A method for identifying key factors affecting load participation in demand response of process industry and a computer device thereof

Abstract

The application discloses a kind of key factor identification method and its computer equipment influencing process industry load participation demand response, it is related to electric power system dispatching and control field, first, the coupling relationship of production link is described using state task network, to maximize the total system profit as optimization goal, build general industrial process-oriented energy optimization scheduling model.Second, facing the interaction scene of power grid and process industry production, the common influencing factors faced in production process are extracted.Based on Morris screening method, the above-mentioned large number of influencing factors are preliminarily and quickly screened, and finally, based on Sobol' global sensitivity analysis method, the global sensitivity coefficient of the influencing factors screened to the three core indexes of economy, productivity and energy efficiency is quantified, so as to identify the key factors influencing the process industry load participation demand response.

Description

Technical Field

[0001] This invention relates to the field of power system dispatching and control, specifically to a method for identifying key factors affecting the participation of process industry loads in demand response and its computer equipment. Background Technology

[0002] With the increasing integration of intermittent renewable energy sources such as wind and solar power, the uncertainty of the power system has increased significantly, and the problem of insufficient flexibility has become increasingly prominent. By regulating user-side resources and coordinating with the existing power grid, the flexibility of the power system can be effectively enhanced. Process industry loads play a crucial role in the power system due to their large capacity, high automation, and flexible adjustment capabilities in some production processes, effectively balancing grid load through peak shaving and valley filling. Statistics show that in 2023, China's industrial electricity consumption reached 5.9779 trillion kWh, accounting for 65% of total social electricity consumption, with industrial (PI) electricity consumption accounting for over 80% of industrial electricity consumption. Its enormous adjustability potential makes it one of the most important demand-side regulation resources. However, industrial users are frequently required to adjust their production plans due to factors such as energy price fluctuations and changes in operating conditions, making it difficult for industrial enterprises to flexibly adjust electricity consumption without affecting production plans, thus limiting their willingness to interact with the grid. Therefore, identifying the key factors influencing the participation of process industry loads in demand response is particularly important for characterizing the sensitivity characteristics of influencing factors in industrial production processes.

[0003] CN119362476A discloses a method, system, medium, and equipment for interactive regulation of industrial load resources and the power grid. It proposes a real-time control method for adjustable loads in industrial production processes, focusing on process links with regulation potential within specific production lines, such as electrolytic aluminum current adjustment and air compressor start-stop sequences. The method dynamically assesses the adjustability of the process chain by analyzing the load operating status and production constraints in real time, and then constructs regulation strategies to respond to power grid frequency regulation and peak shaving commands. Based on process-level state judgment and real-time calculation, it ensures that regulation does not disrupt production continuity. While this patent focuses on industrial load participation in power grid regulation, its method focuses on a single production line process and lacks modeling and analysis of the entire production process. It fails to identify key parameters through the STN model combined with sensitivity analysis.

[0004] CN117788210A presents a method, device, and medium for calculating the global sensitivity of integrated energy systems based on generalized unscented transform. This method proposes a global sensitivity calculation method for input parameters in electro-thermal coupled integrated energy systems. It employs generalized unscented transform to construct samples of input variable perturbations and uses the Sobol approach to approximate the variance contribution rate of the system output, thereby obtaining a first-order sensitivity index. The method requires the input variables to follow a Gaussian distribution and pre-sets their higher-order statistical moments. This scheme is suitable for sensitivity analysis in strongly nonlinear scenarios, and its validation is limited to electro-thermal combined systems. However, this patent only focuses on a single sensitivity index and does not perform preliminary screening of input variables. When the variable dimensionality is high (e.g., dozens of influencing factors may be involved in industrial systems), the computational cost of this method increases significantly.

[0005] Existing technologies typically focus on the macro-level interaction and economic dispatch of industrial loads and the power grid, often neglecting the dynamic coupling characteristics between industrial processes and energy supply. Furthermore, current incentive mechanisms lack a balance between production and response objectives, failing to effectively coordinate the conflicting interests between industrial users and the power grid. In addition, frequent occurrences of influencing factors during production, such as energy price fluctuations and equipment start-up and shutdown failures, make it difficult for industrial enterprises to flexibly respond to power grid dispatch while ensuring the stable operation of key processes. Although global sensitivity analysis has been applied in many fields, its in-depth application in integrated energy system optimization and industrial production dispatch is not yet widespread, limiting the accuracy and practicality of dispatch parameter identification. Summary of the Invention

[0006] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0007] To address the aforementioned technical problems, this invention provides the following technical solution: a method for identifying key factors affecting the participation of process industry load in demand response, comprising:

[0008] Step S1: Based on the energy supply structure and production process of industrial users, construct an industrial integrated energy system architecture that includes an integrated energy system and a state task network.

[0009] Step S2: For various influencing parameters in the production process (assuming they follow a normal distribution), a small-scale input sample set is generated through random sampling. Based on the industrial integrated energy system architecture, an energy optimization scheduling model for general industrial processes is constructed, optimized with profit maximization as the objective function, and the optimized industrial load production index is extracted as the output sample.

[0010] Step S3: Define the influencing factors and comprehensive performance evaluation indicators of the process industry's load production process.

[0011] Step S4: Calculate the basic effects of each input parameter based on the Morris method to perform preliminary screening of the influencing parameters.

[0012] Step S5: Quantify the global sensitivity index of the parameters after screening based on the Sobol' method, and clarify the influence weight of each parameter on the production plan output results.

[0013] As a preferred embodiment of the method for identifying key factors affecting the participation of process industry load in demand response according to the present invention, the industrial integrated energy system in step S1 is analyzed, wherein:

[0014] Industrial integrated energy systems integrate multiple energy forms such as electricity, heat, cooling, and gas, leveraging the synergistic and complementary effects between heterogeneous energy flows to achieve coordinated optimization of energy allocation and production scheduling. This effectively meets the demands of complex industrial production and realizes dynamic coupling between the energy system and industrial processes. Typical industrial integrated energy systems are widely used in energy-intensive industries such as steel, chemicals, and electrolytic aluminum. The structure of an industrial integrated energy system includes an integrated energy system architecture and a State Task Network (STN) architecture. Therefore, this invention constructs an STN representation considering multi-energy coupling to describe the production process and optimize energy consumption.

[0015] The integrated energy system architecture is supply-side, primarily composed of three units: energy input, conversion, and storage. The energy input unit integrates various energy sources such as distributed photovoltaic (PV), public power grid, and natural gas, prioritizing PV power generation and supplementing with grid power when insufficient, ensuring stable and economical power supply. The energy conversion unit, through multi-energy complementarity and synergistic optimization, meets the electricity, heat, and cooling needs of industrial users, including combined heat and power (CHP), gas-fired boilers, waste heat boilers, air conditioning systems, and absorption chillers, significantly improving energy utilization efficiency. The energy storage unit includes electrical energy storage systems and thermal storage tanks, used to balance supply and demand fluctuations. Electrical energy storage systems store surplus electricity and release it during peak periods, while thermal storage tanks store waste heat or cooling energy to meet energy demands during specific periods.

[0016] The state-task network structure diagram, representing the demand side, illustrates the characteristics of a typical industrial manufacturing process. During production, a buffer serves as a storage area between every two stages. As raw materials, As an intermediate product, For the final product. If Production volume is higher than Consumption levels will be monitored, and excess products will be stored in an intermediate buffer. The goal of production planning is to complete the production of a specified quantity of products within a cycle and follow a strict production sequence, combining industrial production and energy supply, flexibly arranging the operation of production equipment, effectively reducing operating costs, and achieving interaction with the power grid.

[0017] As a preferred embodiment of the method for identifying key factors affecting the participation of process industry load in demand response according to the present invention, step S2 is specifically analyzed, wherein:

[0018] S2.1 Define that the distribution of parameters affecting the production process in the process industry follows a normal distribution. A small-scale input sample set is generated by random sampling, and each parameter follows a different parameter range according to the actual situation.

[0019] S2.2 Based on the state-task network structure of the industrial integrated energy system in step S1, and combined with the energy consumption characteristics of production processes, a generalized unified modeling method for industrial production processes is constructed.

[0020] S2.3 Based on the integrated energy system architecture in the industrial integrated energy system of step S1, construct the integrated energy system equipment model;

[0021] S2.4. Based on the response to the power grid interaction demand, construct an industrial production economic optimization scheduling model with the goal of maximizing the total system profit.

[0022] As a preferred embodiment of the method for identifying key factors affecting the participation of process industry load in demand response according to the present invention, in step S2.2, the expression of the industrial production process model is analyzed, wherein:

[0023] (1) Operational constraints in the production process

[0024] Production process operational constraints include production status constraints, minimum production line running time constraints, and dynamic transfer constraints. Among these, production status constraints include:

[0025]

[0026] Each link and each production line can only be in one operating state at time t.

[0027] Minimum production line run time constraints:

[0028]

[0029] Dynamic transitive constraints:

[0030]

[0031] Prevent production interruptions or material accumulation. Ensure that the operating status of upstream machines is promptly responded to by downstream machines.

[0032] in, For each link in t The time and the operating mode of each production line. For production line indexing, ; m is the production process index, ; This is a production process mode index, which refers to the operating mode of process m on production line l. This is a state where production is stopped, meaning no production activities are carried out, and both energy consumption and output are zero. This represents the standard operating state of the production process, i.e., operating at standard or medium load, with a certain output and energy consumption, meeting the needs of routine production. This refers to a high-load operation state in the production process, that is, operating at high load or maximum capacity, resulting in higher output, but also a significant increase in energy consumption. This refers to the minimum start-up time for each production line, each stage, and each production mode.

[0033] (2) Buffer balance constraint

[0034] Buffer balancing constraints include storage constraints and production order constraints. Storage constraints include:

[0035]

[0036]

[0037] The current storage quantity consists of three parts: the storage quantity of the previous period, the input quantity of the upstream process, and the consumption quantity of the current process; Production sequence constraints:

[0038]

[0039] Downstream processes can only operate if there is sufficient inventory in the buffer zone of upstream processes.

[0040] in, Let be the buffer storage capacity of each link in each production line at time t; The amount of raw materials input for each stage of each production line; Let be the production volume of each stage of each production line at time t; This refers to the output per unit time of each stage of each production line in each operating mode. , These represent the minimum and maximum storage capacities of the buffer, respectively. Upper limit of buffer capacity for each stage

[0041] (3) Production constraints

[0042] Production constraints:

[0043]

[0044] By multiplying the productivity in the current state by the state variable, the output constraints in the production process are quantified.

[0045] Final production target constraints:

[0046]

[0047] in, The total output of the final produced products is in units. The target output.

[0048] As a preferred embodiment of the method for identifying key factors affecting the participation of process industry load in demand response according to the present invention, in step S2.3, the equipment model expression of the integrated energy system is analyzed, wherein:

[0049] (1) Multi-current coupling device

[0050] 1) Combined cooling, heating and power (CCHP) unit

[0051]

[0052]

[0053]

[0054] in, , Let be the power generation and cooling power of the CCHP unit at time t, respectively, in kW; The thermal power output of the waste heat boiler is expressed in kW. This refers to the start-up and shutdown status of the CCHP unit; This refers to the gas consumption of the CCHP unit. , , and These are the cooling, heating, and electrical efficiencies and self-dissipation rate of the CCHP unit, respectively. This indicates the low calorific value of natural gas.

[0055] 2) Electric and thermal energy storage

[0056] Electric energy storage:

[0057]

[0058]

[0059]

[0060]

[0061]

[0062]

[0063] In the formula, for t Energy storage capacity at all times; , These are the charging and discharging efficiencies of electrical energy storage, respectively. These are the upper and lower limits of the energy storage capacity, respectively. , These represent the remaining electricity at the end of the energy storage dispatch and the remaining electricity at the beginning of the dispatch, respectively. They are respectively t The state variables of the battery during charging and discharging are constantly monitored. =1、 =0; during discharge =0、 =1.

[0064] Thermal energy storage

[0065]

[0066]

[0067]

[0068]

[0069]

[0070]

[0071] In the formula, for t Storing heat energy at all times; , These are the charging and releasing efficiencies of thermal energy storage, respectively. These are the upper and lower limits of the heat storage capacity for thermal energy storage; , These represent the remaining heat at the end of the thermal energy storage dispatch and the remaining heat at the beginning of the dispatch, respectively. They are respectively t The constant state variables of the heat storage device during charging and discharging, during charging. =1、 =0; when releasing heat =0、 =1;

[0072] 3) Air conditioning and absorption chillers

[0073]

[0074]

[0075]

[0076] In the formula, for t Cooling power of Shike Leng air conditioner, KW; For air conditioning cooling efficiency; for t Power consumption of a 24-hour air conditioner (in KW). , These are the cooling capacity and heat absorption capacity of the refrigeration unit, respectively. The energy conversion efficiency of the refrigeration unit; This represents the upper limit of the refrigeration capacity of the refrigerator.

[0077] 4) Gas-fired boiler

[0078]

[0079]

[0080] in, For GB in t Heating power at any given time, in KW; The amount of natural gas consumed by GB, in m³; The heating efficiency is in GB. , These represent the upper and lower boundaries of GB output, respectively.

[0081] (2) Multi-energy equilibrium constraint

[0082] 1) Power balance

[0083]

[0084] in, Let be the energy consumption of each stage in state p, expressed in kWh.

[0085] 2) Thermal energy balance

[0086]

[0087] in, Let KWh represent the thermal energy consumption of each stage under state p.

[0088] 3) Leng Nengping

[0089]

[0090] in, Let KWh represent the cold energy consumption of each stage under state p.

[0091] As a preferred embodiment of the method for identifying key factors affecting the participation of process industry load in demand response according to the present invention, in step S2.4, the industrial production economic optimization scheduling model is analyzed, wherein:

[0092] Industrial enterprises' willingness to participate in grid interaction is highly dependent on economic returns. Therefore, their optimization goal is to optimize production scheduling and energy management by maximizing total system profit while responding to grid demand.

[0093]

[0094] in, The final sales revenue of the manufactured products is expressed in yuan. The total cost of industrial production is expressed in yuan.

[0095] (1) Total Revenue

[0096]

[0097] in, To optimize cycle length; For production line indexing; The unit price is the selling price of the product, expressed in yuan.

[0098] (2) Total cost

[0099]

[0100] in, For energy purchase costs; For production and maintenance costs; For production costs; For start-up and shutdown costs; For storage costs.

[0101] Energy purchase costs consist of electricity and gas purchase costs. Industrial systems prioritize the use of their own distributed photovoltaic power, and when this power is insufficient, they purchase electricity from the external grid to meet the power needs of equipment in various production processes. Gas purchase costs include three parts: fuel costs for CCHP and GB units, expressed as follows:

[0102]

[0103] in, and These are the costs of purchasing electricity and gas, respectively. Let t be the external grid electricity price, in yuan / kWh; Let t be the power purchased from the external power grid, in KW; The price is for natural gas, in yuan / m3; and Let t represent the gas purchase amounts of CCHP and GB, respectively.

[0104] Considering that although the production line is fully automated, a small number of personnel are still required for inspection and monitoring during the start-up, shutdown and operation of the production line, there are maintenance costs for each production process.

[0105]

[0106] Where m represents the production stage; This is the operation and maintenance cost coefficient.

[0107] Production costs encompass the expenses directly related to each production process in an industrial system, mainly including raw material consumption and equipment depreciation.

[0108]

[0109] in, The unit production cost is expressed in yuan per unit. Let t represent the output of each stage of each production line, in units.

[0110] Start-up and shutdown costs refer to the additional expenses incurred during the start-up and shutdown of a production line. These mainly include increased energy consumption during equipment start-up and shutdown, as well as system debugging costs. Frequent start-up and shutdown operations may still affect the lifespan and stability of the equipment.

[0111]

[0112] in, Indicates the unit cost of the switch; and These indicate the on and off states, respectively.

[0113] Storage costs involve the expenses incurred in warehousing raw materials, intermediate products, or finished goods during the production process. In industrial systems, storage costs also need to consider the impact of inventory backlogs or stockouts on production planning, thereby optimizing inventory management to reduce overall operating costs.

[0114]

[0115] in, Cost per unit of storage.

[0116] As a preferred embodiment of the method for identifying key factors affecting the participation of process industry load in demand response according to the present invention, in step S3, the defined influencing factors and comprehensive performance evaluation indicators are analyzed, wherein:

[0117] (1) Definition of influencing factors of industrial system

[0118] This invention selects unit output production cost Unit product sales price External power grid electricity price Maximum buffer capacity Target output Start-up and shutdown costs Natural gas prices Unit inventory cost and unit production and maintenance costs Nine categories of factors serve as the primary sources of influence. These factors cover multiple core elements affecting system stability and economy, such as cost-benefit analysis during the production process, energy price fluctuations, and equipment scheduling flexibility. They influence electricity demand response by affecting economic efficiency, energy efficiency, and production flexibility.

[0119] Among these factors, unit production cost, unit product sales price, and unit production and maintenance cost reflect fluctuations on both the cost and revenue sides, directly impacting the profit margins of industrial enterprises when responding to the power grid and influencing their economic incentives to participate in the response. Fluctuations in external power grid prices, especially under time-of-use pricing mechanisms, can interfere with the cost-effectiveness of enterprises' off-peak electricity consumption or peak-shaving, making it difficult to optimize response strategies. Uncertainty regarding the upper limit of buffer capacity will affect material synchronization and system scheduling feasibility, and inventory constraints may hinder load transfer. Uncertainty regarding target output will cause enterprises to prioritize production plans, weakening their ability to respond to the power grid. Uncertainty regarding start-up and shutdown costs will increase the economic risk of equipment start-up and shutdown, limiting enterprises' load adjustment capabilities in demand response. Natural gas prices, as a key energy input cost factor in industrial production, especially dominating in heat supply, will affect the energy costs of multi-energy coupled systems due to price fluctuations. Changes in unit inventory costs indirectly weaken enterprises' willingness to adjust production plans to respond to the power grid by increasing storage costs. These uncertainties, through mechanisms such as cost transmission, efficiency reduction, and limited flexibility, constrain the performance of IIES in production-energy synergistic optimization, thereby affecting the willingness and ability of industrial enterprises to participate in demand response.

[0120] (2) Comprehensive performance evaluation indicators

[0121] 1) Economic indicators

[0122] Total system profit is a key indicator for measuring the economic efficiency of industrial production, and it directly reflects the system's net revenue capacity under the influence of various factors.

[0123]

[0124] in, It is the total profit of the industrial production system; and These represent the final sales revenue of the manufactured products and the total industrial production cost, respectively, in yuan.

[0125] Energy procurement costs are a key characteristic of the high energy dependence of industrial systems, especially in industrial park scenarios where fluctuations in energy prices have a significant impact on system operating costs.

[0126]

[0127] in, Energy procurement cost, in yuan; Let t be the external grid electricity price, in yuan / kWh; Let t be the power purchased from the external power grid at time t, in KW. The price is for natural gas, in yuan / m3; Let t be the amount of gas purchased by CHP. ; Let t be the gas purchase amount of the gas-fired boiler. .

[0128] 2) Energy efficiency indicators

[0129] Integrated energy utilization rate is an important indicator for evaluating the energy conversion efficiency of multi-energy complementary systems. It is defined as the ratio of the total effective energy output of the system to the total input of non-renewable energy. In regional integrated energy systems, due to the coupling and conversion of various heterogeneous energy sources such as electricity, heat, and cooling, it is necessary to accurately quantify the overall system performance.

[0130]

[0131]

[0132]

[0133] in, For comprehensive energy utilization rate; This represents the total amount of non-renewable energy consumed by the system during its operating cycle T. This refers to the total effective energy provided by the system within period T.

[0134] 3) Production flexibility indicators

[0135] Inventory levels reflect the buffering capacity of an industrial system between production and demand. Their impact can lead to excessive or insufficient inventory, which in turn affects system costs.

[0136]

[0137] in, For inventory levels, [number] units; The inventory level of each production line and each stage at the end of the scheduling process is [number of units].

[0138] Total production is a direct representation of the production capacity of an industrial system, and its impact affects the system's output capacity. As a fundamental indicator for measuring system performance, total production is closely related to inventory levels and system costs.

[0139]

[0140] in, Total production quantity, in units; The total output of the final produced products is in units.

[0141] As a preferred embodiment of the method for identifying key factors affecting the participation of process industry load in demand response according to the present invention, in step S4, the Morris-based pre-screening method is analyzed. Considering the problem of increased computational cost due to the large number of parameters, and that not all parameters have an equally important impact on the model output, the Morris method is first used to quickly screen out the parameters with the most significant impact.

[0142] The Morris method assesses the impact of each input parameter on the model output by calculating the elementary effect (EE) of each input parameter. For model parameters... and the i-th parameter Its basic effect It can be represented as:

[0143]

[0144] in, The predefined step size reflects the magnitude of parameter perturbation. To comprehensively explore the global characteristics of the parameter space, the Morris method randomly generates multiple trajectories and sequentially changes the value of a single parameter on each trajectory, calculating the corresponding basic effect. Each trajectory contains k+1 sampling points (k being the total number of parameters), thus forming a systematic coverage of the parameter space.

[0145] Calculated based on multiple trajectories The importance of the parameters is quantified using the following two statistical indicators: mean and standard vehicle.

[0146]

[0147]

[0148] in, The number of trajectories; The basic effect of the i-th parameter on the j-th trajectory; This is the average of the absolute values ​​of the basic effects. The larger the value, the more significant the impact of this parameter on the model output. The standard deviation of the basic effect is represented by the value of , and the larger the value, the more interaction there is between this parameter and other parameters.

[0149] As a preferred embodiment of the method for identifying key factors affecting the participation of process industry load in demand response according to the present invention, in step S5, the Sobol' global sensitivity analysis method is analyzed, wherein:

[0150] Compared to the qualitative Morris screening method, the variance-based Sobol's global sensitivity analysis method can quantitatively solve the sensitivity of individual parameters and their interactions in complex nonlinear models. Sobol's global sensitivity analysis method decomposes the total variance of the system output into the variance of individual system inputs and the variance between multiple inputs to calculate the first-order global sensitivity coefficient (FSC) and the total global sensitivity coefficient (TSC) of the parameters affecting the evaluation index in industrial production processes. The sensitivity index is calculated as follows:

[0151]

[0152]

[0153] in, and These are the first-order global sensitivity index and the total global sensitivity index, respectively. and These represent the estimated values ​​of the expected value and variance of the output, respectively. and Generated by simple random sampling respectively A single independent sample set; To remove from vector x The resulting sub-vector.

[0154] Furthermore, the present invention also provides a computer device, including: a memory and a processor; the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0155] Furthermore, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method described.

[0156] The beneficial effects of this invention are as follows:

[0157] This paper proposes an integrated industrial energy system architecture that takes into account the scheduling of typical processes. It comprehensively considers the interaction of multiple energy sources and the dynamic coupling of industrial processes, enhances the STN model, and fully explores the scheduling potential between processes. It adopts a qualitative and quantitative analysis method, using the Morris pre-screening method to efficiently rank parameter sensitivity and achieve preliminary rapid screening of parameters. The Sobol' global sensitivity analysis method is used to accurately calculate the quantitative impact of parameters, deeply analyze their response value, optimize the sensitivity analysis process, and accurately characterize the sensitivity characteristics of influencing factors in industrial production processes. Attached Figure Description

[0158] Figure 1 An industrial integrated energy system structure diagram provided as an embodiment of the present invention for a method of identifying key factors affecting the participation of process industry load in demand response;

[0159] Figure 2 A flowchart illustrating the solution of a method for identifying key factors influencing the participation of process industry load in demand response, provided in one embodiment of the present invention.

[0160] Figure 3 A diagram illustrating the operational status of each stage of a production line, provided as an embodiment of the present invention, illustrates a method for identifying key factors influencing the load participation of process industries in demand response.

[0161] Figure 4 A sensitivity distribution diagram of different input parameters for various output indicators of a method for identifying key factors affecting the participation of process industry load in demand response, provided as an embodiment of the present invention;

[0162] Figure 5 This is a cumulative probability distribution diagram for various scenarios of a method for identifying key factors affecting the participation of process industry load in demand response, provided as an embodiment of the present invention. Detailed Implementation

[0163] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, other embodiments obtained by those skilled in the art without creative effort are all within the protection scope of this invention.

[0164] Example 1

[0165] Reference Figure 1-5 This is the first embodiment of the present invention, which provides a method for identifying key factors affecting the participation of process industry load in demand response, including:

[0166] Step S1: Based on the energy supply structure and production process of industrial users, construct an industrial integrated energy system architecture that includes an integrated energy system and a state task network.

[0167] Step S2: For various influencing parameters in the production process (assuming they follow a normal distribution), a small-scale input sample set is generated through random sampling. Based on the industrial integrated energy system architecture, an energy optimization scheduling model for general industrial processes is constructed, optimized with profit maximization as the objective function, and the optimized industrial load production index is extracted as the output sample.

[0168] Step S3: Define the influencing factors and comprehensive performance evaluation indicators of the process industry's load production process.

[0169] Step S4: Calculate the basic effects of each input parameter based on the Morris method to perform preliminary screening of the influencing parameters.

[0170] Step S5: Quantify the global sensitivity index of the parameters after screening based on the Sobol' method, and clarify the influence weight of each parameter on the production plan output results.

[0171] In step S1 of this invention, the architecture of an integrated industrial energy system is studied.

[0172] Industrial integrated energy systems integrate multiple energy forms such as electricity, heat, cooling, and gas, leveraging the synergistic and complementary effects between heterogeneous energy flows to achieve coordinated optimization of energy allocation and production scheduling. This effectively meets the demands of complex industrial production and realizes dynamic coupling between the energy system and industrial processes. Typical industrial integrated energy systems are widely used in energy-intensive industries such as steel, chemicals, and electrolytic aluminum. The structure of an industrial integrated energy system includes an integrated energy system architecture and a State Task Network (STN) architecture. Therefore, this invention constructs an STN representation considering multi-energy coupling to describe the production process and optimize energy consumption.

[0173] The integrated energy system architecture is supply-side, primarily composed of three units: energy input, conversion, and storage. The energy input unit integrates various energy sources such as distributed photovoltaic (PV), public power grid, and natural gas, prioritizing PV power generation and supplementing with grid power when insufficient, ensuring stable and economical power supply. The energy conversion unit, through multi-energy complementarity and synergistic optimization, meets the electricity, heat, and cooling needs of industrial users, including combined heat and power (CHP), gas-fired boilers, waste heat boilers, air conditioning systems, and absorption chillers, significantly improving energy utilization efficiency. The energy storage unit includes electrical energy storage systems and thermal storage tanks, used to balance supply and demand fluctuations. Electrical energy storage systems store surplus electricity and release it during peak periods, while thermal storage tanks store waste heat or cooling energy to meet energy demands during specific periods.

[0174] The state-task network structure diagram, representing the demand side, illustrates the characteristics of a typical industrial manufacturing process. During production, a buffer serves as a storage area between every two stages. As raw materials, As an intermediate product, For the final product. If Production volume is higher than Consumption levels will be monitored, and excess products will be stored in an intermediate buffer. The goal of production planning is to complete the production of a specified quantity of products within a cycle and follow a strict production sequence, combining industrial production and energy supply, flexibly arranging the operation of production equipment, effectively reducing operating costs, and achieving interaction with the power grid.

[0175] In step S2 of the present invention, specific steps are studied.

[0176] S2.1 Define that the distribution of parameters affecting the production process in the process industry follows a normal distribution. A small-scale input sample set is generated by random sampling, and each parameter follows a different parameter range according to the actual situation.

[0177] S2.2 Based on the state-task network structure of the industrial integrated energy system in step S1, and combined with the energy consumption characteristics of production processes, a generalized unified modeling method for industrial production processes is constructed.

[0178] S2.3 Based on the integrated energy system architecture in the industrial integrated energy system of step S1, construct the integrated energy system equipment model;

[0179] S2.4. Based on the response to the power grid interaction demand, construct an industrial production economic optimization scheduling model with the goal of maximizing the total system profit.

[0180] In step S2.2 of the present invention, the expression of the industrial production process model is studied.

[0181] (1) Operational constraints in the production process

[0182] Production process operational constraints include production status constraints, minimum production line running time constraints, and dynamic transfer constraints. Among these, production status constraints include:

[0183]

[0184] Each link and each production line can only be in one operating state at time t.

[0185] Minimum production line run time constraints:

[0186]

[0187] Dynamic transitive constraints:

[0188]

[0189] Prevent production interruptions or material accumulation. Ensure that the operating status of upstream machines is promptly responded to by downstream machines.

[0190] in, For each link in t The time and the operating mode of each production line. For production line indexing, ; m is the production process index, ; This is a production process mode index, which refers to the operating mode of process m on production line l. This is a state where production is stopped, meaning no production activities are carried out, and both energy consumption and output are zero. This represents the standard operating state of the production process, i.e., operating at standard or medium load, with a certain output and energy consumption, meeting the needs of routine production. This refers to a high-load operation state in the production process, that is, operating at high load or maximum capacity, resulting in higher output, but also a significant increase in energy consumption. This refers to the minimum start-up time for each production line, each stage, and each production mode.

[0191] (2) Buffer balance constraint

[0192] Buffer balancing constraints include storage constraints and production order constraints. Storage constraints include:

[0193]

[0194]

[0195] The current storage quantity consists of three parts: the storage quantity of the previous period, the input quantity of the upstream process, and the consumption quantity of the current process; Production sequence constraints:

[0196]

[0197] Downstream processes can only operate if there is sufficient inventory in the buffer zone of upstream processes.

[0198] in, Let be the buffer storage capacity of each link in each production line at time t; The amount of raw materials input for each stage of each production line; Let be the production volume of each stage of each production line at time t; This refers to the output per unit time of each stage of each production line in each operating mode. , These represent the minimum and maximum storage capacities of the buffer, respectively. Upper limit of buffer capacity for each stage

[0199] (3) Production constraints

[0200] Production constraints:

[0201]

[0202] By multiplying the productivity in the current state by the state variable, the output constraints in the production process are quantified.

[0203] Final production target constraints:

[0204]

[0205] in, The total output of the final produced products is in units. The target output.

[0206] In step S2.3 of the present invention, the model expression of the integrated energy system equipment is studied.

[0207] (1) Multi-current coupling device

[0208] 1) Combined cooling, heating and power (CCHP) unit

[0209]

[0210]

[0211]

[0212] in, , Let be the power generation and cooling power of the CCHP unit at time t, respectively, in kW; The thermal power output of the waste heat boiler is expressed in kW. This refers to the start-up and shutdown status of the CCHP unit; This refers to the gas consumption of the CCHP unit. , , and These are the cooling, heating, and electrical efficiencies and self-dissipation rate of the CCHP unit, respectively. This indicates the low calorific value of natural gas.

[0213] 2) Electric and thermal energy storage

[0214] Electric energy storage:

[0215]

[0216]

[0217]

[0218]

[0219]

[0220]

[0221] In the formula, for t Energy storage capacity at all times; , These are the charging and discharging efficiencies of electrical energy storage, respectively. These are the upper and lower limits of the energy storage capacity, respectively. , These represent the remaining electricity at the end of the energy storage dispatch and the remaining electricity at the beginning of the dispatch, respectively. They are respectively t The state variables of the battery during charging and discharging are constantly monitored. =1、 =0; during discharge =0、 =1.

[0222] Thermal energy storage:

[0223]

[0224]

[0225]

[0226]

[0227]

[0228]

[0229] In the formula, for t Storing heat energy at all times; , These are the charging and releasing efficiencies of thermal energy storage, respectively. These are the upper and lower limits of the heat storage capacity for thermal energy storage; , These represent the remaining heat at the end of the thermal energy storage dispatch and the remaining heat at the beginning of the dispatch, respectively. They are respectively t The constant state variables of the heat storage device during charging and discharging, during charging. =1、 =0; when releasing heat =0、 =1;

[0230] 3) Air conditioning and absorption chillers

[0231]

[0232]

[0233]

[0234] In the formula, for t Cooling power of Shike Leng air conditioner, KW; For air conditioning cooling efficiency; for t Power consumption of a 24-hour air conditioner (in KW). , These are the cooling capacity and heat absorption capacity of the refrigeration unit, respectively. The energy conversion efficiency of the refrigeration unit; This represents the upper limit of the refrigeration capacity of the refrigerator.

[0235] 4) Gas-fired boiler

[0236]

[0237]

[0238] in, For GB in t Heating power at any given time, in KW; The amount of natural gas consumed by GB, in m³; The heating efficiency is in GB. , These represent the upper and lower boundaries of GB output, respectively.

[0239] (2) Multi-energy equilibrium constraint

[0240] 1) Power balance

[0241]

[0242] in, Let be the energy consumption of each stage in state p, expressed in kWh.

[0243] 2) Thermal energy balance

[0244]

[0245] in, Let KWh represent the thermal energy consumption of each stage under state p.

[0246] 3) Cold energy balance

[0247]

[0248] in, Let KWh represent the cold energy consumption of each stage under state p.

[0249] In step S2.4 of this invention, an economic optimization scheduling model for industrial production is studied.

[0250] The willingness of enterprises to participate in grid interaction is highly dependent on economic returns. Therefore, their optimization goal is to optimize production scheduling and energy management by maximizing the total system profit while responding to grid demand.

[0251]

[0252] in, The final sales revenue of the manufactured products is expressed in yuan. The total cost of industrial production is expressed in yuan.

[0253] (1) Total Revenue

[0254]

[0255] in, To optimize cycle length; For production line indexing; The unit price is the selling price of the product, expressed in yuan.

[0256] (2) Total cost

[0257]

[0258] in, For energy purchase costs; For production and maintenance costs; For production costs; For start-up and shutdown costs; For storage costs.

[0259] Energy purchase costs consist of electricity and gas purchase costs. Industrial systems prioritize the use of their own distributed photovoltaic power, and when this power is insufficient, they purchase electricity from the external grid to meet the power needs of equipment in various production processes. Gas purchase costs include three parts: fuel costs for CCHP and GB units, expressed as follows:

[0260]

[0261] in, and These are the costs of purchasing electricity and gas, respectively. Let t be the external grid electricity price, in yuan / kWh; Let t be the power purchased from the external power grid, in KW; The price is for natural gas, in yuan / m3; and Let t represent the gas purchase amounts of CCHP and GB, respectively.

[0262] Considering that although the production line is fully automated, a small number of personnel are still required for inspection and monitoring during the start-up, shutdown and operation of the production line, there are maintenance costs for each production process.

[0263]

[0264] Where m represents the production stage; This is the operation and maintenance cost coefficient.

[0265] Production costs encompass the expenses directly related to each production process in an industrial system, mainly including raw material consumption and equipment depreciation.

[0266]

[0267] in, The unit production cost is expressed in yuan per unit. Let t represent the output of each stage of each production line, in units.

[0268] Start-up and shutdown costs refer to the additional expenses incurred during the start-up and shutdown of a production line. These mainly include increased energy consumption during equipment start-up and shutdown, as well as system debugging costs. Frequent start-up and shutdown operations may still affect the lifespan and stability of the equipment.

[0269]

[0270] in, Indicates the unit cost of the switch; and These indicate the on and off states, respectively.

[0271] Storage costs involve the expenses incurred in warehousing raw materials, intermediate products, or finished goods during the production process. In industrial systems, storage costs also need to consider the impact of inventory backlogs or stockouts on production planning, thereby optimizing inventory management to reduce overall operating costs.

[0272]

[0273] in, Cost per unit of storage.

[0274] In step S3 of this invention, influencing factors and comprehensive performance evaluation indicators are studied.

[0275] (1) Definition of influencing factors of industrial system

[0276] This invention selects unit output production cost Unit product sales price External power grid electricity price Maximum buffer capacity Target output Start-up and shutdown costs Natural gas prices Unit inventory cost and unit production and maintenance costs Nine categories of factors serve as the primary sources of influence. These factors cover multiple core elements affecting system stability and economy, such as cost-benefit analysis during the production process, energy price fluctuations, and equipment scheduling flexibility. They influence electricity demand response by affecting economic efficiency, energy efficiency, and production flexibility.

[0277] Among these factors, unit production cost, unit product sales price, and unit production and maintenance cost reflect fluctuations on both the cost and revenue sides, directly impacting the profit margins of industrial enterprises when responding to the power grid and influencing their economic incentives to participate in the response. Fluctuations in external power grid prices, especially under time-of-use pricing mechanisms, can interfere with the cost-effectiveness of enterprises' off-peak electricity consumption or peak-shaving, making it difficult to optimize response strategies. Uncertainty regarding the upper limit of buffer capacity will affect material synchronization and system scheduling feasibility, and inventory constraints may hinder load transfer. Uncertainty regarding target output will cause enterprises to prioritize production plans, weakening their ability to respond to the power grid. Uncertainty regarding start-up and shutdown costs will increase the economic risk of equipment start-up and shutdown, limiting enterprises' load adjustment capabilities in demand response. Natural gas prices, as a key energy input cost factor in industrial production, especially dominating in heat supply, will affect the energy costs of multi-energy coupled systems due to price fluctuations. Changes in unit inventory costs indirectly weaken enterprises' willingness to adjust production plans to respond to the power grid by increasing storage costs. These uncertainties, through mechanisms such as cost transmission, efficiency reduction, and limited flexibility, constrain the performance of IIES in production-energy synergistic optimization, thereby affecting the willingness and ability of industrial enterprises to participate in demand response.

[0278] (2) Comprehensive performance evaluation indicators

[0279] 1) Economic indicators

[0280] Total system profit is a key indicator for measuring the economic efficiency of industrial production, and it directly reflects the system's net revenue capacity under the influence of various factors.

[0281]

[0282] in, It is the total profit of the industrial production system; and These represent the final sales revenue of the manufactured products and the total industrial production cost, respectively, in yuan.

[0283] Energy procurement costs are a key characteristic of the high energy dependence of industrial systems, especially in industrial park scenarios where fluctuations in energy prices have a significant impact on system operating costs.

[0284]

[0285] in, Energy procurement cost, in yuan; Let t be the external grid electricity price, in yuan / kWh; Let t be the power purchased from the external power grid at time t, in KW. The price is for natural gas, in yuan / m3; Let t be the amount of gas purchased by CHP. ; Let t be the gas purchase amount of the gas-fired boiler. .

[0286] 2) Energy efficiency indicators

[0287] Integrated energy utilization rate is an important indicator for evaluating the energy conversion efficiency of multi-energy complementary systems. It is defined as the ratio of the total effective energy output of the system to the total input of non-renewable energy. In regional integrated energy systems, due to the coupling and conversion of various heterogeneous energy sources such as electricity, heat, and cooling, it is necessary to accurately quantify the overall system performance.

[0288]

[0289]

[0290]

[0291] in, For comprehensive energy utilization rate; This represents the total amount of non-renewable energy consumed by the system during its operating cycle T. This refers to the total effective energy provided by the system within period T.

[0292] 3) Production flexibility indicators

[0293] Inventory levels reflect the buffering capacity of an industrial system between production and demand. Uncertainty about inventory levels can lead to excessive or insufficient inventory, which in turn affects system costs.

[0294]

[0295] in, For inventory levels, [number] units; The inventory level of each production line and each stage at the end of the scheduling process is [number of units].

[0296] Total production is a direct representation of the production capacity of an industrial system, and its impact affects the system's output capacity. As a fundamental indicator for measuring system performance, total production is closely related to inventory levels and system costs.

[0297]

[0298] in, Total production quantity, in units; The total output of the final produced products is in units.

[0299] In step S4 of this invention, a Morris-based pre-screening method is studied.

[0300] Considering the increased computational cost due to the large number of parameters, and the fact that not all parameters have an equally important impact on the model output, the Morris method is first used to quickly screen out the parameters with the most significant impact.

[0301] The Morris method assesses the impact of each input parameter on the model output by calculating the elementary effect (EE) of each input parameter. For model parameters... and the i-th parameter Its basic effect It can be represented as:

[0302]

[0303] in, The predefined step size reflects the magnitude of parameter perturbation. To comprehensively explore the global characteristics of the parameter space, the Morris method randomly generates multiple trajectories and sequentially changes the value of a single parameter on each trajectory, calculating the corresponding basic effect. Each trajectory contains k+1 sampling points (k being the total number of parameters), thus forming a systematic coverage of the parameter space.

[0304] Calculated based on multiple trajectories The importance of the parameters is quantified using the following two statistical indicators: mean and standard vehicle.

[0305]

[0306]

[0307] in, The number of trajectories; The basic effect of the i-th parameter on the j-th trajectory; This is the average of the absolute values ​​of the basic effects. The larger the value, the more significant the impact of this parameter on the model output. The standard deviation of the basic effect is represented by the value of , and the larger the value, the more interaction there is between this parameter and other parameters.

[0308] In step S5 of this invention, the Sobol' global sensitivity analysis method is studied.

[0309] Compared to the qualitative Morris screening method, the variance-based Sobol's global sensitivity analysis method can quantitatively solve the sensitivity of individual parameters and their interactions in complex nonlinear models. Sobol's global sensitivity analysis method decomposes the total variance of the system output into the variance of individual system inputs and the variance between multiple inputs to calculate the first-order global sensitivity coefficient (FSC) and the total global sensitivity coefficient (TSC) of the parameters affecting the evaluation index in industrial production processes. The sensitivity index is calculated as follows:

[0310]

[0311]

[0312] in, and These are the first-order global sensitivity index and the total global sensitivity index, respectively. and These represent the estimated values ​​of the expected value and variance of the output, respectively. and Generated by simple random sampling respectively A single independent sample set; To remove from vector x The resulting sub-vector.

[0313] Example 2

[0314] Reference Figure 1-5 This invention provides a method for identifying key factors affecting the participation of process industry load in demand response, as one embodiment of the present invention. To verify the beneficial effects of the present invention, comparative experiments are conducted for scientific demonstration.

[0315] The example uses a modified Stn model to describe the production process. Raw materials are processed through five stages (m1~m5) on three production lines (l1~l3) to produce finished products. Each stage has different operating states (p0~p2). It is assumed that m1 can only operate under p1 (normally on); m2 and m3 can operate under both p0 and p1 (on / off) at medium load; m4 and m5 can operate simultaneously under all three states (on / off) and can operate under both medium and high loads. Energy consumption parameters under different modes are shown in Table 1. Output per unit time is shown in Table 2. The distribution of influencing factors in the industrial production process is shown in Table 3. Industrial users purchase electricity from the external power grid at peak-valley prices, specifically divided as follows: 1:00-7:00 and 24:00 are valley periods; 11:00-15:00 and 19:00-21:00 are peak periods; and 8:00-10:00, 16:00-18:00 and 22:00-23:00 are normal periods. The optimization scheduling model is solved using the commercial solver GUROBI in the GAMS environment, while the parameter identification model is solved in the MATLAB environment. The optimized operating environment is a Windows 10 system with an Intel Core i7-8750H CPU (2.20GHz) and 8GB of RAM.

[0316] Table 1. Energy consumption parameters under different modes

[0317] Table 1. Energy Consumption Parameters under Different OperatingModes

[0318]

[0319] Table 2 Output per unit time

[0320] Table 2. Production Rate

[0321]

[0322] Table 3 Distribution of Influence Parameters on Industrial Production Processes

[0323] Table 3. Distribution of Uncertainty Parameters in IndustrialProduction Processes

[0324]

[0325] Figure 3This diagram illustrates the operational status of each stage of the production line. Mode P1, with its moderate output and energy consumption, dominates stages M1 and M3-M5. Mode P2 is only activated in stage M5, and due to its high energy consumption, its usage frequency is relatively low, primarily to address short-term high output demands. This scheduling scheme meets production needs while utilizing peak-valley electricity price differences to reduce production costs. Stage M5 typically switches to P2 during off-peak hours (1:00-7:00 and 21:00-24:00) and normal hours (12:00-16:00) to increase output, while during peak hours (12:00-16:00), it remains shut down or only operates in P1 to avoid excessive energy consumption from high-load operation. Furthermore, each production line does not frequently switch to high-load modes over a longer scheduling cycle, primarily maintaining a relatively stable operating state to ensure output and reduce start-up and shutdown costs, effectively balancing energy consumption and electricity prices while considering overall output targets.

[0326] To initially screen model parameters, a global sensitivity analysis was first performed on 11 parameters using the Morris method. This analysis was conducted on each parameter... The statistics are used to calculate μ* and σ, which respectively measure the strength of the parameter's influence and the degree of nonlinearity or interaction effect.

[0327] Considering that a single indicator may not fully reflect the importance of parameters, a fusion indicator is further constructed: after normalizing μ* and σ to the [0,1] interval, a weighted average method is used to calculate the comprehensive importance score, where the weight of μ* is set to 0.7 and the weight of σ is set to 0.3. Finally, the top 25% of parameters are selected, and redundant variables with little impact on the model output are removed. Their sensitivity distribution is as follows: Figure 4 As shown.

[0328] Based on the results of the Morris method and the industrial production background, the following parameters were selected: parameter 1 (unit output production cost), parameter 2 (unit product sales price), parameter 4 (normal electricity price), parameter 5 (off-peak electricity price), parameter 7 (target output) and parameter 8 (start-up and shutdown costs).

[0329] Depend on Figure 4It can be seen that parameters 1 and 2 are the main determinants of total profit, directly related to the profitability of enterprises. Changes in these two types of parameters have a significant impact on the objective function of the production optimization model, especially when costs and market prices fluctuate drastically. Parameters 4 and 5 show high sensitivity in energy purchase costs, inventory, and overall energy utilization rate, reflecting the broad impact of electricity prices on energy costs and utilization efficiency. Due to lower electricity prices during normal and off-peak hours, factories tend to schedule energy-intensive processes during off-peak hours to reduce electricity costs. Therefore, even if the fluctuation range of off-peak electricity prices is small, it may significantly affect production scheduling strategies. Especially when production flexibility is limited, its sensitivity to scheduling optimization may be further enhanced. Parameter 7 is important in total profit and inventory, and is a core constraint in production planning, directly affecting resource allocation and cost structure. Its changes will significantly affect the optimal solution of the model. Parameter 8 is significant in total profit and overall energy utilization rate, affecting equipment scheduling and energy efficiency. The selection of these parameters is not only based on their significance in the Morris method, but also reflects the priority decision points in industrial production.

[0330] Parameter 3 is only important in inventory, with a smaller impact on other indicators and a limited overall influence; Parameter 6 is only important in output, but performs weakly in other indicators, and its effect is limited to the physical constraints of production capacity; Parameters 9-11 have a combined index of less than 0.5 across all indicators, indicating low sensitivity, and therefore are not considered.

[0331] Compared to Morris analysis results, Sobol global sensitivity analysis can reveal the sensitivity of parameters more intuitively. The value of the sensitivity coefficient directly reflects the strength of the influence of each parameter on the model output - the larger the coefficient, the more significant the influence of the parameter on the model output.

[0332] (1) Total system profit

[0333]

[0334] As shown in the table, the overall system profit is most affected by parameter 2 (unit product sales price), followed by parameter 8 (start-up and shutdown costs), while the impact of parameters 1 (unit output production cost), 4-5 (grid electricity price), and 7 (target output) is relatively small. The high sensitivity of parameter 2 stems from its two-way coupling effect in grid response: on the one hand, the sales price directly determines product revenue to offset grid interaction costs; on the other hand, its fluctuations guide the dynamic coordination between production plans and grid peak-shaving periods through market signals. When the grid implements peak electricity prices, high sales prices can support users to bear higher electricity prices to maintain continuous production; conversely, it is necessary to adjust output to match the electricity consumption strategy during low-price periods. The sensitivity of parameter 8 reflects the conflict between grid commands and industrial equipment operation—frequent responses to grid start-up and shutdown demands will lead to increased start-up and shutdown costs. Especially when the cost exceeds a certain critical value, the system will tend to adjust to continuous production mode and exit short-term demand response to avoid high start-up and shutdown costs. The overall global sensitivity coefficients of parameters 4-5 and 7 are both below 0.005 because electricity costs account for a relatively small proportion of the total cost, and the system can reduce the impact of electricity price fluctuations by optimizing production plans. The weakening of the rigid constraint effect of the target output is mainly due to the flexibility of real-time scheduling optimization. These results indicate that the high sensitivity of sales prices to start-up and shutdown costs reveals the need to construct a price-market linkage decision-making model to stabilize demand response revenue, while low electricity prices and output sensitivity confirm the resilience of flexible production scheduling to grid uncertainties.

[0335] (2) Energy procurement costs

[0336]

[0337] Parameter 5 (grid price - off-peak hours) exhibits extremely high sensitivity to energy procurement costs. Since industrial systems tend to schedule energy-intensive processes during off-peak hours to reduce electricity costs, fluctuations in off-peak electricity prices significantly impact energy procurement costs through electricity consumption. While the impact of parameter 4 (grid price - normal operating hours) on energy procurement costs is less significant than that of off-peak prices, it is still considerable. Normal operating prices primarily affect electricity costs during non-off-peak periods; in scenarios with limited production scheduling flexibility, fluctuations in normal operating prices may force the system to adjust production plans, indirectly increasing energy procurement costs. Parameter 7 (target output) has a limited direct impact on energy procurement costs; the flexibility of real-time scheduling optimization weakens the rigid constraint effect of target output. Industrial systems can dynamically adjust production plans based on real-time conditions (such as electricity price fluctuations), thus buffering the impact of output changes on energy costs. The sensitivity coefficients of parameters 1 (unit output production cost), 2 (unit product sales price), and 8 (start-up cost) are extremely low. Unit output production cost and sales price mainly affect production costs and revenue, and are not directly related to energy procurement costs. Although start-up cost affects production planning, its proportion in energy costs is extremely small, and its impact can be mitigated by optimizing start-up and shutdown strategies. Therefore, its impact is negligible compared to the previous three. This sensitivity result proves that the time-series characteristics of electricity prices are the decisive factor driving changes in energy procurement costs.

[0338] (3) Overall energy utilization rate

[0339]

[0340] The overall energy utilization rate is most significantly affected by parameter 5 (grid price - off-peak hours). Due to lower off-peak electricity prices, industrial systems tend to schedule energy-intensive processes during off-peak hours to reduce costs. This strategy leads to a significant increase in energy consumption during off-peak hours, thus affecting the overall energy utilization rate. This response pattern essentially reflects the grid price signal's ability to regulate industrial energy consumption behavior—guiding load shifting through lower off-peak prices achieves both peak shaving and valley filling on the grid side and improves energy utilization on the user side. Parameters 7 (target output) and 8 (start-up and shutdown costs) have relatively smaller impacts on the overall energy utilization rate. This stems from the flexibility of real-time scheduling optimization; the system can maintain stable energy utilization efficiency by adjusting production plans or energy consumption. If the target output increases, the system will overuse energy during off-peak hours, leading to a decrease in energy utilization efficiency. Changes in start-up and shutdown costs affect equipment start-up and shutdown strategies, thereby impacting continuous energy use and utilization efficiency. If the cost of starting and stopping equipment is high, the system will reduce the number of equipment start-ups and shutdowns, extending continuous operating time and thus improving energy efficiency. The impact of parameters 1 (unit output production cost), 2 (unit product sales price), and 4 (grid electricity price - normal operating hours) is minimal. Unit output production cost and sales price mainly affect production costs and revenue, with a weak direct correlation to energy efficiency. Although the normal operating electricity price affects energy costs, its impact on overall energy utilization is significantly weakened in scenarios dominated by off-peak electricity prices, and therefore its influence is negligible compared to the previous three. The above parameter sensitivity distribution characteristics confirm that in the interaction between industrial load and the grid, the external grid electricity price, especially the off-peak electricity price, is the core factor driving the improvement of energy efficiency.

[0341] (4) Inventory level

[0342]

[0343] Inventory levels are most significantly affected by parameter 7 (target output). The setting of inventory levels directly determines the degree of matching between production scale and electricity demand. When the target output is too high, continuous production to meet grid load requirements will lead to inventory accumulation. This not only increases the cost of energy storage but may also force production to continue during peak electricity price periods due to exceeding the buffer capacity limit. Conversely, if the target output decreases, the inventory level may decrease, but this will lead to supply shortages. Secondly, the impact of parameter 8 (start-up and shutdown costs) reflects the conflict between responding to grid demand and industrial operation. High start-up and shutdown costs will inhibit the system's ability to respond to grid regulation commands, prompting enterprises to choose continuous operation mode to avoid frequent start-ups and shutdowns. This strategy of "trading inventory for grid stability" ensures grid demand but at the cost of increased inventory levels. The impacts of parameters 1 (unit output production cost), 2 (unit product sales price), 4 (grid electricity price - normal times), and 5 (grid electricity price - off-peak hours) on inventory levels are all very small and can be ignored compared to the first three.

[0344] (5) Total production

[0345]

[0346] Total production is most affected by parameter 8 (start-up and shutdown costs). When costs rise, industrial users are often forced to maintain continuous production to avoid economic losses from frequent start-ups and shutdowns. This not only rigidifies production constraints but also weakens the system's ability to respond to grid load regulation demands. Conversely, when costs decrease, the production system can respond more flexibly to grid time-of-use pricing signals or regulation commands, dynamically optimizing total production through start-up and shutdown strategies. Parameter 7 (target output) has the next greatest impact on total production. Setting the target output directly limits the upper limit of production scale. However, in actual operation, the system needs to balance the benefits of grid interaction with production constraints, often using flexible production strategies, such as overproduction during off-peak hours and reduced production during peak hours, to balance total production with grid demand. Parameters 1 (unit output production cost), 2 (unit product sales price), and 4-5 (grid electricity price) have extremely low sensitivity coefficients and their impact on total production is negligible.

[0347] To verify the validity of the above Sobol' global sensitivity analysis results, the following scenario was set up:

[0348] Scenario S1: All parameters exhibit random bias;

[0349] Scenario S2: Only critical parameters exhibit random deviations, while non-critical parameters remain fixed at baseline values;

[0350] Scenario S3: Only non-critical parameters exhibit random deviations, while critical parameters remain fixed at the baseline value.

[0351] The Cumulative Distribution Function (CDF) curve reflects the mapping relationship between the model output value and its probability. By observing the changing trends of the CDF curve under different scenarios, the influence of key and non-key parameters on the model output can be assessed. The CDF for different scenarios is calculated using the total system profit as a representative example. The CDF curves for different scenarios are shown below. Figure 5 As shown.

[0352] Figure 5 The cumulative probability distribution plot shows that the CDF curve of S2 highly overlaps with that of S1, indicating that changes in only the key parameters can well reproduce the output distribution when all parameters change. However, the CDF curve of S3 deviates significantly from S1, with a narrower output range, indicating that changes in non-key parameters have a smaller impact on the model output. Therefore, changes in the key parameters are sufficient to explain the main impact of changes in all parameters, validating the effectiveness of Sobol's global sensitivity analysis method.

[0353] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0354] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0355] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0356] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0357] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A method for identifying key factors affecting the participation of process industry load in demand response, characterized in that, include: Step S1: Based on the energy supply structure and production process of industrial users, construct an industrial integrated energy system architecture that includes an integrated energy system and a state task network; Step S2: For various influencing parameters in the production process, a small-scale input sample set is generated through random sampling. Based on the industrial integrated energy system architecture, an energy optimization scheduling model for general industrial processes is constructed. The model is optimized with profit maximization as the objective function, and the optimized industrial load production index is extracted as the output sample. Step S3: Define the influencing factors and comprehensive performance evaluation indicators of the process industry's load production process; Step S4: Calculate the basic effects of each input parameter based on the Morris method, and perform preliminary screening of the influencing parameters; Step S5: Quantify the global sensitivity index of the screened parameters based on the Sobol' method, and clarify the influence weight of each parameter on the production plan output results; The industrial integrated energy system in step S1 includes: The integrated energy system architecture on the supply side includes an energy input unit, an energy conversion unit, and an energy storage unit. The energy input unit integrates distributed photovoltaic, public grid, and natural gas energy, prioritizing photovoltaic power generation and supplementing with grid power when electricity is insufficient. The energy conversion unit meets the electricity, heat, and cooling needs of industrial users through multi-energy complementarity and synergistic optimization, and includes combined heat and power, gas-fired boilers, waste heat boilers, air conditioning, and absorption chiller equipment. The energy storage unit includes an electric energy storage system and a thermal storage tank to balance supply and demand fluctuations. The electric energy storage system stores surplus electrical energy, while the thermal storage tank stores waste heat or cooling energy. The demand-side state-task network structure is used to demonstrate the characteristics of a typical industrial manufacturing process; during the production process, a buffer is set up between every two stages as a storage area. In step S3, the influencing factors and comprehensive performance evaluation indicators are defined as follows: (1) Definition of influencing factors of industrial system Selecting unit output production cost Unit product sales price External power grid electricity price Maximum buffer capacity Target output Start-up and shutdown costs Natural gas prices Unit inventory cost and unit production and maintenance costs Nine categories of factors serve as sources of influence; (2) Comprehensive performance evaluation indicators 1) Economic indicators in, It is the total profit of the industrial production system; and These are the final sales revenue of manufactured products and the total cost of industrial production, respectively. in, For energy procurement costs; Let t be the external grid electricity price; Let t be the power purchased from the external power grid at time t; For natural gas prices; Let t be the amount of gas purchased by CHP; Let t be the gas purchase amount of the gas boiler; 2) Energy efficiency indicators Defined as the ratio of the total effective energy output of the system to the total input of non-renewable energy. in, For comprehensive energy utilization rate; This represents the total amount of non-renewable energy consumed by the system during its operating cycle T. The total effective energy provided by the system within period T; 3) Production flexibility indicators in, Inventory level; This represents the inventory levels at each stage of each production line at the end of the scheduling process. in, Total production; This refers to the total output of the final products.

2. The method for identifying key factors affecting the participation of process industry load in demand response as described in claim 1, characterized in that, Step S2 specifically includes: S2.1 Define that the distribution of parameters affecting the production process in the process industry all follow a normal distribution, and generate a small-scale input sample set through random sampling; S2.2 Based on the state-task network structure of the industrial integrated energy system in step S1, and combined with the energy consumption characteristics of production processes, a generalized unified modeling method for industrial production processes is constructed. S2.3 Based on the integrated energy system architecture in the industrial integrated energy system of step S1, construct the integrated energy system equipment model; S2.

4. Based on the response to the power grid interaction demand, construct an industrial production economic optimization scheduling model with the goal of maximizing the total system profit.

3. According to the method for identifying key factors affecting the participation of process industry load in demand response as described in claim 2, in step S2.2, the expressions of the constructed generalized industrial production process model are as follows: (1) Operational constraints in the production process Production process constraints include production state constraints, minimum production line running time constraints, and dynamic transfer constraints. Among these, production state constraints include: Each link and each production line can only be in one operating state at time t. Minimum production line run time constraints: Dynamic transitive constraints: in, For each link in t The timing and operating mode of each production line, For production line indexing, ; m is the production process index, ; This is a production process mode index, which indicates the operational mode of process m on production line l. This represents a production stoppage state, meaning no production activities are carried out, and both energy consumption and output are zero. This refers to the standard operating state of the production process, i.e., operating at standard or medium load. This refers to a high-load operation state in the production process, i.e., operating at high load or maximum capacity. The minimum start-up time for each production line, each stage, and each production mode; (2) Buffer balance constraint Buffer balancing constraints include storage constraints and production order constraints, where storage constraints are: The current storage quantity consists of three parts: the storage quantity of the previous period, the input quantity of the upstream process, and the consumption quantity of the current process; Production sequence constraints: Downstream processes can only operate if there is sufficient inventory in the buffer zone of upstream processes; in, Let be the buffer storage capacity of each link in each production line at time t; The amount of raw materials input for each stage of each production line; Let t be the production volume of each stage of each production line; This refers to the output per unit time of each stage of each production line in each operating mode. , These represent the minimum and maximum storage capacities of the buffer, respectively. This is the upper limit of the buffer capacity for each stage; (3) Production constraints Production constraints: By multiplying the productivity in the current state by the state variable, the output constraints in the production process can be quantified. Final production target constraints: in, For the total output of the final products, The target output.

4. According to the method for identifying key factors affecting the participation of process industry load in demand response as described in claim 3, in step S2.3, the constructed integrated energy system equipment model expression is as follows: (1) Multi-current coupling device 1) Combined cooling, heating and power (CCHP) units in, , Let be the power generation and cooling power of the CCHP unit at time t, respectively, in kW; The thermal power output of the waste heat boiler is expressed in kW. This refers to the start-up and shutdown status of the CCHP unit; This refers to the gas consumption of the CCHP unit. , , and These are the cooling, heating, and electrical efficiencies and self-dissipation rate of the CCHP unit, respectively. This indicates the low calorific value of natural gas; 2) Energy storage In the formula, for t Energy storage capacity at all times; , These are the charging and discharging efficiencies of electrical energy storage, respectively. These are the upper and lower limits of the energy storage capacity, respectively. , These represent the remaining electricity at the end of the energy storage dispatch and the remaining electricity at the beginning of the dispatch, respectively. They are respectively t The state variables of the battery during charging and discharging are constantly monitored. =1、 =0; during discharge =0、 =1; 3) Thermal energy storage In the formula, for t Storing heat energy at all times; , These are the charging and releasing efficiencies of thermal energy storage, respectively. These are the upper and lower limits of the heat storage capacity for thermal energy storage; , These represent the remaining heat at the end of the thermal energy storage dispatch and the remaining heat at the beginning of the dispatch, respectively. They are respectively t The constant state variables of the heat storage device during charging and discharging, during charging. =1、 =0; when releasing heat =0、 =1; 4) Air conditioning and absorption chillers In the formula, for t Cooling power of Shike Leng air conditioner, KW; For air conditioning cooling efficiency; for t Power consumption of the air conditioner. , These are the cooling capacity and heat absorption capacity of the refrigeration unit, respectively. The energy conversion efficiency of the refrigeration unit; This is the upper limit of the refrigeration capacity of the refrigerator; 5) Gas-fired boiler in, For GB in t Heating power at any given time, in KW; The amount of natural gas consumed by GB, in m³; The heating efficiency is in GB. , These represent the upper and lower boundaries of GB output; (2) Multi-energy equilibrium constraint 1) Power balance in, The energy consumption of each component in state p; 2) Thermal energy balance in, The heat energy consumption of each stage under state p; 3) Cold energy balance in, This represents the cold energy consumption of each component under state p.

5. According to the method for identifying key factors affecting the participation of process industry load in demand response as described in claim 4, in step S2.4, the constructed industrial production economic optimization scheduling model is as follows: The optimization objective, based on responding to grid demand, is to optimize production scheduling and energy management by maximizing total system profit. in, This refers to the sales volume of the final manufactured products; Total industrial production cost; (1) Total Revenue in, To optimize cycle length; For production line indexing; The unit product sales price; (2) Total cost in, For energy purchase costs; For production and maintenance costs; For production costs; For start-up and shutdown costs; For storage costs; Energy purchase costs consist of electricity and gas purchase costs. Industrial systems prioritize the use of their own distributed photovoltaic power, and when the electricity is insufficient, they purchase electricity from the external grid to meet the power needs of equipment in various production processes. Gas purchase costs include three parts: fuel costs for CCHP and GB units, expressed as follows: in, and These are the costs of purchasing electricity and gas, respectively. Let t be the external grid electricity price; Let t be the power purchased from the external power grid at time t; For natural gas prices; and The gas purchase amounts for CCHP and GB at time t are respectively. Operation and maintenance costs of each production line process: Where m represents the production stage; This is the coefficient for operation and maintenance costs; Production costs encompass the expenses directly related to each production process in an industrial system, including raw material consumption and equipment depreciation. in, Unit production cost; Let t be the product output of each stage on each production line; Start-up and shutdown costs refer to the additional expenses incurred during the start-up and shutdown of a production line, including increased energy consumption during equipment startup and shutdown and system commissioning costs. in, Indicates the unit cost of the switch; and These indicate the on and off states, respectively. Storage costs involve the warehousing expenses for raw materials, intermediate products, or finished goods during the production process. In industrial systems, storage costs also consider the impact of inventory backlogs or stockouts on production planning, thereby optimizing inventory management to reduce overall operating costs. in, Cost per unit of storage.

6. The method for identifying key factors affecting the participation of process industry load in demand response according to claim 5, characterized in that, In step S4, the Morris-based pre-screening method is as follows: The Morris method is used to screen for the most significant parameters, and the basic effect (EE) of each input parameter is calculated to assess its impact on the model output. For the model parameters... and the i-th parameter Its basic effect Represented as: in, With a predefined step size to reflect the magnitude of parameter perturbation, the Morris method randomly generates multiple trajectories and sequentially changes the value of a single parameter on each trajectory to calculate the corresponding basic effect. Each trajectory contains k+1 sampling points, where k is the total number of parameters, forming a systematic coverage of the parameter space. Calculated based on multiple trajectories The importance of the parameters is quantified using the following two statistical indicators: mean and standard vehicle. in, The number of trajectories; The basic effect of the i-th parameter on the j-th trajectory; This is the average of the absolute values ​​of the basic effects. The larger the value, the more significant the impact of this parameter on the model output. The standard deviation of the basic effect is represented by the value of , and the larger the value, the more interaction there is between this parameter and other parameters.

7. A computer device, comprising: A memory and a processor; the memory stores a computer program, characterized in that: when the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

Images