Equipment energy efficiency improvement method and system based on multi-scale energy efficiency analysis mechanism and AI fusion optimization

Abstract

The invention discloses an equipment energy efficiency improvement method and system based on a multi-scale energy efficiency analysis mechanism and AI fusion optimization. The method comprises the steps of collecting design data, real-time operation data and related industry knowledge data of target equipment; preprocessing the real-time operation data, and performing diagnostic analysis and early warning on the operation state of the equipment in combination with a preset reasonable range of the operation data; executing short-time scale energy efficiency analysis on the preprocessed operation data based on an energy efficiency mechanism model of the equipment; energy efficiency analysis of medium-time and long-time scales is executed; and fusing the analysis results of the short time scale, the medium time scale and the long time scale, constructing an energy efficiency optimization model in which a mechanism and AI are fused, and based on the energy efficiency optimization model, outputting a regulation and control strategy to optimize the equipment energy efficiency. On the basis of fusion of an energy efficiency mechanism and AI, equipment energy saving can be reliably, efficiently and accurately realized from the global aspect.

Description

Technical Field

[0001] This invention relates to the field of energy-saving technology, and in particular to a method and system for improving equipment energy efficiency based on multi-scale energy efficiency analysis mechanism and AI fusion optimization. Background Technology

[0002] Industrial production consumes a large amount of energy, and energy-consuming equipment is ubiquitous in industrial enterprises. Energy consumption often accounts for more than 30% of production costs. Energy conservation and cost reduction have always been ongoing tasks for industrial enterprises, and improving the energy utilization efficiency (referred to as "energy efficiency") of energy-consuming equipment is the most fundamental direction for energy conservation and cost reduction. Numerous energy-saving technologies and measures have been developed to improve the energy efficiency of energy-consuming equipment, such as high-efficiency energy-consuming equipment technology, automatic control technology, and waste heat and energy recovery and utilization technology. The application of these technologies has greatly improved the energy efficiency level of energy-consuming equipment. For example, the comprehensive energy consumption per ton of steel in my country's steel industry has decreased from 770 kgce in 2003 to approximately 550 kgce in 2024. With the gradual improvement of energy efficiency, further tapping into energy conservation potential is becoming increasingly difficult, and pursuing ultimate energy efficiency is a new challenge posed by the times.

[0003] The development of next-generation computer technologies such as artificial intelligence has provided new technological directions for improving the energy efficiency of energy-consuming equipment. For example, the invention patent application CN202411827850.9, entitled "A Method for Improving the Energy Efficiency of Key Dynamic Equipment in an LNG Receiving Station," proposes to optimize the start-up and shutdown combinations of dynamic equipment and the logistics load allocation scheme by establishing a predictive model and an energy efficiency optimization model to predict the energy consumption of key dynamic equipment. The invention application CN202510396702.4 discloses "A Data-Driven Method and System for Optimizing the Energy Efficiency of Chemical Equipment," which addresses the technical problems of existing chemical equipment energy efficiency optimization being unable to dynamically adapt to complex scenarios, lacking global optimization capabilities, and ignoring the impact of scaling, thus limiting energy efficiency improvement. The invention application CN202410304976.1 discloses "A Method for Judging Equipment Energy Efficiency Balance and Improving Capacity," which compares calculated data with standard energy efficiency to judge energy efficiency, and simultaneously establishes an energy efficiency optimization model to propose energy efficiency improvement measures to improve operational energy efficiency. These technologies and methods are mostly based on mechanisms or data-driven approaches, without organically integrating the two for analysis. They fail to leverage their respective advantages and do not consider the efficient, accurate, and extreme improvement of energy efficiency through energy efficiency analysis at different time scales. Furthermore, they make it difficult to ensure the safe, stable, and reliable production of equipment when controlling multiple objects, thus reducing the applicability of energy efficiency improvement to large energy-consuming equipment such as coke ovens and blast furnaces. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for improving equipment energy efficiency based on multi-scale energy efficiency analysis mechanism and AI fusion optimization, thereby improving the energy efficiency of equipment.

[0005] This invention provides a method for improving equipment energy efficiency based on multi-scale energy efficiency analysis mechanism and AI fusion optimization, including: Collect design data, real-time operational data, and relevant industry knowledge data of the target equipment; The real-time operating data is preprocessed and combined with a preset reasonable range of operating data to perform diagnostic analysis and early warning of the equipment's operating status; Based on the energy efficiency mechanism model of the equipment, short-timescale energy efficiency analysis is performed on the preprocessed operating data to calculate and track the real-time energy efficiency status of the equipment in real time. Perform time-scale energy efficiency analysis and combine the results of short-time-scale energy efficiency tracking analysis to assess whether energy efficiency optimization operations should be triggered; Perform long-term energy efficiency analysis, and use statistical analysis to assess the reasonable range of operational data over a long period of time and analyze the factors affecting energy efficiency; By integrating the analysis results of the short, medium and long time scales, an energy efficiency optimization model that integrates mechanism and AI is constructed. Based on the energy efficiency optimization model, an energy efficiency optimization and control strategy for equipment is output.

[0006] Preferably, the design data includes the target equipment's model, design capacity, rated operating parameters, installed power, basic dimensions, and type parameters; the operating data includes the target equipment's operating environment temperature, ambient humidity, air velocity, actual energy supply parameters, raw material supply parameters, equipment operating parameters, product output and quality, and safety, quality, and environmental protection parameters; the industry knowledge data includes the target equipment's energy efficiency testing and calculation methods, the equipment industry's energy efficiency level, and common raw material physical property parameters.

[0007] Preferably, the energy efficiency mechanism model of the equipment includes one or more of the following: thermal efficiency model, energy utilization efficiency model, energy efficiency model, and unit product energy consumption model. The thermal efficiency model, energy utilization efficiency model, and energy efficiency model are expressed by the following formulas: , In the formula, This refers to thermal efficiency, energy utilization efficiency, or thermal efficiency. This indicates the amount of heat or energy effectively utilized. It represents the total input heat or energy or calorie; The unit product energy consumption model is expressed by the following formula: , In the formula, Q represents the energy consumption per unit of product, Qe represents the total input energy, and W represents the total product output; The formula for calculating the real-time energy efficiency status of coke oven equipment is as follows: , In the formula, η represents the thermal efficiency of the coke oven system. The heat carried out by the red coke in the coking product per ton of coal fed into the furnace, in kJ / t; The heat carried out by the tar in the coking product per ton of coal fed into the furnace, in kJ / t; The heat carried out by crude benzene in the coking product per ton of coal fed into the furnace, in kJ / t; The heat carried away by ammonia in the coking product per ton of coal fed into the furnace, expressed in kJ / t; The heat carried out by the net gas in the coking product per ton of coal fed into the furnace, in kJ / t; Q1 represents the heat carried out by moisture in the coking product per ton of coal fed into the furnace, in kJ / t; Q2 represents the sensible heat carried out by dry coal per ton of coal fed into the furnace, in kJ / t; Q3 represents the sensible heat carried out by moisture in per ton of coal fed into the furnace, in kJ / t. This represents the total heat from all sources, expressed in kJ / t.

[0008] Preferably, the operating data used in the short-timescale energy efficiency analysis, medium-timescale energy efficiency analysis, and long-timescale energy efficiency analysis are all sets of operating data collected within their respective timescales after the previous analysis; wherein, the division of short-time, medium-time, and long-timescales is related to the production characteristics of the equipment, so as to distinguish the impact of operation control, material feeding, and equipment status on energy efficiency.

[0009] Preferably, the energy efficiency analysis performed over a long time scale further includes: Acquire operational data on target preparation over a long timescale; The operational data collected based on the Laida criteria is filtered to remove abnormal data; Cluster analysis was performed on the filtered operational data to select the dataset representing normal production status; The average value of the dataset is calculated based on the dataset of the normal production state, and a preset threshold is extended from the average value in both positive and negative directions to determine the reasonable range of the data. The reasonable range of the data is determined by the following formula: , In the formula, x i These are the operating parameters of the target equipment; x is the average value of the operating parameters after clustering within a preset time period and filtering by the Layda criterion; imax x imin The maximum and minimum values ​​of the operating parameters after clustering within a preset time and filtering by the Laida criterion are α, where α is the expansion coefficient.

[0010] Preferably, the energy efficiency analysis performed over a long time scale further includes: Obtain the operational data and corresponding energy efficiency index values ​​used in the medium-time scale energy efficiency analysis; Linear or nonlinear regression analysis was performed on the operating data and energy efficiency index values ​​to establish a correlation model between energy efficiency impact parameters and energy efficiency indexes; The initial energy efficiency impact pattern generated based on historical data is used as the training basis for the correlation model, and the model parameters are iteratively optimized.

[0011] Preferably, the construction of the multi-scale fusion energy efficiency optimization model includes: We collect energy efficiency analysis data from various short timescales calculated within the medium timescale and construct a training dataset to optimize the dimensional parameters. The training dataset is used to train a model based on machine learning algorithms to generate an initial optimized model. Based on the reasonable range of data determined by long-term energy efficiency analysis, the initial optimization model is constrained and optimized to obtain the optimal values ​​of dimensional parameters, and optimization instructions are output. Substitute the optimal values ​​of the dimensional parameters into the energy efficiency impact law model to predict the adjusted equipment energy efficiency level. Based on actual operating data from medium-timescale energy efficiency analysis, the expected energy-saving effect is calculated and the accuracy of the prediction is verified. The energy efficiency impact law model is a correlation model between energy efficiency impact parameters and energy efficiency indicators established based on medium-time scale energy efficiency analysis.

[0012] This invention also provides an equipment energy efficiency improvement system based on multi-scale energy efficiency analysis mechanism and AI fusion optimization, comprising: The data acquisition module is used to collect design data, real-time operational data, and relevant industry knowledge data of the target equipment; The data preprocessing module is used to preprocess the real-time operating data and, in conjunction with a preset reasonable range of operating data, to perform diagnostic analysis and early warning of the equipment's operating status. The short-timescale energy efficiency analysis module is used to perform short-timescale energy efficiency analysis on preprocessed operating data based on the energy efficiency mechanism model of the equipment, so as to calculate and track the real-time energy efficiency status of the equipment in real time. The medium-timescale energy efficiency analysis module is used to perform medium-timescale energy efficiency analysis and, in conjunction with the results of short-timescale energy efficiency tracking analysis, assess whether to trigger energy efficiency optimization operations. The long-term energy efficiency analysis module is used to perform long-term energy efficiency analysis. It uses statistical analysis to assess the reasonable range of operating data over a long period of time and to analyze factors affecting energy efficiency. The mechanism and AI fusion optimization module is used to integrate the analysis results of the short, medium and long time scales to construct an energy efficiency optimization model that integrates mechanism and AI, and output equipment energy efficiency optimization and control strategies based on the energy efficiency optimization model.

[0013] The present invention also provides an electronic device, comprising: The memory is used to store the processing program; The processor, when executing the processing program, implements the equipment energy efficiency improvement method based on multi-scale energy efficiency analysis mechanism and AI fusion optimization as described in the embodiments of the present invention.

[0014] The present invention also provides a readable storage medium storing a processing program, which, when executed by a processor, implements the equipment energy efficiency improvement method based on multi-scale energy efficiency analysis mechanism and AI fusion optimization as described in the embodiments of the present invention.

[0015] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention achieves the organic integration of mechanism analysis and AI, which is based on mechanism analysis, thereby improving the efficiency and accuracy of energy efficiency optimization. Energy efficiency mechanism analysis is based on the fundamental physical laws such as strict mass conservation and energy balance. Mechanism analysis can accurately reflect the energy use process, making mechanism-based analysis highly interpretable. However, optimization methods based on traditional optimization theory are prone to getting trapped in local optima. AI technology is a technology for prediction and optimization based on machine learning models formed by big data analysis and training. It can reflect global and implicitly related characteristics, but its computational load is large and its interpretability is poor. This technical solution achieves the organic integration of the two by applying machine learning and other AI technologies on the basis of energy efficiency mechanism analysis at different scales. It leverages the advantages of each and reduces the impact of their respective disadvantages, thereby greatly improving the optimization efficiency and accuracy.

[0016] (2) This approach leverages the different roles of energy efficiency analysis at different time scales, improving the efficiency, accuracy, and effectiveness of energy efficiency optimization. This technical solution rigorously defines the energy efficiency mechanism model and proposes the content and different roles of energy efficiency analysis at multiple time scales. Compared to optimization schemes using single-scale energy efficiency analysis, multi-time-scale energy efficiency analysis can reflect the implicit impacts of operation control, material feeding, and equipment status on energy efficiency in different dimensions, thus providing accurate basic data for precise energy efficiency optimization. This is because the impact of operation control on energy efficiency is often reflected in a short time, while the impact of material feeding requires analysis of data from multiple different batches, which often takes a long time. Furthermore, the energy efficiency status of equipment typically changes very slowly without major structural modifications.

[0017] (3) It realizes the organic combination of energy efficiency analysis and energy efficiency optimization at different time scales, and improves the efficiency and accuracy of energy efficiency optimization. Operation control, feeding conditions and equipment status all play a role in the energy efficiency of equipment at different time scales. However, in general, the operation control at which the equipment has the best energy efficiency is different for different equipment statuses and feeding conditions. Therefore, it is difficult to accurately achieve the optimization effect by using energy efficiency analysis at a single time scale. This invention proposes to use energy efficiency analysis at multiple time scales and organically combine it with energy efficiency optimization. This can not only accurately grasp the impact of different equipment statuses, different feeding conditions and different operation controls on energy efficiency, but also accurately obtain the optimization dimension parameters of each through optimization based on energy efficiency analysis at different time scales, thereby improving the efficiency and accuracy of energy efficiency optimization.

[0018] (4) Improved safety and stability of regulation. Through energy efficiency analysis and optimization at different time scales, regulation can be carried out in stages, which can reduce mutual interference between the regulated objects and avoid production instability caused by simultaneous regulation of multiple objects, thus ensuring safe, reliable and stable production to the greatest extent. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the equipment energy efficiency improvement method based on multi-scale energy efficiency analysis mechanism and AI fusion optimization in an embodiment of the present invention. Figure 2 This is a schematic diagram of time scale division in an embodiment of the present invention; Figure 3 This is a flowchart of the energy efficiency optimization process in step S6 of an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] The term "comprising" and its variations as used herein are open-ended inclusion, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below.

[0022] Example 1 like Figure 1As shown, this invention provides a method for improving equipment energy efficiency based on multi-scale energy efficiency analysis mechanism and AI fusion optimization, including: S1: Collect design data, real-time operational data, and relevant industry knowledge data of the target equipment; S2: Preprocess the real-time operating data and, in conjunction with a preset reasonable range of operating data, perform diagnostic analysis and early warning of the equipment's operating status; S3: Based on the energy efficiency mechanism model of the equipment, perform short-timescale energy efficiency analysis on the preprocessed operating data to calculate and track the real-time energy efficiency status of the equipment in real time. S4: Perform energy efficiency analysis on a medium timescale and assess whether to trigger energy efficiency optimization operations by combining the results of energy efficiency tracking analysis on a short timescale. S5: Perform long-term energy efficiency analysis, and conduct statistical analysis to assess the reasonable range of operational data over a long period of time and analyze the factors affecting energy efficiency. S6: Integrate the analysis results of the short, medium and long time scales to construct a multi-scale fusion energy efficiency optimization model, and output the equipment energy efficiency optimization and control strategy based on the energy efficiency optimization model.

[0023] This embodiment, based on a mechanistic model, enables real-time energy efficiency calculation to quickly pinpoint efficiency bottlenecks in the current operating state (such as equipment overload or parameter deviation from optimal values), achieving instant adjustments at the second / minute level to prevent the spread of energy efficiency degradation. It proactively predicts the risk of energy efficiency decline (such as continuous multi-cycle efficiency reduction) through trend analysis, triggering optimization intervention in advance to prevent small fluctuations from evolving into systemic inefficiency. It reveals long-term energy efficiency evolution patterns, supporting strategic transformations (such as equipment selection and process reconfiguration), forming a closed loop of "short-term correction + long-term improvement." It breaks through the limitations of analysis at a single time scale, combining short-term real-time states, medium-term trend characteristics, and long-term statistical patterns to construct an energy efficiency optimization model that more closely reflects real-world operating conditions, avoiding global suboptimal solutions caused by local optimization. It constrains AI training boundaries with physical mechanism models (such as mass conservation and energy balance), solving the traditional AI black-box problem and improving model interpretability and cross-scenario transferability. By automatically filtering abnormal data and dynamically correcting the reasonable range of data based on historical clustering results, the false alarm rate is reduced. It can detect sudden anomalies (such as instantaneous power surges) on a short-term scale and identify gradual fault precursors (such as slow energy consumption increases due to bearing wear) on a medium-term scale, achieving a leap from "reactive maintenance" to "predictive maintenance." Through three core technological innovations—multi-scale spatiotemporal dimension coverage, mechanism-AI dual-drive, and dynamic data governance—it has achieved a leap in equipment energy efficiency from "passive monitoring" to "proactive optimization," and from "experience-driven" to "data intelligence." This significantly reduces enterprise energy costs, improves equipment utilization, and provides a feasible technical path for achieving carbon neutrality goals in the industrial sector.

[0024] To improve equipment energy efficiency in this embodiment, it is essential to first understand the current operating status of the equipment. Therefore, step S1 requires collecting the equipment's design and operational data. Furthermore, the requirements for equipment differ across industries; for example, even with the same heating furnace, the heating process differs between ferrous and non-ferrous metallurgy. Therefore, it is also necessary to collect relevant industry knowledge.

[0025] Preferably, the design data of the equipment in step S1 includes: model, design capacity, rated operating parameters, installed power, basic dimensions, type parameters, etc. These data determine the most basic performance of the equipment. Furthermore, the rated operating parameters include: the environmental conditions of the equipment, energy supply, working system, raw material and fuel conditions, product specifications, and safety, quality, and environmental protection requirements, etc.; the basic dimensions include: the length, width, and height of the equipment, the dimensions of the raw material and fuel inlet / outlet interfaces, and the opening dimensions, etc.; the type parameters include: the principle type, structural type, material type, and logistics type of the equipment, etc.

[0026] Preferably, the operating data of the equipment in step S1 includes: ambient temperature, ambient humidity, and air velocity; actual energy supply parameters; raw material and fuel supply parameters; equipment operating parameters; product output and quality; and safety, quality, and environmental protection parameters. These parameters determine the actual energy utilization efficiency level of the equipment. Energy efficiency improvement and optimization require adjustment of the adjustable parameters, i.e., the parameters of the dimensions to be optimized, to achieve the best energy efficiency level. Furthermore, the actual energy supply parameters include: the flow rate, pressure, and temperature of water, wind, compressed air, and steam supplied to the equipment; the electrical power, frequency, and voltage supplied to the equipment; and the opening degree of the energy control valve. The raw material and fuel supply parameters include: the composition, temperature, and supply quantity of raw materials; and the composition, temperature, flow rate, and pressure of fuel. The safety, quality, and environmental protection parameters include: the temperature and vibration parameters of rotating parts; the pressure, stress, and level parameters of stationary parts; the composition, temperature, pressure, and flow rate parameters of discharged fluids; and the composition, temperature, and quantity parameters of discharged solids.

[0027] Preferably, the industry knowledge mentioned in step S1 includes: equipment energy efficiency testing and calculation methods, equipment industry energy efficiency levels, and common raw material physical property parameters. Equipment energy efficiency testing and calculation methods reflect the computational principles of energy efficiency mechanisms, while equipment industry energy efficiency levels reflect the benchmark and advanced levels achievable by the equipment, providing support for reliable energy conservation.

[0028] Production status early warning can reduce production accidents and energy losses caused by equipment failures. Therefore, in a preferred embodiment, a diagnostic analysis and early warning method is proposed in step S2, including the following steps: judging whether the current operating parameters are within a reasonable range based on the determined reasonable range of data; issuing an early warning signal when the data exceeds the reasonable range; determining abnormal production when key parameters such as output, raw material and fuel consumption, and characteristic operating parameters are abnormal; and analyzing the impact of the operating parameter on energy efficiency based on the influence law obtained from the energy efficiency influencing factor analysis, and giving a diagnostic conclusion. First, judging whether the operating parameters are within a reasonable range is a relatively easy and efficient judgment. Furthermore, early warnings can be given based on the analysis judgment, and the impact of the operating parameter on energy efficiency can be obtained from the energy efficiency influencing factor analysis, giving a diagnostic conclusion. These functions provide a basic guarantee for the safe, reliable, stable, and smooth operation of equipment. In order to determine the reasonable range of data, the state of abnormal production needs to be marked. Here, key parameters are used for judgment, which simplifies the judgment process and maintains consistency with the judgment of the actual production situation of the equipment.

[0029] Energy efficiency mechanism models are the foundation for calculating energy utilization efficiency. The energy efficiency mechanism models of the equipment include one or more of the following: thermal efficiency model, energy utilization efficiency model, thermal efficiency model, and unit product energy consumption model. This embodiment provides their calculation formulas. These energy efficiency mechanism models are strictly based on fundamental physical laws such as mass conservation and energy balance, and can accurately reflect the energy usage process. This makes mechanism-based optimization highly interpretable. At the same time, energy efficiency mechanism analysis does not require complex iterative calculation processes, which can improve the efficiency of computational analysis.

[0030] The thermal efficiency model, energy utilization efficiency model, and thermal efficiency model are expressed by the following formulas: , In the formula, Thermal efficiency (energy utilization efficiency, thermal efficiency) is expressed as a percentage (%). It represents the amount of heat (energy, kJ) that is effectively utilized, expressed in kJ (kgce, kJ). Represents the total input heat (energy, 㶲), kJ (kgce, kJ); The unit product energy consumption model is expressed by the following formula: , In the formula, Q represents the energy consumption per unit product, expressed in kgce / t (or kgce / Nm3, etc.); Qe represents the total input energy, kgce; and W represents the total product output, t (or Nm3, etc.). The specific forms of each item in the energy efficiency mechanism model are closely related to the equipment described. For example, the thermal efficiency of a coke oven can be expressed by the following formula: , In the formula, η represents the thermal efficiency of the coke oven system, expressed as a percentage (%); The heat carried out by the red coke in the coking product per ton of coal fed into the furnace, in kJ / t; The heat carried out by the tar in the coking product per ton of coal fed into the furnace, in kJ / t; The heat carried out by crude benzene in the coking product per ton of coal fed into the furnace, in kJ / t; The heat carried away by ammonia in the coking product per ton of coal fed into the furnace, expressed in kJ / t; The heat carried out by the net gas in the coking product per ton of coal fed into the furnace, in kJ / t; Q1 represents the heat carried out by moisture in the coking product per ton of coal fed into the furnace, in kJ / t; Q2 represents the sensible heat carried out by dry coal per ton of coal fed into the furnace, in kJ / t; Q3 represents the sensible heat carried out by moisture in per ton of coal fed into the furnace, in kJ / t. This represents the total heat from all sources, expressed in kJ / t.

[0031] In a preferred embodiment, the frequency of operational data acquisition can be on the order of seconds or minutes, depending on the specific equipment and the availability of operational data. Figure 2 In a preferred embodiment, this invention provides short-time, medium-time, and long-time time scales, suggesting that these scales are related to the production characteristics of the equipment. This invention also provides suggested time scales for both continuous and non-continuous equipment. For continuous production equipment, such as coke ovens, short-time can be 15 minutes, medium-time can be based on the typical coking cycle of a coke oven, such as 19 hours, and long-time can be one month. Multiple medium-time scales can be set to conduct energy efficiency optimization in different dimensions, such as operation control optimization and feeding optimization. For non-continuous production equipment, such as converters, short-time can be based on the smelting time of a single furnace, such as 35 minutes, medium-time can be based on the average production time of 60 furnaces, and long-time can be one month. The reason for using different short-time, medium-time, and long-time time scales for different equipment is mainly to consider the actual production characteristics of different equipment in different industries. Only in this way can the impact of operation control, feeding conditions, and equipment status on energy efficiency be separated for targeted regulation. For a specific coke oven, its energy efficiency baseline will not change over a relatively long period of time if its equipment status is not significantly modified. Once the specific coal composition and charging frequency are determined, the feeding situation is determined, which basically determines its impact on energy efficiency. At this point, the only things that can be operated and controlled are short-term controllable objects such as combustion load.

[0032] In a preferred embodiment, the operating data used in the short-timescale energy efficiency analysis, medium-timescale energy efficiency analysis, and long-timescale energy efficiency analysis are all sets of operating data collected within their respective timescales after the previous analysis; wherein, the division of short-time, medium-time, and long-timescales is related to the production characteristics of the equipment, so as to distinguish the impact of operation control, material feeding, and equipment status on energy efficiency.

[0033] In a preferred embodiment, step S4, which assesses whether to implement energy efficiency optimization analysis, includes the following steps: determining the energy efficiency development direction based on the energy efficiency change trends at various short time scales within a medium time scale; triggering the energy efficiency optimization analysis function when it is determined that energy efficiency is trending towards deterioration, or when the energy efficiency at a medium time scale is lower than the industry benchmark level. This fully combines medium- and short-time scale energy efficiency analysis, both identifying energy efficiency trends to prevent deterioration and integrating with industry benchmark levels. This ensures that the equipment's energy efficiency level is not lower than the industry benchmark level while reducing the frequency of optimization. Step S4, which combines short-timescale energy efficiency tracking analysis results to assess whether to trigger energy efficiency optimization, includes: Step S4a: Obtaining energy efficiency analysis results for each short-timescale within the medium-timescale, and constructing an energy efficiency change trend sequence; Step S4b: Determining the energy efficiency development direction based on the trend sequence. If a downward trend in energy efficiency is detected for N consecutive short-timescales, it is determined as energy efficiency degradation; Step S4c: When the energy efficiency value at the medium-timescale is lower than a preset industry benchmark level, the energy efficiency optimization analysis function is triggered; Step S4d: When either step S4b or step S4c is met, the energy efficiency optimization process in step S6 is initiated. The trend sequence in step S4a is constructed using a sliding window mechanism, and the duration of the sliding window is consistent with the analysis cycle at the medium-timescale. The threshold for determining energy efficiency degradation in step S4b includes: the energy efficiency decrease exceeding a preset percentage; the number of consecutively decreasing short-timescales reaching a preset number N; and the threshold can be dynamically adjusted according to equipment type and industry characteristics. The industry benchmark level is determined through historical data from long-timescale energy efficiency analysis and is dynamically adjusted as industry energy efficiency standards are updated.

[0034] In a preferred embodiment, performing a long-term energy efficiency analysis in step S5 further includes: Acquire operational data on target preparation over a long timescale; The operational data collected based on the Laida criterion is filtered to remove abnormal data, thus avoiding deviations from the reasonable range of data caused by obvious abnormal data. Cluster analysis is performed on the filtered operational data to select the dataset representing the normal production status. Selecting the dataset representing the normal production status determines the reasonable range of data and avoids the entry of unreasonable data during abnormal production. This can also be combined with the aforementioned judgment on the marking of abnormal production. The average value of the dataset is calculated based on the dataset under normal production conditions. The reasonable range of the data is determined by expanding the dataset outward from the average value in both positive and negative directions using the average value as the center. By expanding the data slightly outward, the optimal energy efficiency level that can be achieved after optimization can be further improved. Since the reasonable range of data plays an important role in diagnostic analysis and early warning, it is often necessary to obtain the initial reasonable range of data through offline analysis based on historical data.

[0035] The reasonable range of the data is determined by the following formula: , In the formula, x i These are the operating parameters of the target equipment; x is the average value of the operating parameters after clustering within a preset time period and filtering by the Layda criterion; imax x imin These are the maximum and minimum values ​​of the operating parameters after clustering within a preset time and filtering by the Laida criterion. α is the expansion coefficient, which defaults to 10% and can be adjusted according to the site conditions.

[0036] In a preferred embodiment, performing a long-term energy efficiency analysis in step S5 further includes: Obtain the operational data and corresponding energy efficiency index values ​​used in the medium-time scale energy efficiency analysis; Linear or nonlinear regression analysis is performed on the operational data and energy efficiency index values ​​to establish a correlation model between energy efficiency influencing parameters and energy efficiency indexes, thereby obtaining the influence law of each energy efficiency influencing parameter on energy efficiency; here, long-term energy efficiency analysis and medium-term energy efficiency analysis are combined, and the influence law of each energy efficiency influencing parameter on energy efficiency is obtained through regression analysis.

[0037] The initial energy efficiency impact patterns generated based on historical data are used as the training basis for the correlation model, and the model parameters are iteratively optimized. Similarly, since impact patterns play an important role in diagnostic analysis and early warning, it is often necessary to obtain them based on historical data analysis in order to obtain the initial impact patterns.

[0038] In a preferred embodiment, step S6, constructing a multi-scale fusion energy efficiency optimization model, includes: We collect energy efficiency analysis data from various short timescales calculated within the medium timescale and construct a training dataset to optimize the dimensional parameters. The training dataset is used to train a model based on machine learning algorithms to generate an initial optimized model. Based on the reasonable range of data determined by long-term energy efficiency analysis, the initial optimization model is constrained and optimized to obtain the optimal values ​​of dimensional parameters, and optimization instructions are output. Substitute the optimal values ​​of the dimensional parameters into the energy efficiency impact law model to predict the adjusted equipment energy efficiency level. Based on actual operating data from medium-timescale energy efficiency analysis, the expected energy-saving effect is calculated and the accuracy of the prediction is verified. The energy efficiency impact model is a correlation model between energy efficiency impact parameters and energy efficiency indicators established based on medium-time-scale energy efficiency analysis. This can be understood as combining... Figure 3 The energy efficiency optimization steps in step S6 are as follows: (1) Collect energy efficiency analysis data of various short time scales calculated within the medium time scale, and train the machine learning model for the optimization dimension parameters; (2) Combine the reasonable range of data of various operating parameters obtained from the long time scale energy efficiency analysis, carry out optimization based on machine learning algorithm, obtain the optimal value of dimension parameters, and output optimization instructions; (3) Predict the energy efficiency level that can be achieved after adjustment based on the energy efficiency impact law obtained from the long time scale energy efficiency analysis, and calculate the expected energy saving effect based on the results of the medium time scale energy efficiency analysis. It can be seen from this that the energy efficiency optimization method further organically combines medium and long time scale energy efficiency analysis, and organically combines energy efficiency mechanism analysis with AI algorithm, utilizing the respective advantages of mechanism analysis and AI technology, which can improve optimization efficiency and accuracy. At the same time, by setting multiple medium time scales, adopting multi-time scale energy efficiency analysis and organically combining it with machine learning-based energy efficiency optimization, it is possible to accurately grasp the impact of different equipment states, different feeding conditions, and different operation controls on energy efficiency, and accurately obtain their respective optimization dimension parameters through optimization based on energy efficiency analysis of different time scales, thereby improving the efficiency and accuracy of energy efficiency optimization. By analyzing and optimizing energy efficiency at different time scales, regulation can be carried out in a hierarchical manner, which can reduce mutual interference between regulated objects and avoid production instability caused by simultaneous regulation of multiple objects, thus ensuring safe, reliable, stable and smooth production to the greatest extent.

[0039] Preferably, when energy efficiency optimization involves multiple dimensions, the parameters of the previous dimension involved in the energy efficiency optimization of the next scale can be optimized by substituting their corresponding optimal values. This design can reduce the optimization difficulty, improve the optimization efficiency, and achieve an organic combination of energy efficiency analysis at different time scales.

[0040] Preferably, the optimization dimension parameters are a subset of the energy efficiency impact parameters in the energy efficiency impact analysis. This achieves an organic combination of long-term energy efficiency analysis and medium-term energy efficiency optimization, ensuring the coordination between the global energy efficiency impact parameters and the optimization dimension parameters.

[0041] This invention proposes a mechanism-AI fusion optimization method for equipment energy efficiency improvement based on multi-scale energy efficiency analysis. By applying AI technology under energy efficiency mechanism analysis at different scales, it achieves an organic integration of mechanism analysis (primarily based on mechanism analysis) and AI, leveraging the advantages of each while mitigating their respective disadvantages, thus improving the efficiency and accuracy of energy efficiency optimization. Compared to optimization schemes using single-scale energy efficiency analysis, multi-timescale energy efficiency analysis can reflect the implicit impacts of operation control, material feeding, and equipment status on energy efficiency, thereby providing accurate basic data for precise energy efficiency optimization. The organic combination of multi-timescale energy efficiency analysis and energy efficiency optimization can accurately grasp the impact of different equipment statuses, material feeding conditions, and operation controls on energy efficiency, and accurately obtain the parameters of each optimization target through optimization based on energy efficiency analysis at different time scales, thereby improving the efficiency and accuracy of energy efficiency optimization. This invention, through energy efficiency analysis and optimization at different time scales, enables hierarchical control, reducing mutual interference between control targets and avoiding production instability caused by simultaneous control of multiple targets, thus maximizing the safety, reliability, and stable operation of production.

[0042] Example 2 Based on the same concept, this invention provides an equipment energy efficiency improvement system based on multi-scale energy efficiency analysis mechanism and AI fusion optimization, comprising: The data acquisition module is used to collect design data, real-time operational data, and relevant industry knowledge data of the target equipment; The data preprocessing module is used to preprocess the real-time operating data and, in conjunction with a preset reasonable range of operating data, to perform diagnostic analysis and early warning of the equipment's operating status. The short-timescale energy efficiency analysis module is used to perform short-timescale energy efficiency analysis on preprocessed operating data based on the energy efficiency mechanism model of the equipment, so as to calculate and track the real-time energy efficiency status of the equipment in real time. The medium-timescale energy efficiency analysis module is used to perform medium-timescale energy efficiency analysis and, in conjunction with the results of short-timescale energy efficiency tracking analysis, assess whether to trigger energy efficiency optimization operations. The long-term energy efficiency analysis module is used to perform long-term energy efficiency analysis. It uses statistical analysis to assess the reasonable range of operating data over a long period of time and to analyze factors affecting energy efficiency. The mechanism and AI fusion optimization module is used to integrate the analysis results of the short, medium and long time scales to construct an energy efficiency optimization model that integrates mechanism and AI. Based on the energy efficiency optimization model, it outputs equipment energy efficiency optimization and control strategies. The implementation principle of the above module has been described in the previous embodiments, so it will not be repeated here.

[0043] Example 3 Based on the same concept, an electronic device is also provided in some embodiments of this application. This electronic device includes a memory and a processor, wherein the memory stores a processing program, and the processor executes the processing program according to instructions. When the processor executes the processing program, the equipment energy efficiency improvement method based on multi-scale energy efficiency analysis mechanism and AI fusion optimization described in the foregoing embodiments is realized.

[0044] In some embodiments of this application, a readable storage medium is also provided. This readable storage medium can be a non-volatile readable storage medium or a volatile readable storage medium. The readable storage medium stores instructions that, when executed on a computer, cause an electronic device containing this readable storage medium to perform the aforementioned equipment energy efficiency improvement method based on multi-scale energy efficiency analysis mechanisms and AI fusion optimization.

[0045] It is understood that the aforementioned equipment energy efficiency improvement methods based on multi-scale energy efficiency analysis mechanisms and AI fusion optimization, if implemented as software functional modules and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0046] Computer-readable storage media may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable storage medium may also be any readable medium other than a readable storage medium that can transmit, propagate, or transfer a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.

[0047] The program code for executing the technical solutions disclosed in this application can be written in any combination of one or more programming languages. These programming languages ​​include object-oriented programming languages—such as Java and C++—and conventional procedural programming languages—such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0048] 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 them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for improving equipment energy efficiency based on multi-scale energy efficiency analysis mechanism and AI fusion optimization, characterized in that, include: Collect design data, real-time operational data, and relevant industry knowledge data of the target equipment; The real-time operating data is preprocessed and combined with a preset reasonable range of operating data to perform diagnostic analysis and early warning of the equipment's operating status; Based on the energy efficiency mechanism model of the equipment, short-timescale energy efficiency analysis is performed on the preprocessed operating data to calculate and track the real-time energy efficiency status of the equipment in real time. Perform time-scale energy efficiency analysis and combine the results of short-time-scale energy efficiency tracking analysis to assess whether energy efficiency optimization operations should be triggered; Perform long-term energy efficiency analysis, and use statistical analysis to assess the reasonable range of operational data over a long period of time and analyze the factors affecting energy efficiency; By integrating the analysis results of the short, medium and long time scales, an energy efficiency optimization model that integrates mechanism and AI is constructed. Based on the energy efficiency optimization model, an energy efficiency optimization and control strategy for equipment is output.

2. The equipment energy efficiency improvement method based on multi-scale energy efficiency analysis mechanism and AI fusion optimization as described in claim 1, characterized in that, The design data includes the target equipment's model, design capacity, rated operating parameters, installed power, basic dimensions, and type parameters; the operational data includes the target equipment's operating environment temperature, ambient humidity, air velocity, actual energy supply parameters, raw material supply parameters, equipment operation parameters, product output and quality, and safety, quality, and environmental protection parameters; the industry knowledge data includes the target equipment's energy efficiency testing and calculation methods, the equipment industry's energy efficiency level, and common raw material properties.

3. The equipment energy efficiency improvement method based on multi-scale energy efficiency analysis mechanism and AI fusion optimization as described in claim 1, characterized in that, The energy efficiency mechanism model of the equipment includes one or more of the following: thermal efficiency model, energy utilization efficiency model, energy efficiency model, and unit product energy consumption model. The thermal efficiency model, energy utilization efficiency model, and energy efficiency model are expressed by the following formulas: ， In the formula, This refers to thermal efficiency, energy utilization efficiency, or thermal efficiency. This indicates the amount of heat or energy effectively utilized. It represents the total input heat or energy or calorie; The unit product energy consumption model is expressed by the following formula: , In the formula, Q represents the energy consumption per unit of product, Qe represents the total input energy, and W represents the total product output; The formula for calculating the real-time energy efficiency status of coke oven equipment is as follows: ， In the formula, η represents the thermal efficiency of the coke oven system. The heat carried out by the red coke in the coking product per ton of coal fed into the furnace, in kJ / t; The heat carried out by the tar in the coking product per ton of coal fed into the furnace, in kJ / t; The heat carried out by crude benzene in the coking product per ton of coal fed into the furnace, in kJ / t; The heat carried away by ammonia in the coking product per ton of coal fed into the furnace, expressed in kJ / t; The heat carried out by the net gas in the coking product per ton of coal fed into the furnace, in kJ / t; Q1 represents the heat carried out by moisture in the coking product per ton of coal fed into the furnace, in kJ / t; Q2 represents the sensible heat carried out by dry coal per ton of coal fed into the furnace, in kJ / t; Q3 represents the sensible heat carried out by moisture in per ton of coal fed into the furnace, in kJ / t. This represents the total heat from all sources, expressed in kJ / t.

4. The equipment energy efficiency improvement method based on multi-scale energy efficiency analysis mechanism and AI fusion optimization as described in claim 1, characterized in that, The operational data used in the short-timescale energy efficiency analysis, medium-timescale energy efficiency analysis, and long-timescale energy efficiency analysis are all sets of operational data collected within their respective timescales after the previous analysis. The division of short-time, medium-time, and long-timescales is related to the production characteristics of the equipment, in order to distinguish the impact of operation control, material feeding, and equipment status on energy efficiency.

5. The equipment energy efficiency improvement method based on multi-scale energy efficiency analysis mechanism and AI fusion optimization according to claim 1, characterized in that, The energy efficiency analysis performed over a long time scale further includes: Acquire operational data on target preparation over a long timescale; The operational data collected based on the Laida criteria is filtered to remove abnormal data; Cluster analysis was performed on the filtered operational data to select the dataset representing normal production status; The average value of the dataset is calculated based on the dataset of the normal production state, and a preset threshold is extended from the average value in both positive and negative directions to determine the reasonable range of the data. The reasonable range of the data is determined by the following formula: ， In the formula, x i These are the operating parameters of the target equipment; x is the average value of the operating parameters after clustering within a preset time period and filtering by the Layda criterion; imax x imin The maximum and minimum values ​​of the operating parameters after clustering within a preset time and filtering by the Layda criterion are α, where α is the expansion coefficient.

6. The equipment energy efficiency improvement method based on multi-scale energy efficiency analysis mechanism and AI fusion optimization according to claim 1, characterized in that, The energy efficiency analysis performed over a long time scale further includes: Obtain the operational data and corresponding energy efficiency index values ​​used in the medium-time scale energy efficiency analysis; Linear or nonlinear regression analysis was performed on the operating data and energy efficiency index values ​​to establish a correlation model between energy efficiency impact parameters and energy efficiency indexes; The initial energy efficiency impact pattern generated based on historical data is used as the training basis for the correlation model, and the model parameters are iteratively optimized.

7. The equipment energy efficiency improvement method based on multi-scale energy efficiency analysis mechanism and AI fusion optimization according to claim 1, characterized in that, The energy efficiency optimization model that integrates the construction mechanism and AI includes: We collect energy efficiency analysis data from various short timescales calculated within the medium timescale and construct a training dataset to optimize the dimensional parameters. The training dataset is used to train a model based on machine learning algorithms to generate an initial optimized model. Based on the reasonable range of data determined by long-term energy efficiency analysis, the initial optimization model is constrained and optimized to obtain the optimal values ​​of dimensional parameters, and optimization instructions are output. Substitute the optimal values ​​of the dimensional parameters into the energy efficiency impact law model to predict the adjusted equipment energy efficiency level. Based on actual operating data from medium-timescale energy efficiency analysis, the expected energy-saving effect is calculated and the accuracy of the prediction is verified. The energy efficiency impact law model is a correlation model between energy efficiency impact parameters and energy efficiency indicators established based on medium-time scale energy efficiency analysis.

8. An equipment energy efficiency improvement system based on multi-scale energy efficiency analysis mechanism and AI fusion optimization, characterized in that, include: The data acquisition module is used to collect design data, real-time operational data, and relevant industry knowledge data of the target equipment; The data preprocessing module is used to preprocess the real-time operating data and, in conjunction with a preset reasonable range of operating data, to perform diagnostic analysis and early warning of the equipment's operating status. The short-timescale energy efficiency analysis module is used to perform short-timescale energy efficiency analysis on preprocessed operating data based on the energy efficiency mechanism model of the equipment, so as to calculate and track the real-time energy efficiency status of the equipment in real time. The medium-timescale energy efficiency analysis module is used to perform medium-timescale energy efficiency analysis and, in conjunction with the results of short-timescale energy efficiency tracking analysis, assess whether to trigger energy efficiency optimization operations. The long-term energy efficiency analysis module is used to perform long-term energy efficiency analysis. It uses statistical analysis to assess the reasonable range of operating data over a long period of time and to analyze factors affecting energy efficiency. The mechanism and AI fusion optimization module is used to integrate the analysis results of the short, medium and long time scales to construct an energy efficiency optimization model that integrates mechanism and AI, and output equipment energy efficiency optimization and control strategies based on the energy efficiency optimization model.

9. An electronic device, characterized in that, include: The memory is used to store the processing program; The processor, when executing the processing program, implements the equipment energy efficiency improvement method based on multi-scale energy efficiency analysis mechanism and AI fusion optimization as described in any one of claims 1 to 8.

10. A readable storage medium, characterized in that, The readable storage medium stores a processing program, which, when executed by a processor, implements the equipment energy efficiency improvement method based on multi-scale energy efficiency analysis mechanism and AI fusion optimization as described in any one of claims 1 to 8.

Images