A multi-process collaborative optimization regulation method and system for a steel enterprise

Abstract

A multi-process collaborative optimization and regulation method and system for steel enterprises is disclosed. Based on the steel enterprise's production architecture and technological flow, adjustable equipment in the steelmaking process is identified; power consumption characteristics of the adjustable equipment are modeled, and dynamic response characteristic parameters are quantified; a hierarchical collaboration mechanism is designed, and a collaborative optimization model is designed and constraint functions are defined based on the cost characteristics of each level; the corresponding level in the hierarchical collaboration mechanism is activated according to the type of grid demand fluctuation, and the role weights of the adjustable equipment in different levels are reallocated; the collaborative optimization model is transformed into a mixed-integer linear programming model, and the updated dynamic response characteristic parameters, constraint functions, and role weights are combined and imported into the optimization solver to output a scheduling scheme. This invention enhances the responsiveness of steel enterprises as adjustable loads, thereby improving their economic efficiency and reliability in interactive processes.

Description

Technical Field

[0001] This invention belongs to the field of industrial energy management and intelligent optimization control technology, specifically a multi-process collaborative optimization adjustment method and system for steel enterprises. Background Technology

[0002] As a high-energy-consuming industry, steel enterprises possess significant internal adjustable load potential (such as in refining and continuous casting processes), enabling energy savings during peak energy demand periods through demand-side response (DSR). Currently, the steel industry has mature optimization systems for production scheduling in key processes such as sintering, hot rolling, and continuous casting, which can improve production efficiency in specific areas.

[0003] However, existing research mainly focuses on modeling single equipment (such as electric arc furnaces and rolling mills) and static analysis of energy and material flows in steel industrial parks. In the analysis and optimization methods, the focus is mainly on time costs, equipment wear and tear or local energy consumption within the process, neglecting the coordinated scheduling of the entire production process and lacking the dynamic coupling capability across processes and multiple energy flows.

[0004] Due to differences in the response characteristics of different equipment, this independent response mode can easily lead to imbalances in material and energy consumption, resulting in production rhythm conflicts, frequent equipment start-ups and shutdowns, and other problems, affecting production efficiency, safety, and energy consumption. This localized optimization method, which lacks a global perspective and inter-process coordination, limits the overall adjustment potential of steel enterprises and cannot effectively reduce energy costs while ensuring output. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a multi-process collaborative optimization and adjustment method and system for steel enterprises. It aims to build an efficient inter-process collaboration mechanism by accurately matching the dynamic response characteristics of equipment with the adjustment task requirements. Under the premise of ensuring the safe production and daily output requirements of steel enterprises remain unchanged, it significantly reduces the total operating cost of the system and improves the economy and safety of enterprises participating in power grid demand response.

[0006] The first aspect of this application discloses a multi-process collaborative optimization and adjustment method for steel enterprises, which adopts the following technical solution: Based on the production structure and process flow of steel enterprises, adjustable equipment in the steelmaking process is identified; power consumption characteristics are modeled for the adjustable equipment, and the dynamic response characteristic parameters of the adjustable equipment are quantified; the power consumption characteristic modeling includes electric arc furnace modeling, ladle refining furnace modeling, continuous casting machine modeling, and rolling mill modeling. A hierarchical collaboration mechanism is designed based on the dynamic response characteristic parameters, including a fast response layer, a main adjustment layer, and a flexible buffer layer; a collaborative optimization model is designed based on the cost characteristics of the hierarchical division, and constraint functions are defined. The collaborative optimization model is piecewise linearized and transformed into a mixed-integer linear programming model; the power grid demand response signal and the operating status of the steel enterprise's internal production equipment are acquired in real time, and the dynamic response characteristic parameters and constraint functions are updated according to the current production stage. Analyze the types of power grid demand fluctuations, activate the corresponding levels in the hierarchical cooperation mechanism according to the fluctuation types, and reallocate the role weights of adjustable equipment in different levels; import the mixed integer linear programming model, the updated dynamic response characteristic parameters, the constraint functions, and the role weights of the production equipment into the optimization solver, and output the scheduling scheme.

[0007] Furthermore, the electric arc furnace modeling constructs a segmented time-varying power model based on the electric arc furnace operation stages to identify the optimal adjustment period for melting; the ladle refining furnace modeling reuses the segmented time-varying power model. The continuous casting machine modeling uses a linear adjustable power model to decompose the total power of the continuous casting machine and optimizes the billet drawing speed by combining the linear constraint relationship between the drawing speed and the pouring temperature. The rolling mill modeling uses a composite power model to decompose the total power of the rolling mill and calculates the adjustable load of the rolling mill by combining the rolling parameters.

[0008] Furthermore, the dynamic response characteristic parameters include response time, interruptible duration, reduceable duration, and adjustability; The response time is the time required for the device to change power from receiving a command; the interruptible duration is the longest time the device can pause operation; the reduceable duration is the longest time the device can operate at reduced power; and the adjustability is the range and magnitude of the device's power adjustment.

[0009] Furthermore, the dynamic response characteristic parameters are quantified according to the power grid control requirements, wherein: The response time of the electric arc furnace is calculated by combining the mechanical response time of the electrode lifting system and the delay of the control system; the response time of the continuous casting machine is calculated based on the response time of the billet speed adjustment system; and the response time of the rolling mill is calculated based on the rolling rhythm and the billet bite and throw cycle.

[0010] Furthermore, in the aforementioned hierarchical collaboration mechanism: The fast response layer is based on the optimal adjustment melting period of the electric arc furnace and is used to receive and execute the first type of power regulation command. The main regulating layer is based on the adjustable load output by the composite power model of the rolling mill, and is used to receive and execute the second type of power regulation command; The flexible buffer layer is based on the billet pulling speed output by the adjustable power model of the continuous casting machine, and is used to receive and execute the third type of power adjustment command.

[0011] Furthermore, based on the cost characteristics of the hierarchical division, optimization objectives are designed and constraint functions are defined, including: A collaborative optimization model is established with the goal of minimizing total production cost. The optimization objective is to minimize the total operating cost of the system. The total operating cost includes the cost of purchasing and selling electricity, the cost of thermal power and gas, the adjustment cost of steel enterprises, and the compensation cost. The compensation cost is expressed as the product of the load difference generated by the steel enterprise's response to the power grid during the dispatch period and the power compensation price paid by the operator to the steel enterprise.

[0012] The constraint functions include internal constraints of steel enterprises and power grid power balance constraints. Internal constraints of steel enterprises include internal power balance constraints, storage constraints, output constraints, and operating time constraints.

[0013] Furthermore, the process of acquiring the power grid demand response signal includes: Real-time demand response data of the power grid is obtained through the power grid interface, including the frequency, amplitude and response time window of demand fluctuations, and the role weight of each device in the fast response layer, the main regulation layer and the flexible buffer layer are assigned.

[0014] Furthermore, the power grid demand signals are identified by type, and a mapping relationship between demand types and hierarchical cooperation mechanisms is established; When the grid frequency deviation is detected to exceed the threshold and / or a fault occurs, the grid demand fluctuation type is determined to be an emergency safety demand; in the hierarchical cooperation mechanism, the fast response layer is activated as the first response level; When receiving time-of-use pricing signals, peak-valley price arbitrage instructions, and / or long-term load transfer plans, the grid demand fluctuation type is determined to be economically regulated demand; in the hierarchical cooperation mechanism, the main regulation layer is activated as the first response level. When the power grid issues a resource allocation plan and / or a renewable energy consumption instruction, the power grid demand fluctuation type is determined to be a flexible adjustment type of demand; in the hierarchical cooperation mechanism, the flexible buffer layer is activated as the first response level.

[0015] The second aspect of this application discloses a multi-process collaborative optimization and adjustment system for steel enterprises, which implements the multi-process collaborative optimization and adjustment method as described in the first aspect of this application. The system includes: Adjustable equipment modeling module; used to identify adjustable equipment in the steelmaking process based on the production architecture and process flow of steel enterprises; to model the power consumption characteristics of adjustable equipment and quantify the dynamic response characteristic parameters of adjustable equipment; the power consumption characteristic modeling includes electric arc furnace modeling, ladle refining furnace modeling, continuous casting machine modeling and rolling mill modeling; The collaborative optimization design module is used to design a hierarchical collaborative mechanism based on the dynamic response characteristic parameters, including a fast response layer, a main adjustment layer, and a flexible buffer layer; and to design a collaborative optimization model and define constraint functions based on the cost characteristics of the hierarchical division. The preprocessing module is used to perform piecewise linearization on the collaborative optimization model, transforming it into a mixed-integer linear programming model; and to acquire power grid demand response signals and the operating status of production equipment within steel enterprises in real time, updating the dynamic response characteristic parameters and constraint functions according to the current production stage. The scheduling module is used to analyze the types of power grid demand fluctuations, activate the corresponding levels in the hierarchical cooperation mechanism according to the fluctuation types, and reallocate the role weights of adjustable equipment in different levels. The mixed integer linear programming model, the updated dynamic response characteristic parameters, the constraint functions, and the role weights of the production equipment are imported into the optimization solver to output the scheduling scheme.

[0016] The beneficial effects of this invention are that, compared with the prior art, The most significant difference between this patent and existing technologies lies in its breakthrough from the limitations of traditional methods that isolate and optimize steel production equipment or respond to grid commands only within a single process. It innovatively integrates the production characteristics of both the "long process" (blast furnace – converter – refining – continuous casting – hot rolling) and the "short process" (electric arc furnace – refining – continuous casting – rolling) of steel manufacturing. It constructs a unified collaborative optimization framework that takes into account the structural differences between the two types of processes, equipment response characteristics, and energy coupling relationships. Based on the inherent differentiated dynamic response characteristics of different equipment (such as electric arc furnaces, continuous casting machines, and rolling mills), it constructs a three-layer collaborative optimization architecture of "rapid response – main force adjustment – ​​flexible buffer". It also deeply integrates process operation constraints, multi-energy flow coupling relationships, and grid demand response commands to form a multi-dimensional dynamic collaborative model of "process-energy-grid", realizing collaborative scheduling across processes, time scales, and energy media.

[0017] The resulting technical effects include: By complementing the scheduling of long and short process resources and employing a three-tiered collaborative mechanism, the system can more efficiently utilize off-peak electricity prices and avoid peak electricity prices to participate in the electricity market; significantly reduce the total operating costs of enterprises; and efficiently release the potential for flexible regulation throughout the entire process while ensuring safe production and daily output targets, achieving a win-win situation of "ensuring production and reducing costs"; enhance the reliability and economy of steel enterprises participating in the electricity market as adjustable loads, enabling them to respond more accurately to time-of-use pricing and grid dispatch signals; and systematically address the shortcomings of existing methods in terms of full-process collaboration, dynamic coupling of energy flow, and real-time response mechanisms, achieving a fundamental leap from local static optimization to global dynamic collaboration. Attached Figure Description

[0018] Figure 1A system framework diagram for multi-process power grid coordinated control; Figure 2 A multi-process production diagram for steel enterprises; Figure 3 The load diagram shows the independent response of each device to the superior's instructions. Figure 4 To optimize the load diagram in consideration of equipment response characteristics; Figure 5 To consider the response of electricity load and price in order to optimize collaborative regulation; Detailed Implementation 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, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this invention.

[0019] To promote the utilization of renewable resources and achieve demand-side response, steel enterprises, as a high-energy-consuming industry, possess enormous potential for adjustable loads. Existing research mainly focuses on modeling individual equipment—electric arc furnaces and rolling mills—or on the static synergistic analysis of energy and material flows in steel industrial parks.

[0020] Currently, the steel industry has established a relatively mature optimization system for production process scheduling, achieving significant progress in key processes such as sintering, hot rolling, and continuous casting. However, most of these optimization methods are limited to static scenarios of single processes, mainly focusing on time costs, equipment wear and tear, or local energy consumption within a process, failing to achieve coordinated scheduling of the entire production process. With the advancement of energy price marketization reforms, time-of-use pricing mechanisms, as an important policy tool to guide industrial users to participate in demand-side response, have demonstrated their leverage effect in optimizing energy costs. Nevertheless, this method has not yet been fully applied to scenarios involving cross-process and multi-energy flow coordination, particularly for modeling the dynamic coupling of energy flow throughout the ironmaking-steelmaking-rolling process and integrating real-time coordination mechanisms between grid demand response commands and production flexibility parameters.

[0021] Furthermore, the steel manufacturing process is complex due to the inherent material and energy transformation processes, including multi-component, multi-phase, multi-level, and multi-scale characteristics. The varying sets of relationships between processes further complicate the static and dynamic operational structures of the process. While progress has been made in automation and interface technologies, the lack of cross-process collaboration from a holistic perspective remains a challenge. This leads to problems such as uncertainties in transportation routes and fluctuations in molten steel quality, directly impacting enterprise production efficiency, product quality, and energy consumption. Existing optimization techniques and algorithms can improve local production efficiency and resource utilization to some extent, but the lack of a global perspective and cross-process collaboration capabilities makes it difficult to comprehensively and accurately handle the complex relationships and information transmission between processes, limiting the overall optimization effect. Therefore, in-depth research into the collaborative scheduling of the entire steel industry production process, quantitative analysis of the dynamic game relationship between electricity price fluctuations and production flexibility, and the construction of a multi-dimensional coupling model of "process-energy-grid" are urgent problems to be solved.

[0022] However, existing technologies have significant shortcomings: each production device often responds independently after receiving scheduling instructions from higher authorities, lacking a collaborative mechanism between processes. Because different devices possess drastically different dynamic response characteristics, this independent response mode easily leads to imbalances in material and energy consumption between processes, causing production rhythm conflicts and frequent equipment start-ups and shutdowns. This not only limits the overall adjustment potential of steel enterprises but may also affect production safety and efficiency due to delayed and conflicting instruction transmission, failing to effectively reduce energy costs while ensuring output. Therefore, there is an urgent need for an optimized adjustment method that can fully consider the inherent differences in equipment response characteristics and, based on this, achieve efficient multi-process collaboration.

[0023] Example 1 As an embodiment of this application, refer to Figure 1 and Figure 2 . Figure 1 The exhibition showcases the internal production process of steel enterprises, the interaction structure between self-owned power plant equipment and the external power grid. The left side shows the core processes of steel production; the right side shows the interaction between energy supply and the power grid. Gas boilers and thermal power units provide self-owned electricity / heat for steel production and are the core of internal energy supply. Figure 2 It showcases the multi-process production structure of steel enterprises.

[0024] First, clarify the production system architecture and processes of the steel enterprise, and identify the core adjustable equipment: electric arc furnace (EAF), ladle refining furnace (LF), continuous casting machine (CC), and rolling mill (MILL). Analyze the power consumption characteristics of the adjustable equipment and assess its ability to respond to grid dispatch.

[0025] (1) Electric Arc Furnace Modeling; As the core of the steelmaking process, the electric arc furnace is the most representative adjustable electrical load in steel enterprises. Its adjustment potential is mainly reflected in the rapid and wide-range adjustment of active power during the melting period through electrode raising and lowering, with an adjustment range of 30%-70% of its rated capacity. In coordinated scheduling, the electric arc furnace is more suitable for participating in short-term, rapid frequency regulation services or as a rotating reserve, rather than long-term continuous power regulation. For the electric arc furnace, a segmented time-varying model is established to identify the melting period with the greatest adjustment potential. Specifically: (1) In the formula: Let t be the total power of all electric arc furnaces. Let t be the power of the j-th electric arc furnace at time t, and m be the number of electric arc furnaces.

[0026] (2) In the formula: The moment is when the electric arc furnace is energized and the arc is ignited; The electric arc furnace is in the melting stage at all times; The electric arc furnace is in the oxidation stage at all times; The electric arc furnace is in the reduction stage at all times; The electric arc furnace will be in a stopped state after a certain time. This refers to the rated power of the electric arc furnace during operation. The coefficients represent the random power fluctuations of the electric arc furnace at different stages and times. It is a random process with zero mean and large variance (instantaneous power fluctuation).

[0027] (2) The ladle refining furnace uses the electric arc heating principle to achieve the heating of molten steel and fine-tuning of its composition. Its core heating mechanism has significant technical similarities with that of the electric arc furnace. Specifically, both utilize the high temperature of the electric arc generated between the graphite electrode and the furnace charge as a heat source. According to the "Iron and Steel Metallurgy" (Steelmaking section, Metallurgical Industry Press) and relevant industry standards, the ladle refining furnace is mainly composed of core components such as the refining furnace transformer, electrode lifting and adjusting device, and water-cooled furnace cover, and has typical submerged arc heating characteristics.

[0028] Given the similarity in structure and working principle, the segmented time-varying power modeling method and dynamic response characteristic parameter quantification method proposed in this patent for electric arc furnaces are also applicable to ladle refining furnaces in terms of power consumption characteristic modeling. The main differences between the two lie in the process stages and operating cycles: conventional electric arc furnaces are mainly used for scrap steel melting, and their operating cycles cover the melting, oxidation, and reduction periods, with drastic power fluctuations and phased changes; while ladle refining furnaces are mainly used for secondary refining of molten steel, and their operating cycles are mainly the heating and holding period, with relatively stable operating conditions.

[0029] Despite the aforementioned technological differences, both methods achieve rapid and continuous adjustment of active power through electrode lifting mechanisms, and their adjustment mechanisms are identical. Therefore, the segmented time-varying power model established in this application is compatible with the operating characteristics of the ladle refining furnace. In practical implementation, the main operating conditions of the ladle refining furnace can be mapped to the melting period in the electric arc furnace model, thereby directly reusing the dynamic response characteristic parameter quantification logic described in this patent without the need to reconstruct an independent mathematical model.

[0030] (3) Continuous casting machine modeling; In the continuous casting process system model, the billet speed is a core decision variable, and its value determines the duration of the casting operation. This duration, together with the power of the main drive system, constitutes the total energy consumption function of the continuous casting machine. Therefore, by optimizing the setting of the billet speed, production rhythm and energy costs can be effectively managed. For the continuous casting machine, a linear adjustable model is established, decomposing the total power into fixed, variable, and auxiliary parts. Specifically: (3) In the formula: Total power of all continuous casting machines The power of a single continuous casting machine. Index for the number of continuous casting machines For the fixed power consumption of the continuous casting machine, The power consumption of a continuous casting machine is variable and depends on the billet casting speed. This refers to the power consumption of the auxiliary system of the continuous casting machine. The constant coefficients, For the billet drawing speed of the continuous casting machine, , These are the upper and lower limits of the throwing speed.

[0031] The setting of the casting speed is critically constrained by the pouring temperature. The intrinsic relationship is that the pouring temperature, by influencing the growth rate of the solidified shell, determines the maximum allowable casting speed to maintain a safe shell thickness. This quantitative relationship is a core component of the continuous casting process optimization and automatic control model. That is, the casting speed is linearly constrained by the pouring temperature, expressed as: (4) In the formula: For pouring temperature, , These represent the minimum and maximum casting temperatures. d and e represent the linear constant coefficients between the casting speed and the casting temperature.

[0032] (4) Rolling mill modeling; The load of the rolling mill is essentially the high-power electric motor driving the rolls, and its core task is to plasticize the steel billet at high temperature. This plasticizing process exhibits distinct impact load characteristics: when the steel billet bites into the roll gap, the driving power of the rolling mill will rise sharply and instantaneously, forming a significant power peak; while at the moment the steel billet is thrown away, the power drops sharply. The roughing mill and finishing mill connected in series on the production line, although having different process objectives, share this dynamic physical essence. In order to accurately characterize this dynamic process and transform it into a controllable resource that can participate in the overall plant collaborative optimization, this paper adopts a composite power model that integrates the metal plastic deformation mechanism, transmission system dynamics, and operating status. This model decomposes the total power into deformation power, friction loss, and auxiliary system power consumption, and its expression can be systematically described as: (5) In the formula: For the full power of the rolling mill, Let be the power of the l-th rolling mill at time t. Metal deformation power is the main component of the work done by the rolling mill, and its value depends on the product specifications and processes. This is inherent equipment wear and tear, and is related to whether or not steel is rolled. To support the system's power, its power output fluctuates with the production schedule, starting and stopping accordingly. The overall efficiency coefficient, This is the amount of pressure applied. For strip / billet width, It is the resistance to deformation of metals, and is related to the type of steel and temperature. Let t be the rolling speed at time t.

[0033] Then, based on the model constructed above, and according to the power grid regulation requirements, its dynamic response characteristic parameters are quantified, including response time (the time required for the equipment to change power from receiving a command); interruptible / reducible duration (the longest time the equipment can suspend or reduce power operation without affecting product quality and safety); and adjustability (the range and magnitude of adjustable equipment power). By comprehensively scoring the above four indicators, different types of power grid needs can be accurately met.

[0034] Table 1 Equipment Response Characteristics of Steel Enterprises

[0035] Response time: For electric arc furnaces, the response time is based on the mechanical response time of the electrode lifting system and the delay of the control system. The electrode lifting response time is approximately 5-30 seconds, and the control system delay is approximately 2-5 seconds, resulting in a combined response time of 7-35 seconds. For continuous casting machines, the response time is based on the billet speed adjustment system, including the frequency converter response and transmission system inertia, and ranges from 30-120 seconds. For rolling mills, the response time is based on the rolling rhythm and billet bite / throw cycle, requiring the completion of the current rolling cycle before power adjustment can be performed, with a timeframe of 120-300 seconds.

[0036] The duration of interruption and the duration of reduction are measured based on the proportion of the adjustable stage to the total furnace length in the model. Electric arc furnaces are limited by the smelting process stages; the interruption time during the melting period should not exceed 15 minutes, otherwise it will affect the furnace temperature stability and smelting cycle. Continuous casting machines and rolling mills are continuous operating equipment; interruption will lead to the risk of steel leakage and cannot be interrupted. Electric arc furnaces can operate at reduced power during the melting period for approximately 30-60 minutes. Continuous casting machines can reduce power by decreasing the billet drawing speed, and rolling mills can reduce power by adjusting the rolling rhythm, for durations up to several hours.

[0037] Adjustability is measured by the proportion of adjustable power in the total power. The adjustment range of electric arc furnace can reach 30%-70% of the rated capacity. The adjustment range of continuous casting machine billet speed is limited, and the power adjustment range is about 10%-30%. Rolling mill can achieve greater power adjustment by adjusting rolling speed and reduction, with an adjustment range of 20%-80%.

[0038] Example 2 As an embodiment of this application, the technical solution implementation method for steel enterprises to participate in the collaborative optimization scheduling of the power system is disclosed.

[0039] A collaborative adjustment command transmission and cooperation mechanism is constructed. Based on the quantified equipment response characteristics shown in Table 1 of Example 1, a hierarchical cooperation mechanism is designed, including a rapid response layer, a main adjustment layer, and a flexible buffer layer. Furthermore, the optimization objective—demand-side response compensation cost—is designed, and constraint functions—electric arc furnace, continuous casting machine, and rolling mill constraints—are defined.

[0040] In a further embodiment, the fast response layer is mainly composed of an electric arc furnace (EAF). Its core quantitative characteristics are extremely short response time and a considerable range of rapid active power adjustment (30%-70% of rated capacity). However, the interruptible duration is limited by the smelting process stage, and frequent adjustments can lead to electrode losses and harmonic problems. It can be regarded as a fast-adjustment resource, which can be used as a rotating standby to maintain the instantaneous power balance of the system.

[0041] The fast response layer is based on the melting period adjustment potential and millisecond-level response capability identified in the segmented time-varying power model of the electric arc furnace. It is used to receive the first type of power adjustment instructions (real-time or second-level fast power adjustment instructions) from the dispatch center and undertakes fast adjustment tasks such as frequency regulation and standby.

[0042] In a further implementation, the main adjustment layer is represented by a rolling mill (MILL). Its core quantitative characteristics are huge adjustability and long reduction duration, but due to the inertia of the motor and transmission system, the response time is relatively slow, and the adjustment process needs to be strictly matched with the bite and throw rhythm of the billet, exhibiting a typical impact load curve.

[0043] Its composite power model accurately depicts the relationship between its power and rolling speed and billet specifications, enabling refined and predictable power shaping. This not only allows it to undertake large-scale, planned load transfers, but more importantly, through coordination with upstream continuous casting processes, it optimizes start-up and shutdown sequences, significantly improving the reliability and economy of regulation. It can be considered a stable regulation resource, responsible for handling most of the planned load adjustments within the day.

[0044] The main regulating layer is based on the accurate characterization of the relationship between power and rolling speed and billet specifications in the composite power model of the rolling mill, as well as its planned regulation capability. It is used to receive the second type of power regulation instructions (planned load adjustment instructions) from the dispatch center and undertake intraday load transfer tasks such as peak shaving and valley filling.

[0045] In a further embodiment, the flexible buffer layer is centered around a continuous casting machine (CC). Its core quantitative characteristics are a high degree of operational continuity, low amplitude and slow rate of power regulation, but its regulation can be sustained for a long time and is crucial for maintaining material balance between upstream and downstream processes.

[0046] By optimizing the billet drawing speed within the limits allowed by its process, energy consumption can be smoothly adjusted, acting as a power buffer and material buffer in the production process on timescales of tens of minutes to hours. When the fast response layer and the main adjustment layer are activated, the flexible buffer layer absorbs or releases a small amount of power by fine-tuning the drawing speed. This not only smooths out fluctuations in the total power of the entire plant but also ensures that the output of continuously cast billets dynamically matches the rolling demand, preventing material accumulation or material shortages. It can be responsible for fine-tuning and energy buffering on a minute to hourly scale, ensuring the stability and continuity of system operation.

[0047] The flexible buffer layer is based on the linear adjustable power model of the continuous casting machine and the linear constraint relationship between the billet speed and the pouring temperature. It is used to receive the third type of power adjustment instructions (minute-level to hour-level power fine-tuning instructions) from the dispatch center and undertake the tasks of material balance buffering and power smoothing.

[0048] The rapid response layer corresponds to the frequency regulation service market, the main regulation layer corresponds to the energy market and the standby market, and the flexible buffer layer ensures the stability and robustness of the entire plant when participating in the market.

[0049] Furthermore, a collaborative optimization model is established with the goal of minimizing total production cost. The optimization objective is to minimize the total system operating cost, which is divided into three parts: electricity purchase / sale cost, thermal power / gas power cost, and steel enterprise adjustment cost. The power compensation cost to incentivize steel enterprises to participate in demand response is also considered.

[0050] The objective function can be expressed as: (6) (7) In the formula: For the cost of purchasing electricity from the grid, For time-of-use electricity pricing, Power purchased from the power grid; For the cost of electricity sold by the power grid, Power sold to the power grid; For the operating costs of thermal power units, For thermal power unit i in Constant effort , and is the coal consumption characteristic coefficient of thermal power unit i; The start-up and shutdown cost of thermal power unit i; For thermal power unit i in The running state variables at any given time. =1 indicates that the device is powered on. =0 indicates that the device is powered off; For the cost of gas fuel, For gas prices, For gas turbines in Electricity generation at any given moment For the power generation efficiency of gas turbines, The calorific value of the gas. To compensate for the cost of demand-side response, The power compensation price paid by the operator to the steel company at time t; For steel companies It responds to load differences generated by the power grid in real time.

[0051] Furthermore, constraints are established to meet the normal production process and maximize the approximation to actual production.

[0052] Establish internal constraints for steel companies as follows: (1) Establishing power balance constraints: The total power consumption of the steel enterprise at time t is expressed as the sum of the power consumption of all equipment at time t: (8) In the formula: Let be the power consumption of the steel company at time t. This represents the other stable production load power of the steel enterprise at time t.

[0053] (2) Establishing storage constraints: Due to the limited warehouse storage capacity corresponding to each task, in order to balance the flexibility of load scheduling and operational safety of steel enterprises within a limited space, upper and lower limits need to be set for the storage quantity of materials: the upper limit constraint is derived from capacity, while the lower limit constraint is used to establish safety reserves to cope with emergencies. Its mathematical expression is: (9) In the formula: and These represent the lower and upper limits for material r storage, respectively.

[0054] (3) Establishing production constraints: The primary prerequisite for steel enterprises to respond to power grid dispatch is to ensure that they achieve their daily production targets. This production constraint can be expressed as: (10) In the formula: Let t be the quantity of finished steel products. This refers to the daily output requirements of steel companies.

[0055] (4) Establish runtime constraints. The equipment in steel enterprises is produced in furnace batches. To ensure the continuity of operation, each piece of equipment requires a maximum and minimum continuous operating time. Due to the different production processes, the maximum and minimum continuous operating times are also different.

[0056] (11) In the formula: The continuous operating time of the electric arc furnace. , These represent the minimum and maximum continuous operating times of the electric arc furnace, respectively.

[0057] In a further specific implementation, in addition to considering the internal constraints of steel enterprises, grid constraints are added; that is, grid power balance constraints are established: (12) In the formula: Let be the power exchange between the power grids at time t. Let t be the power generation capacity of the thermal power unit. Let t be the power generation capacity of the gas turbine. The actual value of the normal load at time t. This represents the power consumption of the steel company at time t.

[0058] Example 3 As an embodiment of this application, after the construction of the collaborative optimization model is completed, the theoretical model is transformed into executable production scheduling instructions, which specifically includes the following sub-steps: S1. Model Linearization and Data Preparation First, the collaborative optimization model established in step 3 is preprocessed. Since the cost of the thermal power unit in the objective function (Equation 7) is a quadratic term, it needs to be transformed into a mixed-integer linear programming (MILP) problem using mature techniques such as piecewise linearization for efficient solution. Simultaneously, a multi-timescale collaborative optimization objective function (corresponding to Equations 6 and 7) integrating the characteristics of both long and short processes is constructed, and all necessary input data are prepared, including: Electricity price signal: Based on the time-of-use electricity price data of large industrial users, a sequence of purchase / sale electricity prices is formed within the scheduling cycle; Equipment parameters: Physical constraint parameters such as rated power, response time, and upper and lower limits of regulation capacity of each piece of equipment are extracted from the equipment model in step 1; Production plan: The final product output requirements and the initial inventory of each intermediate material within the scheduling cycle are defined; Power grid parameters: Cost coefficients, output and ramp limits of thermal power units and gas turbines, gas prices, etc.

[0059] S2. Obtain grid demand response data in real time through the grid interface, including demand fluctuation frequency, fluctuation amplitude, and response time window, and assign the role weight of each device in the fast response layer, main regulation layer, and flexible buffer layer.

[0060] S3. Real-time acquisition of operating status data of each production equipment (electric arc furnace, continuous casting machine, rolling mill) through sensor network, including equipment load rate, current production stage (melting / oxidation / reduction, etc.), equipment response time, and adjustable capacity parameters.

[0061] S4. Identify the current production stage and dynamically update the equipment response characteristic parameters; based on the equipment operating status data, i.e. S3, determine the current production stage, dynamically update the equipment response characteristic parameters according to the current production stage, generate a dynamic response characteristic vector, and embed long-process rigid constraints and short-process flexible constraints.

[0062] The dynamic response characteristic vector of the electric arc furnace is shown below: (13) In the formula, The vector represents the dynamic response characteristics of the electric arc furnace. The dynamic response characteristic vector differs at different stages. 1 represents the dynamic response characteristic in Table 1. The coefficients represent the random power fluctuations of the electric arc furnace at different stages and times.

[0063] Rigid constraints: (14) In the formula, considering the operating characteristics of electric arc furnaces and refining furnaces, the electric arc furnace must be in a certain permissible operating state. This represents the value of the j-th electric arc furnace when it is in the o-th position at time t. The value is 1 if it is in the o-th position, and 0 otherwise.

[0064] Flexibility constraints include the number of ladle transfer furnaces and the production constraints of continuous casting machines: (15) In the formula, This indicates the quantity of molten steel produced by the converter. This represents the amount of molten steel consumed by the ladle refining furnace at time t. Let t be the amount of molten steel present in the ladle transfer furnace at time t. The maximum amount of molten steel that can be stored in a ladle transfer furnace. This indicates the amount of refined steel used at time t under the operating speed of the continuous casting machine. This represents the amount of refined steel produced by the ladle refining furnace at time t. This indicates the maximum amount of refined steel that can be processed on a continuous casting machine.

[0065] S5. Analyze the characteristics of power grid demand fluctuations and generate dynamic hierarchical decision-making instructions. First, identify the type of power grid demand signals and establish a mapping relationship between demand types and hierarchical cooperation mechanisms. Although different demand types emphasize different response levels, the three levels are always in a state of coordinated cooperation to maintain system balance. When the system detects a grid frequency deviation exceeding a threshold or an emergency safety situation such as a sudden fault, it determines the demand as an emergency safety-type demand. At this time, the fast response layer is activated as the leading response layer (or the first response layer), using its millisecond to second-level response speed to provide emergency power support. The main regulation layer and the flexible buffer layer prepare to provide subsequent energy replenishment simultaneously. When the system receives time-of-use pricing signals, peak-valley arbitrage instructions, or long-term load transfer plans, it determines the demand as an economic regulation-type demand. At this time, the main regulation layer, as the leading response layer, achieves significant load shifting and valley filling by adjusting the production rhythm. The fast response layer is responsible for smoothing high-frequency fluctuations during the regulation process, and the flexible buffer layer provides basic load support. When the grid issues medium- and long-term flexible resource allocation plans or renewable energy consumption instructions, the system determines the demand as a flexible adjustment-type demand. At this time, the flexible buffer layer, as the leading response layer, uses its long-duration regulation capability to adapt to slowly changing power curves. The fast response layer and the main regulation layer cooperate to fine-tune to ensure that production continuity is not affected.

[0066] The instructions not only include power setpoints, but also, for independent instructions not set by equipment, adjustments are made according to the specific smelting stage window for electric arc furnaces; for rolling mills, the start-up time is delayed and the heating furnace insulation strategy is compensated; and for continuous casting, the casting speed curve is dynamically adjusted to match changes in the upstream steel output rhythm. Based on the analysis results of power grid demand fluctuation characteristics and the operating status data of production equipment acquired by sensors, namely S2 and S3, dynamic hierarchical decision instructions are generated to redistribute the role weights of each piece of equipment in the rapid response layer, the main adjustment layer, and the flexible buffer layer.

[0067] S6. Model Solving and Solution Generation The preprocessed MILP model, all input data, and solver configuration (constraints, objective function, initial values, time interval accuracy, maximum solver time, and role weights at different levels) are imported into the optimization solver environment. This invention preferably uses the Yalmip toolbox as the modeling language and calls the Cplex solver for solving. The solver will output a complete, time-discrete optimal scheduling scheme, which includes: Optimal power settings for each key production equipment (EAF, CC, MILL, etc.) during each scheduling period; Optimal output of self-owned thermal power units and gas turbines at each time period; The optimal power purchase / sale for steel companies and the external power grid in each time period.

[0068] S7. Scheme Analysis and Command Issuance The raw data output by the solver is parsed and its format converted. For continuous casting machines, the optimal power needs to be back-mapped to specific billet drawing speed commands; for rolling mills, the total power curve needs to be decomposed into specific start / stop sequences for roughing and finishing rolling processes. Finally, a standardized set of scheduling instructions for each equipment control system is generated and distributed in real time to the field controllers through the steel company's Energy Management System (EMS) or Manufacturing Execution System (MES).

[0069] S8. Execution Monitoring and Rolling Optimization During the execution of the scheduling plan, the system continuously collects the actual operating data of each device and compares it with the optimized plan. If the actual state deviates significantly from the plan due to unforeseen circumstances—such as equipment failure or order changes—the rolling optimization mechanism is triggered: taking the current moment as a new starting point, based on the latest system state and future forecast information, steps S6-S8 are re-executed to generate and distribute an updated rolling scheduling plan. The flexible buffer layer automatically compensates, the main adjustment layer redistributes the load, and the initial state and constraint boundaries of the optimization model for the next time period are updated. This ensures the robustness and adaptability of the entire coordinated adjustment process.

[0070] Finally, the mixed-integer linear programming model was solved using the Yalmip toolbox and the Cplex solver. The solution results are as follows: Figure 3 , Figure 4 , Figure 5 As shown in Table 2, the electricity prices are as follows. During peak electricity price periods, the system, through a collaborative mechanism, prioritizes the scheduling of loads with strong adjustability that can be reduced, and utilizes the rapid response capability of the electric arc furnace for fine-tuning, thereby reducing high-priced electricity purchases and even increasing electricity sales to the grid to obtain economic benefits. During off-peak electricity price periods, the system ensures the stable operation of basic production. The simulation data in Table 3 verifies the effectiveness of this method; the total operating cost of Scenario 2 is significantly lower than that of Scenario 1.

[0071] Table 2 Time-of-use Electricity Prices for Large Industrial Users

[0072] Table 3. Economic Costs of Industrial Park Operation

[0073] As an embodiment of this application, a multi-process collaborative optimization and adjustment system for steel enterprises is disclosed. Employing the specific implementation method described above, the system includes: Adjustable equipment modeling module; used to identify adjustable equipment in the steelmaking process based on the production architecture and process flow of steel enterprises; to model the power consumption characteristics of adjustable equipment and quantify the dynamic response characteristic parameters of adjustable equipment; the power consumption characteristic modeling includes electric arc furnace modeling, ladle refining furnace modeling, continuous casting machine modeling and rolling mill modeling; The collaborative optimization design module is used to design a hierarchical collaborative mechanism based on the dynamic response characteristic parameters, including a fast response layer, a main adjustment layer, and a flexible buffer layer; and to design a collaborative optimization model and define constraint functions based on the cost characteristics of the hierarchical division. The preprocessing module is used to perform piecewise linearization on the collaborative optimization model, transforming it into a mixed-integer linear programming model; and to acquire power grid demand response signals and the operating status of production equipment within steel enterprises in real time, updating the dynamic response characteristic parameters and constraint functions according to the current production stage. The scheduling module is used to analyze the types of power grid demand fluctuations, activate the corresponding levels in the hierarchical cooperation mechanism according to the fluctuation types, and reallocate the role weights of adjustable equipment in different levels. The mixed integer linear programming model, the updated dynamic response characteristic parameters, the constraint functions, and the role weights of the production equipment are imported into the optimization solver to output the scheduling scheme.

[0074] As an embodiment of this application, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded onto the processor, it employs the specific implementation described in the multi-process collaborative optimization and adjustment method above.

[0075] As an embodiment of this application, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, adopts the specific embodiments described above.

[0076] 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 multi-process collaborative optimization and adjustment method for steel enterprises, characterized in that, include: Based on the production structure and process flow of steel enterprises, identify the adjustable equipment in the steelmaking process; Power consumption characteristics modeling is performed for adjustable equipment, and the dynamic response characteristic parameters of the adjustable equipment are quantified; the power consumption characteristics modeling includes electric arc furnace modeling, ladle refining furnace modeling, continuous casting machine modeling, and rolling mill modeling; A hierarchical collaboration mechanism is designed based on the dynamic response characteristic parameters, including a fast response layer, a main adjustment layer, and a flexible buffer layer; a collaborative optimization model is designed based on the cost characteristics of the hierarchical division, and constraint functions are defined. The collaborative optimization model is piecewise linearized and transformed into a mixed-integer linear programming model; the power grid demand response signal and the operating status of the steel enterprise's internal production equipment are acquired in real time, and the dynamic response characteristic parameters and constraint functions are updated according to the current production stage. Analyze the types of power grid demand fluctuations, activate the corresponding levels in the hierarchical cooperation mechanism based on the fluctuation types, and reallocate the role weights of adjustable equipment in different levels. The mixed-integer linear programming model, the updated dynamic response characteristic parameters, the constraint functions, and the role weights of the production equipment are imported into the optimization solver to output a scheduling scheme.

2. The multi-process collaborative optimization and adjustment method for steel enterprises according to claim 1, characterized in that, The electric arc furnace modeling constructs a segmented time-varying power model based on the electric arc furnace operation stages to identify the optimal melting period adjustment; the ladle refining furnace modeling reuses the segmented time-varying power model. The continuous casting machine modeling uses a linear adjustable power model to decompose the total power of the continuous casting machine and optimizes the billet drawing speed by combining the linear constraint relationship between the drawing speed and the pouring temperature. The rolling mill modeling uses a composite power model to decompose the total power of the rolling mill and calculates the adjustable load of the rolling mill by combining the rolling parameters.

3. The multi-process collaborative optimization and adjustment method for steel enterprises according to claim 1, characterized in that, The dynamic response characteristic parameters include response time, interruptible duration, reduceable duration, and adjustability. The response time is the time required for the device to change power from receiving a command; the interruptible duration is the longest time the device can pause operation; the reduceable duration is the longest time the device can operate at reduced power; and the adjustability is the range and magnitude of the device's power adjustment.

4. The multi-process collaborative optimization and adjustment method for steel enterprises according to claim 1, characterized in that, The dynamic response characteristic parameters are quantified according to the power grid control requirements, wherein: The response time of the electric arc furnace is calculated by combining the mechanical response time of the electrode lifting system and the delay of the control system; the response time of the continuous casting machine is calculated based on the response time of the billet speed adjustment system; and the response time of the rolling mill is calculated based on the rolling rhythm and the billet bite and throw cycle.

5. The multi-process collaborative optimization and adjustment method for steel enterprises according to claim 1, characterized in that, In the aforementioned hierarchical collaboration mechanism: The fast response layer is based on the optimal adjustment melting period of the electric arc furnace and is used to receive and execute the first type of power regulation command. The main regulating layer is based on the adjustable load output by the composite power model of the rolling mill, and is used to receive and execute the second type of power regulation command; The flexible buffer layer is based on the billet pulling speed output by the adjustable power model of the continuous casting machine, and is used to receive and execute the third type of power adjustment command.

6. The multi-process collaborative optimization and adjustment method for steel enterprises according to claim 1, characterized in that, Based on the cost characteristics of the different levels, design optimization objectives and define constraint functions, including: A collaborative optimization model is established with the goal of minimizing total production cost. The optimization objective is to minimize the total operating cost of the system. The total operating cost includes the cost of purchasing and selling electricity, the cost of thermal power and gas, the adjustment cost of steel enterprises, and the compensation cost. The compensation cost is expressed as the product of the load difference generated by the steel enterprise's response to the power grid during the dispatch period and the power compensation price paid by the operator to the steel enterprise.

7. A multi-process collaborative optimization and adjustment method for steel enterprises according to any one of claims 1 or 6, characterized in that, Based on the cost characteristics of the different levels, the optimization objective is designed and constraint functions are defined, which also includes: The constraint functions include internal constraints of steel enterprises and power grid power balance constraints. Internal constraints of steel enterprises include internal power balance constraints, storage constraints, output constraints, and operating time constraints.

8. The multi-process collaborative optimization and adjustment method for steel enterprises according to claim 1, characterized in that, The process of acquiring the power grid demand response signal includes: Real-time demand response data of the power grid is obtained through the power grid interface, including the frequency, amplitude and response time window of demand fluctuations, and the role weight of each device in the fast response layer, the main regulation layer and the flexible buffer layer are assigned.

9. The multi-process collaborative optimization and adjustment method for steel enterprises according to claim 1, characterized in that, The analysis of power grid demand fluctuation types, and the activation of the corresponding level in the hierarchical cooperation mechanism based on the fluctuation type, include: Type identification of power grid demand signals, and establishment of a mapping relationship between demand types and hierarchical cooperation mechanisms; When the grid frequency deviation is detected to exceed the threshold and / or a fault occurs, the grid demand fluctuation type is determined to be an emergency safety demand; in the hierarchical cooperation mechanism, the fast response layer is activated as the first response level; When receiving time-of-use pricing signals, peak-valley price arbitrage instructions, and / or long-term load transfer plans, the grid demand fluctuation type is determined to be economically regulated demand; in the hierarchical cooperation mechanism, the main regulation layer is activated as the first response level. When the power grid issues a resource allocation plan and / or a renewable energy consumption instruction, the power grid demand fluctuation type is determined to be a flexible adjustment type of demand; in the hierarchical cooperation mechanism, the flexible buffer layer is activated as the first response level.

10. A multi-process collaborative optimization and adjustment system for steel enterprises, executing the multi-process collaborative optimization and adjustment method as described in any one of claims 1-9, characterized in that, The system includes: Adjustable equipment modeling module; used to identify adjustable equipment in the steelmaking process based on the production architecture and process flow of steel enterprises; to model the power consumption characteristics of adjustable equipment and quantify the dynamic response characteristic parameters of adjustable equipment; the power consumption characteristic modeling includes electric arc furnace modeling, ladle refining furnace modeling, continuous casting machine modeling and rolling mill modeling; The collaborative optimization design module is used to design a hierarchical collaborative mechanism based on the dynamic response characteristic parameters, including a fast response layer, a main adjustment layer, and a flexible buffer layer; and to design a collaborative optimization model and define constraint functions based on the cost characteristics of the hierarchical division. The preprocessing module is used to perform piecewise linearization on the collaborative optimization model, transforming it into a mixed-integer linear programming model; and to acquire power grid demand response signals and the operating status of production equipment within steel enterprises in real time, updating the dynamic response characteristic parameters and constraint functions according to the current production stage. The scheduling module is used to analyze the types of power grid demand fluctuations, activate the corresponding levels in the hierarchical cooperation mechanism according to the fluctuation types, and reallocate the role weights of adjustable equipment in different levels. The mixed integer linear programming model, the updated dynamic response characteristic parameters, the constraint functions, and the role weights of the production equipment are imported into the optimization solver to output the scheduling scheme.

11. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the computer program is loaded into the processor, it implements the multi-process collaborative optimization and adjustment method according to any one of claims 1-9.

12. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the multi-process collaborative optimization and adjustment method according to any one of claims 1-9.

Images