;

[0079] Figure 2 The flowchart of the abnormal production scenario identification model in the multi-scenario multi-energy dynamic optimization and control method based on complex energy systems provided by the present invention;

[0080] Figure 3 A flowchart of the production scenario model matching process for steel enterprises is provided in an embodiment of a multi-scenario, multi-energy dynamic optimization and control method based on a complex energy system provided by the present invention.

[0081] Figure 4 A schematic diagram of the energy system structure of an iron and steel enterprise in an embodiment of a multi-scenario, multi-energy dynamic optimization and control method based on a complex energy system provided by the present invention;

[0082] Figure 5 A flowchart illustrating the modeling and identification process of production scenarios in the energy system of an iron and steel enterprise, as provided in this invention, for an embodiment of a multi-scenario, multi-energy dynamic optimization and control method based on a complex energy system;

[0083] Figure 6 The flowchart of the multi-scenario multi-energy dynamic optimization and control method for the energy system of an iron and steel enterprise is provided in an embodiment of the present invention. Detailed Implementation

[0084] To better explain and facilitate understanding of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0085] like Figure 1 As shown: This embodiment discloses a multi-scenario, multi-energy dynamic optimization and control method for complex energy systems, including the following steps:

[0086] S1. Obtain production plans and production data;

[0087] S2. Preprocess the obtained production plan and production data;

[0088] S3. Input the preprocessed data into the production scene recognition model to determine the corresponding production scene;

[0089] S4. Match the corresponding multi-energy coupling optimization scheduling model according to the determined production scenario;

[0090] S5. Formulate and implement scheduling schemes based on the corresponding multi-energy coupling optimization scheduling model.

[0091] In detail, S1 also includes: obtaining energy production and consumption data of the enterprise's energy topology network structure and energy system through the enterprise integrated data integration platform server, and determining the dispatchable equipment;

[0092] Obtain design information and scheduling parameters for schedulable equipment, as well as production and maintenance plans within the scheduling cycle;

[0093] The historical energy production and consumption data include at least: boiler gas consumption, boiler steam production, generator steam intake and extraction, generator power generation, real-time gas holder position and gas release, and gas generation.

[0094] In this embodiment, S2 further includes:

[0095] The enterprise's integrated data platform server acquires production data of various equipment within the enterprise's gas, steam, and power systems, as well as production and maintenance plans for future scheduling cycles.

[0096] The collected real-time data is normalized to linearly map the data to the range of 0-1. The transformation function is expressed as follows:

[0097]

[0098] Where x represents the collected real-time data, min represents the minimum value in the dataset, and max represents the maximum value in the dataset.

[0099] In this embodiment, S3 further includes:

[0100] The preprocessed data is imported into the scene recognition model to perform scene recognition based on the characteristics of the data.

[0101] Scene recognition models are divided into planned scene recognition models and abnormal scene recognition models:

[0102] When production is carried out according to plan, the production scenario is determined based on the planning model;

[0103] When an abnormal production scenario occurs, the scenario is identified based on the abnormal scenario identification model.

[0104] like Figure 2 and Figure 5 The diagram shows the operation flowchart of the abnormal production scenario identification model of the energy system of steel enterprises and the modeling and identification flowchart of the production scenario of the energy system of steel enterprises in this embodiment.

[0105] This abnormal production scenario identification model is based on machine learning. First, a large amount of historical production data is collected and preprocessed. Then, production scenario feature templates are input into the scenario identification model for training, generating a trained model. Based on historical production data from steel enterprises, production scenario features are extracted for the energy consumption characteristics of various energy equipment and the operating characteristics of equipment under various typical production scenarios, constructing feature templates for various production scenarios.

[0106] The preprocessed test dataset is put into the trained model to test the training results of the model. If it is unsatisfactory, return to the previous step, adjust the scene feature parameters and retrain.

[0107] After passing the test, the real-time production data is imported into the trained scene recognition model, and the current production scene is determined through the scene recognition model.

[0108] It should be noted that the scene recognition model uses the CH coefficient to evaluate the abnormal scene recognition model. A higher Calinski-Harabaz score S indicates better clustering performance. The formula for calculating S is as follows:

[0109]

[0110] In formula (2), tr represents the trace of the matrix, and B k W is the covariance matrix between categories. k Let be the covariance matrix of the data within each category, m be the number of samples in the training set, and k be the number of target categories.

[0111] like Figure 3 The diagram shown is a flowchart of the steel enterprise production scenario model matching process of the present invention. In this embodiment, S4 further includes:

[0112] The output results of the production scene recognition model are labeled accordingly and stored in the database;

[0113] By using pattern matching, tags and production scenarios are paired up, and a multi-energy coupling optimization scheduling model is selected for the corresponding production scenario.

[0114] S401. Input the real-time energy medium production and consumption data of the enterprise's energy system into the scene recognition model, and the production scene recognition model outputs the corresponding identification label according to the recognition result;

[0115] S402. Input the production scenario identifier into the production scenario library for matching. If the matching is successful, select the multi-energy coupling optimization scheduling model under the corresponding production scenario. If the matching fails, update the production scenario library.

[0116] like Figure 4 The diagram shows the structure of the energy system for steel enterprises according to the present invention. The design parameters of various energy equipment under each production scenario are obtained. Based on the energy production and consumption of the energy equipment in the energy system, the input and output variable sets are determined. The efficiency characteristic equation of the energy equipment is fitted by combining the equipment process model and historical production and consumption data. At the same time, considering the energy medium usage price, energy supply and demand constraints and energy equipment constraints in the scheduling cycle of the energy system, a multi-energy coupling optimization scheduling model is established for each production scenario.

[0117] In detail, the multi-energy coupled optimization scheduling model for each production scenario is established, including the following steps:

[0118] Obtain the production and consumption data of various energy media in the enterprise's energy system, determine the input and output of each energy conversion device, as well as the thermodynamic properties of each energy medium, and calculate the energy conversion efficiency of the energy device through the positive balance thermal efficiency calculation method;

[0119] Based on the correspondence between the energy conversion efficiency and the operating load of each energy conversion device within a certain period, the operating load of the energy conversion device and the energy conversion efficiency under the corresponding load are characterized as quadratic functions by the method of quadratic curve fitting.

[0120] Obtain the design parameters of various energy equipment in different production scenarios, and establish a multi-energy medium coupling optimization scheduling model for each production scenario based on the energy usage price, energy supply and demand constraints, and energy equipment constraints of the multi-energy system during the scheduling cycle.

[0121] The multi-energy coupling optimization control model described in this embodiment includes an objective function and constraints;

[0122] The objective function of the model is expressed as:

[0123]

[0124] In formula (3), T is the number of time periods contained in a scheduling cycle, k represents the number of fuel types, and C k Let F represent the price of the k-th fuel, B represent the number of dispatchable boilers, and F represent the price of the k-th fuel. bi,k,t C represents the consumption of the k-th type of fuel in boiler bi during time period t. bp,t P represents the amount of electricity purchased during time period t. bp,t C represents the purchased electricity price during time period t. bi,k ΔF represents the penalty coefficient when boiler bi adjusts k types of gas. bi,k,t C represents the adjustment amount of boiler bi to k types of gas during time period t. gd,k Let F represent the penalty price for releasing the k-th type of gas. gd,k,t This represents the amount of gas released during time period t.

[0125] The constraints of the models under different production scenarios in this embodiment are as follows:

[0126] Energy system energy demand constraints:

[0127]

[0128]

[0129]

[0130] Formula (4) represents where F represents the total amount of blast furnace gas or coke oven gas consumed by each boiler. release F represents the amount of blast furnace gas or coke oven gas emitted. sale F represents the amount of gas delivered. g This indicates the surplus amount of blast furnace gas or coke oven gas.

[0131] Formula (5) represents where This represents the total steam 's' produced by all boilers. F represents the total steam extracted by each steam turbine. s This indicates the demand for steam (s).

[0132] Formula (6) represents where P represents the total power generated by all generators. pur / sale This indicates purchased or transmitted electricity, and P represents the electricity demand of the energy system.

[0133] Boiler constraints:

[0134]

[0135]

[0136] F b,i,min ≤F b,i ≤F b,i,max (9)

[0137] Formula (7) represents the boiler's calorific value constraint on the mixed gas, where q k q represents the calorific value of gas k. bi,min and q bi,max These represent the upper and lower limits of the calorific value of the mixed gas consumed by each boiler bi;

[0138] Formula (8) represents the material balance constraint of a fuel boiler, indicating the relationship between boiler feedwater, steam production, and blowdown, where F b,fw F represents the boiler's feedwater volume. b,s F represents the amount of steam s produced by the boiler. b,sew Indicates the boiler's blowdown volume;

[0139] Formula (9) represents the upper and lower limits of boiler gas consumption, water supply, and steam generation;

[0140] Steam turbine constraints:

[0141] F st,in,min ≤F st,in ≤F st,in,max (10)

[0142] F st,ex,min ≤F st,ex ≤F st,ex,max (11)

[0143] P st,min ≤P st ≤P st,max (12)

[0144] Formulas (10)-(12) represent the upper and lower limits of the turbine inlet steam flow, turbine exhaust steam flow, and turbine extraction steam flow;

[0145] Gas holder storage constraints:

[0146]

[0147] -ΔF i gasmax ≤H i,t -H i,t-1 ≤ΔF i gasmax (14)

[0148] Formula (13) represents the maximum and minimum allowable capacity of gas stored in gas holder i;

[0149] Formula (14) represents the maximum allowable throughput of gas holder i per unit time, ΔF i gasmax This indicates the maximum allowable range of change for gas holder i within a unit of time.

[0150] Equipment operation penalty coefficient settings:

[0151] P i >P j >P k (15)

[0152] C i,g <C j,g <C k,g (16)

[0153] For high-priority equipment, a smaller penalty coefficient is set for fuel load fluctuations. When energy medium production and consumption fluctuate, these equipment are adjusted first, with a larger adjustment range. The priority and penalty coefficients for the three types of equipment are P, respectively. i P j P k and C i,g C j,g C k,g As shown in formulas (15) and (16).

[0154] In different production scenarios, various energy production equipment, energy conversion equipment, and energy storage equipment will be adjusted accordingly. At this time, the model constraints will be adjusted accordingly to adapt to the equipment operation status of the current production scenario. For example, in the blast furnace shutdown production scenario, the operating status, quantity, and priority of power generation equipment will be adjusted accordingly, as will the optimal operating position of the gas holder and the maximum allowable throughput per unit time.

[0155] like Figure 6The diagram shows the flowchart of the multi-scenario, multi-energy dynamic optimization and control method for the energy system of a steel enterprise in this embodiment. Based on scenario identification, a mathematical model for the corresponding scenario is selected. A mixed-integer nonlinear programming (MINLP) model with the objective of optimizing economic cost is established in the Python environment using the Pyomo software package. The ipopt solver is called to solve the model, obtaining the optimized allocation results of various energy media on various energy devices in the production scenario. The Pyomo software package refers to a Python optimization modeling tool, which is an open-source software package based on Python that supports the calculation and analysis of various optimization models. The debugged coupled optimization scheduling model is deployed on the enterprise's internal energy scheduling server. Based on the multi-energy scheduling scheme given by the model, the enterprise's energy management personnel make corresponding adjustments to the multi-energy system in different production scenarios.

[0156] This embodiment considers multi-energy coupled optimization scheduling under various production scenarios in steel enterprises; it analyzes the energy supply and demand rules under various production scenarios in steel enterprises, as well as the operation of various energy equipment under typical production scenarios, and models various production scenarios. For different production scenarios, corresponding energy equipment operation constraints are set, and a multi-energy coupled optimization scheduling model is established for various production scenarios. The scheduling rules of the model are more consistent with the actual operating conditions of equipment under different production scenarios in enterprises.

[0157] A production scenario identification model is established, which is divided into a planning model and an abnormal scenario identification model. The planning model identifies production scenarios based on the company's actual production and maintenance plans, while the abnormal scenario identification model is based on machine learning to identify production scenarios. Based on the output results of the production scenario identification model, a scenario label is added to each production scenario. The scenario label is used to select the corresponding optimized scheduling model.

[0158] Taking the multi-energy system of steel enterprises as a case study, this paper uses the Python programming language to build a production scenario identification model for steel enterprises and perform coupled optimization modeling under various production scenarios. It proposes multi-scenario multi-energy coupled optimization control technology and methods to enhance the robustness of traditional models and provide necessary technical support for the digital transformation of steel enterprises.

[0159] The technical principles of the present invention have been described above with reference to specific embodiments. These descriptions are merely for explaining the principles of the invention and should not be construed as limiting the scope of protection of the invention in any way. Based on this explanation, those skilled in the art can conceive of other specific embodiments of the invention without creative effort, and these embodiments will all fall within the scope of protection of the present invention.

Claims

1. A multi-scenario, multi-energy dynamic optimization and control method for complex energy systems, characterized in that, Includes the following steps: S1. Obtain production plans and production data; S2. Preprocess the obtained production plan and production data; S3. Input the preprocessed data into the production scene recognition model to determine the corresponding production scene; S4. Match the corresponding multi-energy coupling optimization scheduling model according to the determined production scenario; S5. Formulate and implement scheduling schemes based on the corresponding multi-energy coupling optimization scheduling model; Before S1, the following steps are also included: establishing a multi-energy coupling optimization scheduling model for each production scenario; Includes the following steps: Obtain the production and consumption data of various energy media in the enterprise's energy system, determine the input and output of each energy conversion device, as well as the thermodynamic properties of each energy medium, and calculate the energy conversion efficiency of the energy device through the positive balance thermal efficiency calculation method; Based on the correspondence between the energy conversion efficiency and the operating load of each energy conversion device within a certain period, the operating load of the energy conversion device and the energy conversion efficiency under the corresponding load are characterized as quadratic functions by the method of quadratic curve fitting. Obtain the design parameters of various energy equipment in each production scenario, and establish a multi-energy medium coupling optimization scheduling model for each production scenario based on the energy usage price, energy supply and demand constraints, and energy equipment constraints of the multi-energy system during the scheduling cycle. The multi-energy coupled optimization scheduling model includes an objective function and constraints; The objective function of the model is expressed as: （3） In formula (3), T is the number of time periods contained in a scheduling cycle, and K represents the number of fuel types. C k Let B represent the price of the k-th fuel, and let B represent the number of dispatchable boilers. F bi,k,t This represents the consumption of the k-th type of fuel by boiler bi during time period t. C bp,t This represents the amount of electricity purchased during time period t. P b,t This represents the purchased electricity price within time period t. This represents the penalty coefficient when boiler bi adjusts to k types of gas. This represents the adjustment amount of boiler bi to k types of gas during time period t. C gd,k This represents the penalty price for releasing the k-th type of gas. F gd,k,t This represents the amount of gas released during time period t.

2. The method according to claim 1, characterized in that, S1 further includes: obtaining energy production and consumption data of the enterprise's energy topology network structure and energy system through the enterprise's integrated data platform server, and determining the schedulable equipment; Obtain design information and scheduling parameters for schedulable equipment, as well as production and maintenance plans within the scheduling cycle; The energy production and consumption data include at least: boiler gas consumption, boiler steam production, generator steam intake and extraction, generator power generation, real-time gas holder position and gas release, and gas generation.

3. The method according to claim 2, characterized in that, S2 further includes: The collected real-time data is normalized to linearly map the data to the range of 0-1. The transformation function is expressed as follows: （1） Where x represents the collected real-time data, min represents the minimum value in the dataset, and max represents the maximum value in the dataset.

4. The method according to claim 3, characterized in that, S3 further includes: The preprocessed data is imported into the scene recognition model to perform scene recognition based on the characteristics of the data. Scene recognition models are divided into planned scene recognition models and abnormal scene recognition models: When production is carried out according to plan, the production scenario is determined based on the planning model; When an abnormal production scenario occurs, the scenario is identified based on the abnormal scenario identification model.

5. The method according to claim 4, characterized in that, S4 further includes: The output results of the production scene recognition model are labeled accordingly and stored in the database; By using pattern matching, tags and production scenarios are paired up, and a multi-energy coupling optimization scheduling model is selected for the corresponding production scenario.

6. The method according to claim 5, characterized in that, S401. Input the real-time energy medium production and consumption data of the enterprise's energy system into the scene recognition model, and the production scene recognition model outputs the corresponding identification label according to the recognition result; S402. Input the production scenario identifier into the production scenario library for matching. If the matching is successful, select the multi-energy coupling optimization scheduling model under the corresponding production scenario. If the matching fails, update the production scenario library.

7. The method according to claim 6, characterized in that, The scene recognition model uses the CH coefficient to evaluate the abnormal scene recognition model. A higher Calinski-Harabaz score S indicates better clustering performance. The formula for calculating S is as follows: （2） In formula (2), tr represents the trace of the matrix. The covariance matrix between categories, Let be the covariance matrix of the data within each category, m be the number of samples in the training set, and k be the number of target categories.

8. The method according to claim 1, characterized in that, The constraints of the models under different production scenarios are as follows: Energy system energy demand constraints: （4） （5） （6） Formula (4) represents where This represents the total amount of blast furnace gas or coke oven gas consumed by all boilers. This indicates the amount of blast furnace gas or coke oven gas released. This indicates the amount of gas delivered. This indicates the surplus amount of blast furnace gas or coke oven gas. Formula (5) represents where This represents the total steam 's' produced by all boilers. This represents the total steam output 's' from all the steam turbines. This indicates the demand for steam (s). Formula (6) represents where This represents the total power generated by all generators. This indicates purchased or transmitted electricity. This indicates the electricity demand of the energy system; Boiler constraints: （7） （8） （9） Formula (7) represents the boiler's calorific value constraint on the mixed gas, where q k Indicates gas k calorific value, q bi,min and q bi,max These represent the upper and lower limits of the calorific value of the mixed gas consumed by boiler bi; Formula (8) represents the material balance constraint of a fuel boiler, indicating the relationship between boiler feedwater, steam production, and blowdown. F b,fw,t This indicates the boiler's water supply. This indicates the amount of steam (s) produced by the boiler. Indicates the boiler's blowdown volume; Formula (9) represents the upper and lower limits of boiler gas consumption, water supply, and steam generation; Steam turbine constraints: （10） （11） （12） Formulas (10)-(12) represent the upper and lower limits of the turbine inlet steam flow, turbine exhaust steam flow, and turbine extraction steam flow; Gas holder storage constraints: （13） （14） Formula (13) represents the maximum and minimum allowable capacity of gas stored in gas holder i; Formula (14) represents the maximum allowable throughput of gas holder i per unit time. This indicates the maximum allowable range of change for gas holder i within a unit of time. Equipment operation penalty coefficient settings: （15） （16） For high-priority equipment, a small penalty coefficient is set for fuel load fluctuations. When energy medium production and consumption fluctuate, these equipment are adjusted first, and the adjustment range is larger. The priority and penalty coefficients for the three types of equipment are P, respectively. i P j P k and , , As shown in formulas (15) and (16).

Images