A steel enterprise demand hybrid prediction and collaborative optimization method and system based on production plan fusion
By combining a production plan-based demand hybrid forecasting method with 5G networks and mechanistic models, the problems of forecast accuracy and real-time control in the power demand management of steel enterprises have been solved, achieving efficient demand control and grid dispatch response, and significantly reducing production costs.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- AUTOMATION RES & DESIGN INST OF METALLURGICAL IND
- Filing Date
- 2026-02-11
- Publication Date
- 2026-06-05
AI Technical Summary
Steel companies lack systematic, predictive, and real-time control capabilities in electricity demand management. Traditional forecasting models struggle to accurately capture nonlinear and abrupt load characteristics, resulting in insufficient load forecasting accuracy and an inability to achieve precise demand control and grid dispatch response.
A demand-mixed forecasting method based on production planning integration is adopted. Power system data is collected through 5G mobile network and wired network, and the load is classified and predicted by combining the mechanism model. A Gantt chart linked with the load is generated for visualization, and an adjustment strategy for power consumption process is generated according to the optimization and adjustment strategy.
It significantly improves the accuracy and practicality of load forecasting, achieving full-process coverage from pre-forecasting to in-process control, reducing the monthly peak impact load by more than 8%, and forming a complete demand control management system.
Smart Images
Figure CN122155004A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of planning and scheduling technology, and in particular to a method and system for mixed forecasting and collaborative optimization of demand in steel enterprises based on production planning integration. Background Technology
[0002] In recent years, with the increasing demands for energy conservation and carbon reduction in the steel industry, the importance of energy cost control for steel enterprises has become increasingly prominent. Electricity is an essential energy source for the production process of steel enterprises; however, traditional electricity management models rely on manual experience, making it difficult to cope with complex production scheduling and real-time changes in electricity prices, resulting in room for optimization of electricity costs for steel enterprises.
[0003] Against the backdrop of the ongoing deepening of electricity market reforms, the two-part tariff is a commonly used billing method in the electricity market, which divides electricity charges into two parts: a per-unit charge and a basic charge. The per-unit charge is the fee paid by the user for the total electricity consumed. It is calculated based on the total kilowatt-hours (kWh) used by the user within a billing cycle (usually one month). One of the main billing methods for the basic charge is based on maximum demand. This method uses the user's maximum demand for the month as a benchmark and calculates the basic charge at an agreed-upon unit capacity price. Demand refers to the average power measured within a specific interval (usually 15 minutes); while maximum demand (MD) refers to the maximum value among all demand values within a billing cycle (usually one month).
[0004] With the continuous optimization of the electricity market mechanism, demand management has transformed from an auxiliary management tool into a core component of enterprise energy management. However, most steel companies remain in a passive response phase, resulting in a lack of systematic, predictive, and real-time control capabilities, making it difficult to achieve precise demand control and respond to grid dispatch demands. The root cause lies in two aspects: First, steel production involves numerous impact loads, such as the dramatic fluctuations in the start-up, shutdown, and operating power of equipment like electric furnaces and rolling mills. Traditional general-purpose forecasting models struggle to accurately capture these nonlinear and abrupt characteristics, leading to insufficient load forecasting accuracy. Second, existing demand management systems primarily focus on post-event monitoring, failing to deeply integrate with production plans. This hinders proactive dynamic optimization and real-time closed-loop control during operations, thus limiting management efficiency. Furthermore, traditional demand management relies heavily on manual experience for load estimation and control. While these methods offer some flexibility and interpretability, they often exhibit lag, difficulty in quantification, and significant limitations imposed by personnel experience when facing multi-variable, high-frequency impact load scenarios.
[0005] Therefore, there is an urgent need to build a comprehensive demand management system covering the entire process from "pre-event forecasting to in-event control to post-event analysis." The main objective of this invention is to provide steel enterprises with a new type of demand management system that is highly accurate, practical, and can be integrated with production. Summary of the Invention
[0006] Based on the above analysis, the present invention aims to provide a method and system for mixed forecasting and collaborative optimization of electricity demand in steel enterprises based on production planning integration, so as to accurately predict electricity demand and save production costs.
[0007] On the one hand, this invention provides a method for mixed forecasting and collaborative optimization of steel enterprise demand based on production planning integration, including the following steps:
[0008] S1: Use 5G mobile network and wired network to collect power system electricity consumption data, and classify and predict the power load based on the mechanism model; S2: Time series overlay forecast results, estimating demand values for future periods; S3: Synchronize user production plans and generate Gantt charts linked to electricity load for visualization; S4: Generate an electricity consumption process adjustment strategy based on the optimization and adjustment strategy.
[0009] Furthermore, the electrical load includes electricity consumption load, fluctuating load, and generation load; The electrical load refers to the collection of all production units within the factory area; The fluctuating load refers to the collection of auxiliary power consumption units; The power generation load refers to the units within the plant area that have the ability to generate electricity by connecting to the grid.
[0010] Furthermore, the electrical load is a stable load and an impulse load; The stable load refers to production equipment with a flat power curve, a power fluctuation rate of less than 10% per minute, and continuous operation. The impact load refers to production equipment whose power fluctuates drastically and intermittently within seconds / minutes, with a volatility often >50%. The impact load includes refining furnaces, hot rolling production lines, and long product production lines.
[0011] Furthermore, the method for predicting the electricity consumption of the refining furnace load is as follows: First, the calculation basis for different refining furnaces is obtained based on the production plan. During the power-on period [ [Inside, the power consumption of different refining furnaces] Take its calculation basis During non-power-off periods, its electrical load is 0; this is expressed as: .
[0012] Furthermore, the method for predicting the electricity consumption of the hot rolling line load is as follows: Based on the production plan Gantt chart in the Manufacturing Execution System (MES), the rolling cycle is... The rolling process is divided into n steel billets; for any steel billet i, its power state is defined as follows: steel biting period From the time the steel billet is recorded as leaving the heating furnace, a fixed process time is maintained; during this period, the load power remains at a constant high value. ; Non-steel-biting period The time interval within the same rolling cycle that does not fall under any billet biting period; during this period, the load power is a constant low value. ; Based on the above power model, the formula for calculating the power consumption of the hot rolling line load for a rolling cycle containing n steel billets is as follows: .
[0013] Furthermore, the method for predicting the electricity consumption of long product production lines is as follows: Based on the production plan Gantt chart in the Manufacturing Execution System (MES), the rolling cycle of each production line is defined, and the power status is defined as follows: Production period This refers to the continuous time period from the start to the end of the rolling mill, as determined by the production Gantt chart; during this period, the load power remains constant at the rated empirical value of the production line. ; Non-production periods: All time outside of the planned rolling cycle, during which the load power is zero; Based on the above power model, the formula for calculating the total electricity consumption E of any long product production line within the prediction range is as follows: .
[0014] Furthermore, the power generation load is composed of gas power generation, saturated steam power generation, and other power generation loads. The calculation method for the power generation of a saturated steam power generation load is as follows: For the i-th converter out of N converters, its power generation in one smelting cycle... Average power generation within for:
[0015] The steel production of the converter during the cycle, in tons (t). The smelting cycle of the converter, in hours (h); The steam-to-electricity conversion coefficient, in kWh / t. Total power generation The power of each converter was obtained by superimposing it over time:
[0016] The effective time for the i-th converter to be in the above smelting state within the prediction window.
[0017] Furthermore, purchased electricity load Calculated using the following mathematical model:
[0018] in, Indicates the predetermined calculation period; This represents the sum of electricity consumption of all stable loads within the calculation period τ; This represents the sum of electricity consumption for all impact loads within the calculation period τ. This represents the sum of electricity consumption of all fluctuating loads within the calculation period τ; This represents the sum of the power generation of all generating loads within the calculation period τ.
[0019] Furthermore, based on the purchased electricity load The slip method is used to calculate the electricity demand, and the calculation method is as follows:
[0020] in, , For the start and end times of the sliding window, This represents the demand value at the end of the sliding window operation. This refers to the length of the sliding window.
[0021] On the other hand, the present invention provides a steel enterprise demand hybrid forecasting and collaborative optimization system based on production planning integration, for implementing the method described in the present invention; The system includes a data acquisition module, a power load forecasting module, a maximum demand forecasting module, and an optimization and adjustment strategy module. The data acquisition module is used to collect electricity consumption data and production plan data at various metering points within the factory area in real time using 5G mobile network and wired network. The power load forecasting module is used to perform sub-forecasts of future power loads for stable loads, impact loads, and fluctuating loads, and to generate a forecast curve of the total purchased power load of the entire plant by accumulating the results of each sub-forecast. The maximum demand prediction module is used to predict the maximum demand value within a future target period on a rolling basis. The optimization and adjustment strategy module is used to generate scheduling instructions based on the controllability characteristics of various loads when the maximum demand forecast exceeds the preset demand threshold, so as to collaboratively suppress the final maximum demand.
[0022] Compared with the prior art, the present invention can achieve at least one of the following beneficial effects: 1. This invention innovatively constructs a production planning hybrid forecasting architecture, which effectively solves the industry-wide technical challenges of traditional single forecasting models in the face of multiple processes and complex load fluctuations in steel production, and significantly improves the robustness and adaptability of the forecasting system.
[0023] 2. This invention elevates forecasting from a purely data-driven level to a production and operational level, making the forecast results more closely aligned with actual production, thereby greatly improving the accuracy and practicality of forecasting.
[0024] 3. The integrated intelligent decision-making system provided by this invention realizes full-process coverage from "monitoring" to "early perception" and then to "demand over-limit suggestion", forming a complete demand control management.
[0025] 4. This invention closely integrates with the specific scenarios and core processes of steel production, possessing a high degree of professionalism and irreplaceability, resulting in significant technical effects. Practical application in a steel enterprise demonstrates that adopting this invention's system can reduce the maximum monthly impact load by more than 8%.
[0026] In this invention, the above-described technical solutions can be combined with each other to achieve more preferred combinations. Other features and advantages of this invention will be set forth in the following description, and some advantages may become apparent from the description or be learned by practicing the invention. The objects and other advantages of this invention can be realized and obtained from what is particularly pointed out in the description and drawings. Attached Figure Description
[0027] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts.
[0028] Figure 1 This is a logic diagram of the demand control system; Figure 2 This is a structural diagram of the prediction method; Figure 3 This is a schematic diagram for calculating electricity demand. Figure 4 A diagram illustrating the prediction results. Detailed Implementation
[0029] The preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which constitute a part of the present invention and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.
[0030] In recent years, with the increasing demands for energy conservation and carbon reduction in the steel industry, the importance of energy cost control for steel enterprises has become increasingly prominent. Electricity is an essential energy source for the production process of steel enterprises; however, traditional electricity management models rely on manual experience, making it difficult to cope with complex production scheduling and real-time changes in electricity prices, resulting in room for optimization of electricity costs for steel enterprises.
[0031] Against the backdrop of the ongoing deepening of power market reforms, two-part tariffs are a commonly used billing method in the electricity market. As the power market mechanism continues to optimize, demand management has shifted from a supplementary management tool to a core component of enterprise energy management. However, most steel companies remain in a passive response phase, resulting in a lack of systematic, predictive, and real-time control capabilities, making it difficult to achieve precise demand control and respond to grid dispatch demands. This, in turn, leads to higher electricity demand and increased production costs.
[0032] Therefore, this invention provides a method for mixed demand forecasting and collaborative optimization of steel enterprises based on production planning integration, comprising the following steps: S1: Use 5G mobile network and wired network to collect power system electricity consumption data, and classify and predict the power load based on the mechanism model; S2: Time series overlay forecast results, estimating demand values for future periods; S3: Synchronize user production plans and generate Gantt charts linked to electricity load for visualization; S4: Generate an electricity consumption process adjustment strategy based on the optimization and adjustment strategy.
[0033] Compared with existing technologies, this invention innovatively constructs a production planning-based hybrid forecasting architecture, effectively solving the industry-wide technical challenges of traditional single forecasting models in dealing with the multi-process and complex load fluctuations of steel production, and significantly improving the robustness and adaptability of the forecasting system. Furthermore, it elevates forecasting from a purely data-driven level to a production and operational level, making the forecast results more closely aligned with actual production, thereby greatly improving the accuracy and practicality of the forecasts.
[0034] The integrated intelligent decision-making system provided by this invention achieves full-process coverage from "monitoring" to "early perception" and then to "demand overrun suggestions," forming a complete demand control and management system. Closely integrated with the specific scenarios and core processes of steel production, its technical effects are significant. Practical application in a steel company shows that after adopting this system, the monthly peak impact load can be reduced by more than 8%.
[0035] Specifically, the electrical load includes electricity consumption load, fluctuating load, and generation load; The electrical load refers to the collection of all production units within the factory area; The fluctuating load refers to the collection of auxiliary power consumption units; The power generation load refers to the units within the plant area that have the ability to generate electricity by connecting to the grid.
[0036] It should be noted that in this invention, electrical load is divided into power consumption load, fluctuating load, and power generation load. Power consumption load refers to the collection of production units within the plant area, mainly including stable load and impact load. Stable load refers to production equipment with a smooth power curve, a power fluctuation rate of <10% per minute, and continuous operation, such as blast furnaces, sintering machines, pelletizing plants, oxygen generators, air compressors, and raw material warehouses. Impact load refers to production equipment with drastic, intermittent power fluctuations on a second / minute scale, often exceeding 50%. Impact load includes refining furnaces, hot rolling lines, and long product lines.
[0037] Fluctuating loads refer to a collection of auxiliary power units, mainly consisting of low-power electrical equipment and loads whose operating modes are not highly coupled with the core production processes. These mainly include factory lighting, circulating water pumps, ventilation systems, and non-production office power.
[0038] The power generation load refers to the units within the plant area that have the capacity to generate electricity connected to the grid. These units utilize waste heat, waste pressure, or solar energy to feed electricity into the grid. The power generation capacity ranges from 0 to several megawatts and can be further subdivided according to the energy source, including gas-fired power generation load, saturated steam power generation load, and other power generation loads. The power generation load varies depending on the specific production unit; some units do not have power generation units, so their power generation load is 0.
[0039] Specifically, the method for predicting the electricity consumption of refining furnaces is as follows: First, the calculation basis for different refining furnaces is obtained based on the production plan. During the power-on period [ [Inside, the power consumption of different refining furnaces] Take its calculation basis During non-power-off periods, its electrical load is 0; this can be expressed as: .
[0040] in This is the power-on start time; This is the time when the power-on process ends.
[0041] Specifically, the method for predicting the electricity consumption of hot rolling lines is as follows: Based on the production plan Gantt chart in the Manufacturing Execution System (MES), the rolling cycle is... The rolling process is divided into n steel billets; for any steel billet i, its power state is defined as follows: steel biting period From the time the steel billet is recorded as leaving the heating furnace, a fixed process time is maintained; during this period, the load power remains at a constant high value. ; Non-steel-biting period The time interval within the same rolling cycle that does not fall under any billet biting period; during this period, the load power is a constant low value. ; Based on the above power model, the formula for calculating the power consumption of the hot rolling line load for a rolling cycle containing n steel billets is as follows: .
[0042] Specifically, the method for predicting the electricity consumption of long product production lines is as follows: Based on the production plan Gantt chart in the Manufacturing Execution System (MES), the rolling cycle of each production line is defined, and the power status is defined as follows: Production period This refers to the continuous time period from the start to the end of the rolling mill, as determined by the production Gantt chart; during this period, the load power remains constant at the rated empirical value of the production line. ; Non-production periods: All time outside of the planned rolling cycle, during which the load power is zero; Based on the above power model, the formula for calculating the total electricity consumption E of any long product production line within the prediction range is as follows: .
[0043] It should be noted that for fluctuating loads, although the individual power of such loads is not large, the total amount is considerable. To achieve effective prediction of their electricity consumption, this invention adopts a data-driven rolling prediction model. The training and updating of this model rely on the steel plant's historical electricity consumption data for the previous three months. This time window was chosen because three months of data can effectively cover different production intensities and seasonal cycles, ensuring the representativeness of the learning samples, while avoiding model lag caused by too early historical data. This allows for accurate capture of short-term patterns of fluctuating loads (such as diurnal and weekly cycles), and the rolling mechanism continuously optimizes the prediction accuracy of future short-term loads.
[0044] Calculation steps: Collect and store the steel plant's load history data for the past three months, organized by eight-hour shifts, including: shift type identifier: morning shift, afternoon shift, night shift; timestamp (accurate to 15 minutes); actual load value; date type: ordinary weekday, weekend, holiday; and seasonal identifier.
[0045] The prediction model is based on the following calculation formula: F(t) = w1× A1(t) + w2× A2(t) + w3× M(t) in: F(t) is the predicted load value for the predicted shift at time t; A1(t) is the actual load value of the most recent similar shift at time t; A2(t) represents the actual load value of the same type of shift at time t last week; M(t) represents the average load value of the same period of the past three months at time t; w1, w2, and w3 are weighting coefficients that satisfy w1 + w2 + w3 = 1.
[0046] The weighting coefficients are determined based on the principle of time proximity: w1 takes values of 0.4-0.6, w2 takes values of 0.2-0.4, and w3 takes values of 0.1-0.3.
[0047] Specifically, the power generation load is composed of gas power generation, saturated steam power generation, and other power generation loads. The calculation method for the power generation of a saturated steam power generation load is as follows: For the i-th converter out of N converters, its power generation in one smelting cycle... Average power generation within for:
[0048] The steel production of the converter during the cycle, in tons (t). The smelting cycle of the converter, in hours (h); The steam-to-electricity conversion coefficient, in kWh / t. Total power generation The power of each converter was obtained by superimposing it over time:
[0049] The effective time for the i-th converter to be in the above smelting state within the prediction window.
[0050] It should be noted that the power generation of a saturated steam generator set is proportional to the total saturated steam flow rate produced by each converter per unit time. The steam flow rate is directly related to the smelting cycle and single-furnace output of the converter. Therefore, the total power generation can be characterized as the sum of the steam power contributed by each converter.
[0051] It should be noted that the power generation of the gas-fired power generation load is obtained directly from the linked energy system in real time.
[0052] The power generation for other power generation loads is calculated by multiplying the predetermined equipment load empirical value by its operating time, which is determined based on the actual operating status of the production equipment.
[0053] Specifically, purchased electricity load Calculated using the following mathematical model:
[0054] in, Indicates the predetermined calculation period; This represents the sum of electricity consumption of all stable loads within the calculation period τ; This represents the sum of electricity consumption for all impact loads within the calculation period τ. This represents the sum of electricity consumption of all fluctuating loads within the calculation period τ; This represents the sum of the power generation of all generating loads within the calculation period τ.
[0055] Based on purchased electricity load The slip method is used to calculate the electricity demand, and the calculation method is as follows:
[0056] in, , For the start and end times of the sliding window, This represents the demand value at the end of the sliding window operation. This refers to the length of the sliding window.
[0057] It should be noted that, referring to Figure 3 When calculating demand, the system is based on the purchased electricity load. The collection of real data, and the forecast of the purchased electricity load for the next 8 hours. The prediction results are sorted according to a fixed sampling interval (usually 1 minute), resulting in a series of data points arranged in chronological order. Then, in a given length... Within a sliding window (typically 15 minutes), there are 15 sampling points. Then, the purchased electricity load... Perform integration, then take the average, and denote it as... Demand value at the end For example, the demand for purchased electricity from 0:01 to 0:15 is calculated as the demand value at time 0:15. Then, slide the window step forward. (Usually 1 minute), calculate the next one again. The demand value within (i.e., 0:02-0:16).
[0058] Reference Figure 4 Because electricity prices are lower during off-peak hours, steel companies often activate multiple power loads simultaneously during these periods to save production costs, leading to a sudden surge in maximum demand. To address this, this system, based on a predictive model, can simulate and adjust the production sequence of the refining furnace before demand exceeds the control limit, delaying production for a period. It simultaneously generates a demand curve showing the "simulated adjustment" and visualizes the control effect, providing operators with precise decision support and achieving closed-loop control of maximum demand.
[0059] This invention also provides a steel enterprise demand hybrid forecasting and collaborative optimization system based on production planning integration, used in the method described in this invention; The system includes a data acquisition module, a power load forecasting module, a maximum demand forecasting module, and an optimization and adjustment strategy module. The data acquisition module is used to collect electricity consumption data and production plan data at various metering points within the factory area in real time using 5G mobile network and wired network. The power load forecasting module is used to perform sub-forecasts of future power loads for stable loads, impact loads, and fluctuating loads, and to generate a forecast curve of the total purchased power load of the entire plant by accumulating the results of each sub-forecast. The maximum demand prediction module is used to predict the maximum demand value within a future target period on a rolling basis. The optimization and adjustment strategy module is used to generate scheduling instructions based on the controllability characteristics of various loads when the maximum demand forecast exceeds the preset demand threshold, so as to collaboratively suppress the final maximum demand.
[0060] The integrated intelligent decision-making system provided by this invention achieves full-process coverage from "monitoring" to "early perception" and then to "demand overrun suggestions," forming a complete demand control management system. Practical application in a steel company shows that after adopting the system of this invention, the monthly peak impact load can be reduced by more than 8%.
[0061] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.
Claims
1. A method for mixed demand forecasting and collaborative optimization of steel enterprises based on production planning integration, characterized in that, Includes the following steps: S1: Use 5G mobile network and wired network to collect power system electricity consumption data, and classify and predict the power load based on the mechanism model; S2: Time series overlay forecast results, estimating demand values for future periods; S3: Synchronize user production plans and generate Gantt charts linked to electricity load for visualization; S4: Generate an electricity consumption process adjustment strategy based on the optimization and adjustment strategy.
2. The method for mixed demand forecasting and collaborative optimization of steel enterprises based on production planning integration as described in claim 1, characterized in that, The electrical load includes electricity consumption load, fluctuating load, and generation load; The electrical load refers to the collection of all production units within the factory area; The fluctuating load refers to the collection of auxiliary power consumption units; The power generation load refers to the units within the plant area that have the ability to generate electricity by connecting to the grid.
3. The method for mixed demand forecasting and collaborative optimization of steel enterprises based on production planning integration as described in claim 2, characterized in that, The electrical loads are classified as stable loads and impulse loads; The stable load refers to production equipment with a flat power curve, a power fluctuation rate of less than 10% per minute, and continuous operation. The impact load refers to production equipment whose power fluctuates drastically and intermittently within seconds / minutes, with a volatility often >50%. The impact load includes refining furnaces, hot rolling production lines, and long product production lines.
4. The method for mixed demand forecasting and collaborative optimization of steel enterprises based on production planning integration as described in claim 3, characterized in that, The method for predicting the electricity consumption of refining furnaces is as follows: First, the calculation basis for different refining furnaces is obtained based on the production plan. During the power-on period [ [Inside, the power consumption of different refining furnaces] Take its calculation basis During non-power-off periods, its electrical load is 0; this is expressed as: 。 5. The method for mixed demand forecasting and collaborative optimization of steel enterprises based on production planning integration as described in claim 3, characterized in that, The method for predicting the electricity consumption of hot rolling line load is as follows: Based on the production plan Gantt chart in the Manufacturing Execution System (MES), the rolling cycle is... The rolling process is divided into n steel billets; for any steel billet i, its power state is defined as follows: steel biting period From the time the steel billet is recorded as leaving the heating furnace, a fixed process time is maintained; during this period, the load power remains at a constant high value. ; Non-steel-biting period The time interval within the same rolling cycle that does not fall under any billet biting period; during this period, the load power is a constant low value. ; Based on the above power model, the formula for calculating the power consumption of the hot rolling line load for a rolling cycle containing n steel billets is as follows: 。 6. The method for mixed demand forecasting and collaborative optimization of steel enterprises based on production planning integration as described in claim 3, characterized in that, The method for predicting the electricity consumption of long product production lines is as follows: Based on the production plan Gantt chart in the Manufacturing Execution System (MES), the rolling cycle of each production line is defined, and the power status is defined as follows: Production period This refers to the continuous time period from the start to the end of the rolling mill, as determined by the production Gantt chart; during this period, the load power remains constant at the rated empirical value of the production line. ; Non-production periods: All time outside of the planned rolling cycle, during which the load power is zero; Based on the above power model, the formula for calculating the total electricity consumption E of any long product production line within the prediction range is as follows: 。 7. The method for mixed demand forecasting and collaborative optimization of steel enterprises based on production planning integration as described in claim 1, characterized in that, The power generation load consists of gas power generation, saturated steam power generation, and other power generation loads. The calculation method for the power generation of a saturated steam power generation load is as follows: For the i-th converter out of N converters, its power generation in one smelting cycle... Average power generation within for: The steel production of the converter during the cycle, in tons (t). The smelting cycle of the converter, in hours (h); The steam-to-electricity conversion coefficient, in kWh / t. Total power generation The power of each converter was obtained by superimposing it over time: The effective time for the i-th converter to be in the above smelting state within the prediction window.
8. The method for mixed demand forecasting and collaborative optimization of steel enterprises based on production planning integration as described in claim 1, characterized in that, Purchased electricity load Calculated using the following mathematical model: in, Indicates the predetermined calculation period; This represents the sum of electricity consumption of all stable loads within the calculation period τ; This represents the sum of electricity consumption for all impact loads within the calculation period τ. This represents the sum of electricity consumption of all fluctuating loads within the calculation period τ; This represents the sum of the power generation of all generating loads within the calculation period τ.
9. The method for mixed demand forecasting and collaborative optimization of steel enterprises based on production planning integration as described in claim 8, characterized in that, Based on purchased electricity load The slip method is used to calculate the electricity demand, and the calculation method is as follows: in, , For the start and end times of the sliding window, This represents the demand value at the end of the sliding window operation. This refers to the length of the sliding window.
10. A mixed demand forecasting and collaborative optimization system for steel enterprises based on production planning integration, characterized in that, Used to implement the method according to any one of claims 1-9; The system includes a data acquisition module, a power load forecasting module, a maximum demand forecasting module, and an optimization and adjustment strategy module. The data acquisition module is used to collect electricity consumption data and production plan data at various metering points within the factory area in real time using 5G mobile network and wired network. The power load forecasting module is used to perform sub-forecasts of future power loads for stable loads, impact loads, and fluctuating loads, and to generate a forecast curve of the total purchased power load of the entire plant by accumulating the results of each sub-forecast. The maximum demand prediction module is used to predict the maximum demand value within a future target period on a rolling basis. The optimization and adjustment strategy module is used to generate scheduling instructions based on the controllability characteristics of various loads when the maximum demand forecast exceeds the preset demand threshold, so as to collaboratively suppress the final maximum demand.