Method and system for intelligent management and control of energy in the production process of high-transparency ultra-thin photovoltaic glass

Abstract

The application relates to the field of photovoltaic glass production, and discloses an energy intelligent management and control method for a high-transmittance ultra-thin photovoltaic glass generation process, which comprises the following steps: collecting melting parameters and light transmittance in a photovoltaic glass melting process in real time; calculating melting heat demand of the photovoltaic glass according to a preset melting heat calculation model and the melting parameters; inputting the melting parameters and the melting heat demand into a preset neural network, wherein a hidden layer of the neural network is used to calculate a nonlinear residual term; correcting the melting heat demand based on the nonlinear residual term; detecting controllable current process parameters in the photovoltaic glass melting process; calculating minimum melting heat demand and corresponding target process parameters when the photovoltaic glass meets a target light transmittance requirement; and executing the photovoltaic glass melting process. The application constructs an energy intelligent management and control system with precise sensing, intelligent diagnosis and collaborative optimization, and effectively solves the problems of low energy efficiency and difficult quality control in the production of high-transmittance ultra-thin photovoltaic glass.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic glass production, and more specifically, to a method and system for intelligent energy management and control in the production process of high-transmittance ultrathin photovoltaic glass. Background Technology

[0002] With the accelerated pace of global energy structure transformation, solar energy, as a clean and renewable energy source, has received widespread attention for its development and utilization. Photovoltaic glass, as the core encapsulation material for solar cell modules, not only provides mechanical protection and environmental isolation, but its optical performance directly affects the photoelectric conversion efficiency of photovoltaic modules. In recent years, the market demand for high-transmittance, ultra-thin photovoltaic glass has been increasing, placing higher demands on the manufacturing process of photovoltaic glass.

[0003] In the production of photovoltaic glass, the melting process is a crucial step that determines its final performance. This process involves melting, clarifying, and homogenizing the batch materials at high temperatures to form a molten glass that meets specific optical and physical performance requirements. The melting process typically takes place in a high-temperature furnace, consuming a significant amount of energy, accounting for over 70% of the total energy consumption in photovoltaic glass production. Therefore, achieving energy-efficient utilization in the melting process while ensuring high light transmittance and ultra-thin product quality has become a pressing technical challenge in the photovoltaic glass manufacturing industry.

[0004] Currently, energy management in photovoltaic glass melting processes largely relies on experience or simple heat calculations based on physical models. These methods typically estimate the required melting heat using heat balance equations based on fundamental parameters such as raw material ratios, output rates, and target temperatures. However, the actual melting process is a highly nonlinear, strongly coupled, and complex system, influenced by various dynamic factors such as furnace thermal efficiency, combustion conditions, atmosphere control, and the flow state of the molten glass. Traditional heat calculation models struggle to accurately capture these complex nonlinear relationships, leading to significant deviations in heat demand estimations. This often necessitates substantial adjustments by operators based on experience, resulting in energy waste or impacting product quality.

[0005] Therefore, there is an urgent need for an intelligent energy management method and system for the production process of high-transmittance ultrathin photovoltaic glass that can accurately calculate the heat demand for photovoltaic glass melting and dynamically control the energy consumption of photovoltaic glass production. Summary of the Invention

[0006] In order to solve the above-mentioned technical problems, this application is made to provide a method and system for intelligent energy management and control in the production process of high-transmittance ultrathin photovoltaic glass, which can accurately calculate the heat demand for photovoltaic glass melting and dynamically manage the energy consumption of photovoltaic glass production.

[0007] In a first aspect, the present invention provides an intelligent energy management method for the production process of high-transmittance ultrathin photovoltaic glass, comprising: real-time acquisition of melting parameters and transmittance during the photovoltaic glass melting process; calculation of the melting heat requirement of the photovoltaic glass based on a preset melting heat calculation model and the melting parameters; inputting the melting parameters and the melting heat requirement into a preset neural network, wherein the neural network is a dual-input gated recurrent unit structure, wherein the main path is used to input the melting parameters, the auxiliary path is used to input the melting heat requirement, and the hidden layer is used to perform feature fusion on the melting parameters and the melting heat requirement, and calculate the melting heat requirement based on the fused features. The following steps are taken: Calculate a nonlinear residual term, which is used to compensate for defects in the calculation results of the melting heat calculation model; correct the melting heat requirement based on the nonlinear residual term; detect the controllable current process parameters in the photovoltaic glass melting process; calculate the minimum melting heat requirement and corresponding target process parameters for the photovoltaic glass to meet the target transmittance requirement based on the transmittance and the preset target transmittance, the current process parameters and corresponding critical values, and the melting heat requirement; control the energy supply of the photovoltaic glass melting process according to the target process parameters to execute the photovoltaic glass melting process.

[0008] Optionally, the aforementioned intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process includes real-time acquisition of melting parameters and transmittance during the photovoltaic glass melting process, including real-time acquisition of the temperature of the photovoltaic glass. Liquid temperature change rate Raw material mass flow rate Specific heat capacity and the electro-fusion power using electrodes ; Calculate the melting heat requirement of the photovoltaic glass according to the preset melting heat calculation model and the melting parameters, including: according to the formula Calculate the required melting heat ,in, The preset baseline energy conversion efficiency coefficient, This indicates the production year of the electrode. This indicates that the electrode is subject to the preset production years. The attenuation coefficient of the effect This indicates the service life of the electrode. This indicates that the electrode is subject to the temperature of the photovoltaic glass as preset. The attenuation coefficient of the effect The preset reference temperature, It represents the multiplication operation.

[0009] Optionally, the aforementioned intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process further includes, in addition to, calculating the melting heat requirement of the photovoltaic glass based on a preset melting heat calculation model and the melting parameters: detecting the external temperature of the furnace where the photovoltaic glass is located; calculating the nonlinear heat loss of the photovoltaic glass melting process based on the external temperature of the furnace, as well as the material and ventilation of the furnace; and correcting the melting heat requirement of the photovoltaic glass based on the nonlinear heat loss of the photovoltaic glass melting process. Specifically, before inputting the melting parameters and the melting heat requirement into a preset neural network, the method further includes: training the neural network and calculating the value of a preset loss function during the training process; adjusting the parameters of the main path and the parameters of the auxiliary path based on the gradient change of the loss function value; and adjusting the reference power conversion efficiency coefficient based on the gradient change of the loss function value. The electrode is subject to the number of years of production. The attenuation coefficient of the influence The electrode is affected by the temperature of the photovoltaic glass. The attenuation coefficient of the influence or the reference temperature Adjustments will be made.

[0010] Optionally, the aforementioned intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process includes feature fusion of the melting parameters and the melting heat requirement, comprising: according to a preset time window. The first feature vector is obtained by independently linearly projecting the melting parameters of the main path. According to the time window The second feature vector is obtained by independently linearly projecting the melting heat requirement of the auxiliary path. According to the time window The corresponding first feature vector Second eigenvector Calculate the time window Corresponding dynamic fusion intensity ,in For the sigmoid function, In the time window Previous time window The first feature vector generated, Used for concatenating vectors , These are preset parameters; based on the time window. Corresponding dynamic fusion intensity For the time window The corresponding first feature vector Second eigenvector Perform feature fusion.

[0011] Optionally, the aforementioned intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process further includes, before controlling the energy supply of the photovoltaic glass melting process according to the target process parameters to execute the photovoltaic glass melting process, the method further includes: calculating the difference between the target process parameters and the current process parameters; generating a transition process parameter sequence within a preset transition time using S-curve interpolation based on the difference between the target process parameters and the current process parameters; executing the photovoltaic glass melting process according to the transition process parameter sequence during the transition time; and after the transition time, proceeding to the step of controlling the energy supply of the photovoltaic glass melting process according to the target process parameters to execute the photovoltaic glass melting process.

[0012] Optionally, the aforementioned intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process further includes, before generating the transition process parameter sequence within a preset transition time, the following steps: detecting the thickness and conveying speed of the photovoltaic glass; calculating the speed-to-thickness ratio of the photovoltaic glass based on the thickness and conveying speed, which represents the tensile strain rate borne per unit thickness of the photovoltaic glass; and calculating the temperature of the photovoltaic glass... The temperature change rate of the photovoltaic glass is calculated; based on the speed-to-thickness ratio and temperature change rate of the photovoltaic glass, the maximum stress change rate of the photovoltaic glass is calculated to represent the stress growth rate of the photovoltaic glass per unit time; and the transition time is set based on the maximum stress change rate of the photovoltaic glass.

[0013] Optionally, the aforementioned intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process further includes: detecting the vibration frequency of the furnace using vibration sensors deployed on the furnace; detecting abnormal points on the furnace using infrared equipment; determining a fault in the furnace based on the vibration frequency and abnormal points; when the fault is determined to be a structural resonance frequency shift, reducing the furnace's energy supply by a preset first limit and activating a preset backup pump for energy supply; when the fault is determined to be a defect in the transmission components of the furnace, reducing the furnace's energy supply by a preset second limit and activating a preset diesel generator for energy supply, wherein the second limit is higher than the first limit.

[0014] Optionally, the aforementioned intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process, wherein determining the fault of the melting furnace based on its vibration frequency and abnormal points includes: performing a Fourier transform on the vibration frequency of the melting furnace to extract its time-frequency characteristics; extracting the dynamic trajectory of the dominant frequency component from the time-frequency characteristics of the melting furnace; calculating the resonance trend level between the dynamic trajectory of the dominant frequency component and the natural frequency of the melting furnace; and calculating the area of ​​the abnormal region of the melting furnace based on the abnormal points on the melting furnace. And calculate the area growth rate of the abnormal area of ​​the furnace, wherein This indicates the number of the anomalies. The number of infrared bands for the abnormal points mentioned by the user. Represents the step function. For the first through the infrared device The band in the first Coordinates of anomaly points The radiation value detected at that location, , For the first The mean and standard deviation of each band. The unit area of ​​the abnormal point; when the resonance trend level is higher than a preset first threshold and the growth rate of the abnormal area is lower than a preset second threshold, the fault of the furnace is determined to be that the furnace has a structural resonance frequency shift; when the resonance trend level is lower than the first threshold and the growth rate of the abnormal area is higher than the second threshold, the fault of the furnace is determined to be that there is a defect in the transmission component in the furnace.

[0015] Optionally, the aforementioned intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process, wherein when the fault of the melting furnace is determined to be a structural resonant frequency shift, the energy supply to the melting furnace is reduced by a preset first amount, and a preset backup pump is activated to supply energy, includes: calculating the main resonant frequency range of the melting furnace based on the dynamic trajectory of the main frequency component in the time-frequency characteristics of the melting furnace, the natural frequency of the melting furnace, and the resonance trend level; comparing the operating frequency of the backup pump with the main resonant frequency range of the melting furnace; prohibiting the backup pump from supplying power when the operating frequency of the backup pump is within the main resonant frequency range of the melting furnace; and controlling the backup pump to supply power according to the current when the operating frequency of the backup pump is outside the main resonant frequency range of the melting furnace. Provide power, of which , This indicates the preset DC bias current. The amplitude of the pulse current of the standby pump is [value missing]. The function is of width rectangular pulse, Indicates the current time. The standby pump is the first The pulse start time of each pulse current.

[0016] Optionally, in the aforementioned intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process, when it is determined that the malfunction of the melting furnace is due to a defect in the transmission components within the melting furnace, the energy supply to the melting furnace is reduced according to a preset second limit, and a preset diesel generator is activated to supply energy, including: based on the area of ​​the abnormal region of the melting furnace... Calculate the fault repair time of the defect in the transmission component. Based on the melting heat requirement, calculate the heat gap caused by the defect in the transmission component. ; Calculate the starting index of the diesel generator. ,in , The preset weighting coefficients, The duration of the defect in the transmission component. The natural base, The unit time power generation cost of the diesel generator; the start-up index of the diesel generator. When the energy level exceeds a preset threshold, the diesel generator is activated to supply energy.

[0017] Secondly, this invention provides an intelligent energy management system for the high-transmittance ultrathin photovoltaic glass manufacturing process, comprising: a data acquisition module for real-time acquisition of melting parameters and transmittance during the photovoltaic glass melting process; a demand module for calculating the melting heat demand of the photovoltaic glass based on a preset melting heat calculation model and the melting parameters; and a residual module for inputting the melting parameters and the melting heat demand into a preset neural network, wherein the neural network is a dual-input gated recurrent unit structure, wherein the main path is used to input the melting parameters, the auxiliary path is used to input the melting heat demand, and the hidden layer is used to perform feature fusion on the melting parameters and the melting heat demand, and calculate nonlinearity based on the fused features. The system comprises: a nonlinear residual term, used to compensate for defects in the calculation results of the melting heat calculation model; a correction module, which corrects the melting heat requirement based on the nonlinear residual term; a detection module, which detects the controllable current process parameters in the photovoltaic glass melting process; a process module, which calculates the minimum melting heat requirement and corresponding target process parameters for the photovoltaic glass to meet the target transmittance requirement based on the transmittance and the preset target transmittance, the current process parameters and the corresponding critical values, and the melting heat requirement; and a melting module, which controls the energy supply of the photovoltaic glass melting process according to the target process parameters to execute the photovoltaic glass melting process.

[0018] The above-described technical solutions of the present invention have at least one or more of the following beneficial effects: According to the technical solution of the present invention, by deeply integrating the mechanism model and deep learning technology, an energy intelligent management and control system integrating precise perception, intelligent diagnosis and collaborative optimization is constructed, which effectively solves the problems of low energy efficiency and difficult quality control in the production of high-transmittance ultra-thin photovoltaic glass, and has significant economic and social benefits. Attached Figure Description

[0019] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0020] Figure 1 A flowchart of an energy intelligent management method for the high-transmittance ultrathin photovoltaic glass manufacturing process according to an embodiment of this application; Figure 2 This is a partial flowchart of an energy intelligent management method for the high-transmittance ultrathin photovoltaic glass manufacturing process according to an embodiment of this application; Figure 3 This is another layout flowchart of the energy intelligent management method for the high-transmittance ultrathin photovoltaic glass manufacturing process according to an embodiment of this application; Figure 4 This is another partial flowchart of the energy intelligent management method for the high-transmittance ultrathin photovoltaic glass manufacturing process according to an embodiment of this application; Figure 5 This is a block diagram of an energy intelligent management and control system for the high-transmittance ultrathin photovoltaic glass manufacturing process according to an embodiment of this application. Detailed Implementation

[0021] Some embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0022] like Figure 1 As shown, one embodiment of the present invention provides an intelligent energy management method for the production process of high-transmittance ultrathin photovoltaic glass, comprising: Step S110: Real-time acquisition of melting parameters and light transmittance during the photovoltaic glass melting process.

[0023] Step S120: Calculate the melting heat requirement of photovoltaic glass based on the preset melting heat calculation model and melting parameters.

[0024] Step S130: Input the melting parameters and melting heat requirements into a preset neural network. The neural network is a dual-input gated recurrent unit structure. The main path is used to input the melting parameters, the auxiliary path is used to input the melting heat requirements, the hidden layer is used to perform feature fusion on the melting parameters and melting heat requirements, and calculate the nonlinear residual term based on the fused features. The nonlinear residual term is used to represent the defect compensation of the calculation result of the melting heat calculation model.

[0025] In this embodiment, a dual-input gated recurrent unit (GRU) neural network structure is constructed. The melting heat demand calculated based on the physical model is used as an auxiliary path input, which is then fused with the main path input of the real-time acquired melting parameters in the hidden layer. This design not only preserves the interpretability and basic accuracy of the physical model, but also utilizes the powerful nonlinear fitting capability of neural networks to effectively identify and quantify the systematic biases caused by traditional heat calculation models neglecting complex dynamic factors (such as heat loss fluctuations, combustion efficiency changes, and kiln aging). By calculating and outputting nonlinear residual terms for defect compensation, dynamic correction of the initial heat demand is achieved, significantly improving the accuracy of heat demand prediction and its adaptability to actual operating conditions.

[0026] Step S140: Correct the melting heat requirement based on the nonlinear residual term.

[0027] Step S150: Detect the controllable current process parameters in the photovoltaic glass melting process.

[0028] Step S160: Based on the light transmittance and the preset target light transmittance, the current process parameters and the corresponding critical values, and the melting heat requirement, calculate the minimum melting heat requirement and the corresponding target process parameters for the photovoltaic glass to meet the target light transmittance requirement.

[0029] In this embodiment, real-time transmittance detection data and preset target values ​​are directly incorporated into the optimization decision-making process. Taking into account the current process status, safety threshold constraints, and precise heat requirements corrected by a neural network, the minimum melting heat requirement and corresponding target process parameters are calculated to meet the high transmittance quality requirements. This breaks the traditional control model where quality assurance and energy saving are separated, achieving a closed-loop optimization goal of "achieving optimal quality with the lowest energy consumption," making it particularly suitable for the production of high-transmittance ultra-thin photovoltaic glass with stringent optical performance requirements.

[0030] Step S170: Control the energy supply of the photovoltaic glass melting process according to the target process parameters to execute the photovoltaic glass melting process.

[0031] In this embodiment, by accurately predicting and dynamically correcting the melting heat, and based on this, solving for the minimum energy consumption process scheme that satisfies quality constraints, the "overheating" phenomenon existing in traditional experience-based control can be avoided, reducing unnecessary energy waste. Executing the melting process according to the calculated optimal target process parameters ensures that energy consumption per unit product is minimized while guaranteeing product quality, significantly reducing production costs and aligning with the development direction of green manufacturing.

[0032] According to the technical solution of this embodiment, by deeply integrating the mechanism model and deep learning technology, an energy intelligent management and control system integrating precise perception, intelligent diagnosis and collaborative optimization is constructed, which effectively solves the problems of low energy efficiency and difficult quality control in the production of high-transmittance ultra-thin photovoltaic glass, and has significant economic and social benefits.

[0033] One embodiment of the present invention provides an intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process. Compared with the aforementioned embodiments, the intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process in this embodiment includes step S110: Real-time temperature monitoring of photovoltaic glass Liquid temperature change rate Raw material mass flow rate Specific heat capacity and the electro-fusion power using electrodes .

[0034] Step S120 includes: (1) According to the formula Calculate the heat of melting ,in, The preset baseline energy conversion efficiency coefficient, Indicates the production year of the electrode. This indicates that the preset electrode is subject to the production life. The attenuation coefficient of the effect Indicates the service life of the electrode. This indicates the preset electrode temperature relative to the photovoltaic glass. The attenuation coefficient of the effect The preset reference temperature, It represents the multiplication operation.

[0035] In this embodiment, by real-time acquisition of key physical parameters such as temperature, liquid temperature change rate, raw material mass flow rate, specific heat capacity, and electric melting power, a composite heat calculation formula incorporating the contributions of sensible heat absorption and effective electrical energy conversion was constructed. This formula not only considers the basic thermal energy required for material heating but, more importantly, introduces the actual conversion efficiency of electrode electrical energy to glass melt thermal energy. This upgrades the initial estimation of heat demand from a static empirical model to a dynamic physical model based on real-time operating conditions, significantly improving the scientific rigor and accuracy of the calculation starting point.

[0036] In this embodiment, an innovative nonlinear decay model for the power conversion efficiency coefficient is proposed. This model comprehensively considers the material aging caused by the service life of the electrode and the dual effects of the glass melt temperature on conductivity and heat radiation loss. This model can dynamically evaluate the actual working efficiency of the electrode at different service stages and different process temperatures, avoiding excessive or insufficient heat input due to neglecting electrode decay, thus extending the electrode's service life and ensuring the stability of the melting process.

[0037] (2) Detect the external temperature of the furnace where the photovoltaic glass is located.

[0038] (3) Calculate the nonlinear heat loss of the photovoltaic glass melting process based on the external temperature of the melting furnace, as well as the material and ventilation of the melting furnace.

[0039] (4) Based on the nonlinear heat loss of the photovoltaic glass melting process, the melting heat requirement of photovoltaic glass is corrected.

[0040] In this embodiment, the detection of external ambient temperature was added, and a nonlinear heat loss calculation model was established by combining the thermal conductivity characteristics of the furnace material and the on-site ventilation conditions. Since heat dissipation from the furnace surface is not a simple linear relationship but is affected by multiple factors such as temperature difference, wind speed, and insulation layer condition, traditional fixed heat loss coefficients are difficult to accurately reflect the actual situation. This embodiment dynamically calculates and corrects this nonlinear heat loss, making the estimation of total melting heat demand closer to the actual energy consumption and reducing control deviations caused by environmental fluctuations (such as seasonal changes, diurnal temperature differences, and changes in workshop airflow).

[0041] According to the technical solution of this embodiment, by introducing an electrode performance decay model and a nonlinear heat loss dynamic compensation mechanism, the physical calculation kernel of melting heat demand is greatly enriched and improved, so that the energy intelligent management and control method is based on a more scientific and comprehensive mechanism. This not only improves the prediction accuracy and robustness, but also provides strong technical support for the refined, green and intelligent management of high-transmittance ultra-thin photovoltaic glass production.

[0042] like Figure 2As shown, one embodiment of the present invention provides an intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process. Compared with the aforementioned embodiments, the intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process in this embodiment further includes the following steps before step S130: Step S210: Train the neural network and calculate the value of the preset loss function during the training process.

[0043] Step S220: Adjust the parameters of the main path and the auxiliary path according to the gradient change of the loss function value.

[0044] Step S230: Based on the gradient change of the loss function value, adjust the baseline power conversion efficiency coefficient. Electrodes are subject to production lifespan The attenuation coefficient of the influence The temperature of the electrodes is affected by the photovoltaic glass. The attenuation coefficient of the influence Or reference temperature Adjustments will be made.

[0045] Traditional physical models typically rely on laboratory calibration or empirical settings for key parameters, making it difficult to adapt to long-term dynamic factors in actual production, such as equipment aging, raw material fluctuations, and environmental changes. This embodiment innovatively treats these physical model parameters as learnable variables, adjusting them inversely based on changes in the loss function gradient during neural network training. This allows the physical model to continuously "self-calibrate" based on massive amounts of real-world operating data, automatically approximating the optimal parameter combination under real-world conditions, significantly improving the universality and long-term prediction accuracy of the melting heat calculation model.

[0046] According to the technical solution of this embodiment, a self-learning and self-correcting intelligent control system is constructed by introducing a joint gradient optimization mechanism of physical model parameters and neural network parameters. This system not only achieves a deeper understanding of the nonlinear characteristics of the melting process, but also significantly improves the accuracy, adaptability, and practicality of the model through bidirectional feedback between mechanisms and data. This provides an innovative technical path for the energy-efficient, stable, and intelligent control of the high-transmittance ultrathin photovoltaic glass production process.

[0047] like Figure 3 As shown, one embodiment of the present invention provides an intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process. Compared with the aforementioned embodiments, the intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process in this embodiment includes the feature fusion of melting parameters and melting heat requirements in step S130, which includes: Step S310, according to the preset time window The first eigenvector is obtained by independently linearly projecting the melting parameters of the main path. .

[0048] Step S320, according to the time window The second eigenvector is obtained by independently linearly projecting the melting heat requirement of the auxiliary path. .

[0049] Step S330, according to the time window The corresponding first feature vector Second eigenvector Calculate the time window Corresponding dynamic fusion intensity ,in For the sigmoid function, In the time window Previous time window The first feature vector generated, Used for concatenating vectors , These are the preset parameters.

[0050] In this embodiment, instead of the traditional simple splicing or fixed-weight fusion method, a dynamic fusion strength calculation method based on the sigmoid gating mechanism is proposed. This fusion strength is not fixed but is determined by the current main path features, auxiliary path features, and historical features of the main path. This design allows the system to dynamically adjust the relative importance of melting parameters (main path) and mechanistic model output (auxiliary path) in feature fusion based on real-time operating conditions and historical trends. For example, it relies more on real-time data during periods of drastic process fluctuations and trusts the calibrated physical model more during steady-state operation, thus achieving more intelligent and robust information fusion.

[0051] In this embodiment, the first feature vector of the previous time window is specifically introduced as input. This design gives the fusion decision a "memory" function, enabling it to capture the changing trends of melting parameters (such as the rate of temperature rise and the slope of electrode power adjustment). It not only focuses on the current state but also senses the dynamic evolution direction of the process, thereby predicting potential energy consumption changes or quality risks in advance and making more forward-looking compensation decisions, effectively improving the control's response speed and stability.

[0052] Step S340, according to the time window Corresponding dynamic fusion intensity For the time window The corresponding first feature vector Second eigenvector Perform feature fusion.

[0053] According to the technical solution of this embodiment, by constructing a dynamic gating fusion mechanism that integrates historical context information, intelligent and adaptive integration of main path (data-driven) and auxiliary path (mechanism-driven) information is achieved. This not only significantly improves the scientificity and effectiveness of feature fusion, but also endows the energy intelligent management and control system with stronger environmental perception, dynamic adaptability, and decision-making transparency, providing key technical support for achieving precise, efficient, and stable operation of the high-transmittance ultrathin photovoltaic glass melting process.

[0054] like Figure 4 As shown, one embodiment of the present invention provides an intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process. Compared with the aforementioned embodiments, the intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process in this embodiment further includes the following steps before step S170: Step S410: Calculate the difference between the target process parameters and the current process parameters.

[0055] Step S420: Based on the difference between the target process parameters and the current process parameters, use S-curve interpolation to generate a sequence of transition process parameters within a preset transition time.

[0056] In the melting and forming process of high-transmittance ultrathin photovoltaic glass, drastic changes in process parameters such as temperature, pressure, and edge pulling force can easily generate uneven thermal or mechanical stress within the glass, leading to defects such as microcracks, warping, and excessive waviness, severely affecting the product's light transmittance and mechanical strength. In this embodiment, by calculating the difference between the target process parameters and the current process parameters, and using S-curve interpolation to generate a sequence of transitional process parameters, the parameter adjustment process exhibits a smooth characteristic of "slow start, stable middle, and gentle finish," avoiding the impact of step changes and fundamentally reducing the quality risks induced by process abrupt changes, thus significantly improving product yield.

[0057] The transition time can be calculated as follows: (1) Detect the thickness and transmission speed of photovoltaic glass.

[0058] (2) Calculate the speed-to-thickness ratio of the photovoltaic glass based on its thickness and transmission speed, which is used to represent the tensile strain rate per unit thickness of the photovoltaic glass.

[0059] (3) Based on the temperature of the photovoltaic glass Calculate the temperature change rate of the photovoltaic glass.

[0060] (4) Calculate the maximum stress change rate of photovoltaic glass based on the speed-to-thickness ratio and temperature change rate of photovoltaic glass, which is used to represent the stress growth rate of photovoltaic glass per unit time.

[0061] (5) Set the transition time according to the maximum stress change rate of the photovoltaic glass.

[0062] This embodiment abandons the crude approach of fixed transition time and proposes an intelligent strategy that dynamically calculates the transition time based on the real-time status of photovoltaic glass. By detecting glass thickness, conveying speed, temperature, and temperature change rate, key physical indicators such as "speed-to-thickness ratio" (characterizing tensile strain rate) and "maximum stress change rate" (characterizing stress growth rate per unit time) are calculated. Based on these, the ability of the glass to withstand stress changes under current operating conditions is assessed, and a safe and reasonable transition time is automatically set. For example, when the glass is thin, the speed is high, or the temperature gradient is large, the transition time is automatically extended to ensure slow stress release, achieving true "on-demand adjustment."

[0063] Step S430: During the transition time, perform the photovoltaic glass melting process according to the transition process parameter sequence.

[0064] After the transition time, proceed to step S170.

[0065] According to the technical solution of this embodiment, by introducing an intelligent transition control mechanism based on S-curve interpolation and dynamic stress assessment, the physical properties of materials are deeply integrated into the process control logic, achieving a closed-loop upgrade from "static optimization" to "dynamic and stable execution." This technology not only effectively solves the quality problems caused by process jumps in the production of high-transmittance ultra-thin photovoltaic glass, but also improves production efficiency while ensuring high product quality through adaptive and refined transition management, providing key process execution guarantees for intelligent manufacturing of photovoltaic glass.

[0066] One embodiment of the present invention provides a method for intelligent energy management in the production process of high-transmittance ultrathin photovoltaic glass. Compared with the foregoing embodiments, the method for intelligent energy management in the production process of high-transmittance ultrathin photovoltaic glass in this embodiment further includes: (1) The vibration frequency of the furnace is detected by a vibration sensor deployed on the furnace.

[0067] (2) Detect abnormal points on the furnace using infrared equipment.

[0068] (3) Determine the fault of the furnace based on the vibration frequency and abnormal points of the furnace.

[0069] In this embodiment, by deploying vibration sensors to monitor the resonant frequency shift of the furnace structure in real time, and combining this with infrared equipment to detect local abnormal high-temperature points, potential equipment hazards can be detected in advance.

[0070] (4) When the fault of the furnace is determined to be a structural resonance frequency shift, the energy supply of the furnace will be reduced according to the preset first limit, and the preset backup pump will be activated to supply energy.

[0071] (5) When it is determined that the fault of the furnace is a defect in the transmission component, the energy supply of the furnace is reduced according to the preset second limit, and the preset diesel generator is activated to supply energy. The second limit is higher than the first limit.

[0072] According to the technical solution of this embodiment, a differentiated energy supply strategy is adopted for different fault types: when the structure resonates and shifts, the energy supply is adjusted according to the preset first limit (such as a reduction of 10%-15%) and the backup pump is switched to avoid equipment overload damage; when the transmission component is defective, the load is reduced according to the higher second limit (such as a reduction of 20%-25%), and the diesel generator is activated to maintain the stability of key process parameters to prevent serious accidents such as glass liquid solidification caused by sudden power failure.

[0073] One embodiment of the present invention provides an intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process. Compared with the previous embodiments, the intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process in this embodiment determines the fault of the melting furnace based on the vibration frequency and abnormal points of the melting furnace, including: (1) Perform Fourier transform on the vibration frequency of the melting furnace to extract the time-frequency characteristics of the melting furnace.

[0074] In this embodiment, an adaptive Fourier neural operator is used to enhance the frequency domain features of the vibration signal, which can identify minute frequency shifts.

[0075] (2) Dynamic trajectory of the main frequency component from the time-frequency characteristics of the melting furnace.

[0076] (3) Calculate the resonance trend level between the dynamic trajectory of the main frequency component in the time-frequency characteristics of the melting furnace and the natural frequency of the melting furnace.

[0077] In this embodiment, by calculating the resonance trend between the dynamic trajectory of the main frequency component and the natural frequency, structural resonance shift can be predicted in advance.

[0078] (4) Calculate the area of ​​the abnormal region of the furnace based on the abnormal points on the furnace. And calculate the growth rate of the abnormal area of ​​the melting furnace, where Indicates the number of outliers. The number of infrared bands for user anomaly points. Represents the step function. For the first infrared device The band in the first Coordinates of anomaly points The radiation value detected at that location, , For the first The mean and standard deviation of each band. The unit area is the outlier.

[0079] In this embodiment, the abnormal region area calculation model solves the problem of traditional single-band detection being susceptible to environmental interference by using multi-band data fusion and dynamic threshold adjustment: Standardization processing significantly improves the robustness of detection results to environmental temperature fluctuations. Real-time calculation of the abnormal region area growth rate is achieved through unit area integration, providing data support for predicting fault development trends. Based on multi-band infrared radiation data, an abnormal region area calculation model is constructed using a step function, which can accurately identify temperature anomalies.

[0080] (5) When the resonance trend level is higher than the preset first threshold and the abnormal area growth rate is lower than the preset second threshold, the fault of the furnace is determined to be that the furnace has a structural resonance frequency shift.

[0081] (6) When the resonance trend level is lower than the first threshold and the abnormal area growth rate is higher than the second threshold, the fault of the furnace is judged to be a defect in the transmission component in the furnace.

[0082] According to the technical solution of this embodiment, by integrating vibration spectrum analysis and infrared thermal imaging technology, accurate diagnosis of the furnace operation status and intelligent control of energy supply are achieved.

[0083] One embodiment of the present invention provides an intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process. Compared with the aforementioned embodiments, the intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process in this embodiment, when determining that the furnace malfunction is a structural resonant frequency shift, reduces the furnace's energy supply by a preset first limit and activates a preset backup pump for energy supply, including: (1) Calculate the main resonance frequency range of the furnace based on the dynamic trajectory of the main frequency component in the time-frequency characteristics of the furnace, the natural frequency of the furnace, and the resonance trend level. (2) Compare the operating frequency of the standby pump with the main resonant frequency range of the furnace.

[0084] (3) When the operating frequency of the standby pump is in the main resonant frequency range of the furnace, the standby pump shall not be powered.

[0085] In this embodiment, when the operating frequency of the standby pump falls into the main resonance range of the melting furnace, the power supply is immediately cut off to avoid the mechanical resonance amplification effect.

[0086] (4) When the operating frequency of the standby pump is outside the main resonant frequency range of the furnace, control the standby pump according to the current. Provide power, of which , This indicates the preset DC bias current. This refers to the amplitude of the pulse current of the standby pump. The function is of width rectangular pulse, Indicates the current time. Indicates the number of standby pumps The pulse start time of each pulse current.

[0087] In this embodiment, the pulse power supply strategy described in the above formula is adopted. By superimposing DC bias and pulse current, the pump operating frequency is kept stable within a safe range. In the formula, the pulse width parameter is matched to the thermal inertia characteristics of the melting furnace, thereby reducing the pump's energy consumption.

[0088] According to the technical solution of this embodiment, by integrating vibration spectrum analysis and adaptive current control technology, the risk of furnace resonance and the precise control of energy supply are effectively avoided.

[0089] One embodiment of the present invention provides an intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process. Compared with the aforementioned embodiments, the intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process in this embodiment, when determining that the furnace malfunction is due to a defect in the transmission components within the furnace, reduces the energy supply to the furnace according to a preset second limit and activates a preset diesel generator for energy supply, including: (1) Based on the area of ​​the abnormal zone of the melting furnace Calculate the fault repair time for defects in transmission components. .

[0090] (2) Calculate the heat gap caused by defects in the transmission components based on the melting heat requirements. .

[0091] (3) Calculate the starting index of the diesel generator. ,in , The preset weighting coefficients, This refers to the duration of the defect in the transmission component. The natural base, This represents the unit time cost of generating electricity from a diesel generator.

[0092] In this embodiment, a fault repair time prediction model can be established based on the area of ​​the abnormal region and combined with historical maintenance data, and a heat gap analysis model can be established based on the defect area. The calculation model of the diesel generator start-up index integrates thermodynamic and economic constraints, taking into account the impact of power generation costs.

[0093] (4) The starting index of the diesel generator When the energy level exceeds a preset threshold, the diesel generator will be activated to supply energy.

[0094] According to the technical solution of this embodiment, by establishing an intelligent decision-making model for defects in transmission components, precise control of energy supply and production continuity assurance are achieved.

[0095] One embodiment of the present invention provides a method for intelligent energy management in the production process of high-transmittance ultrathin photovoltaic glass. Compared with the foregoing embodiments, the method for intelligent energy management in the production process of high-transmittance ultrathin photovoltaic glass in this embodiment further includes: (1) Obtain the historical maintenance time of the furnace failure.

[0096] (2) Calculate the expected repair time of the furnace based on the degree of the furnace failure and the historical maintenance time.

[0097] In this embodiment, an expected repair time prediction model that combines historical maintenance time with the current fault level is introduced to scientifically assess the maintenance cycle.

[0098] (3) Determine whether the energy reserves of the standby pump or diesel generator meet the energy supply during the expected repair time.

[0099] (4) If the judgment result is negative, trigger an early warning for the energy reserves of the standby pump or diesel generator.

[0100] According to the technical solution of this embodiment, the system automatically determines whether the energy reserves of the backup pump or diesel generator are sufficient to support the emergency power supply needs during the repair period. If not, an energy reserve warning is triggered in advance to remind maintenance personnel to replenish fuel or switch to backup equipment in a timely manner, ensuring that critical production links remain powered and operational during fault handling, significantly improving the system's resilience and the level of intelligent energy dispatch.

[0101] like Figure 5 As shown, one embodiment of the present invention provides an intelligent energy management system for the high-transmittance ultrathin photovoltaic glass manufacturing process, comprising: The data acquisition module 510 collects the melting parameters and light transmittance in the photovoltaic glass melting process in real time.

[0102] Demand module 520 calculates the melting heat requirement of photovoltaic glass based on the preset melting heat calculation model and melting parameters.

[0103] The residual module 530 inputs the melting parameters and melting heat requirements into a preset neural network. The neural network is a dual-input gated recurrent unit structure, in which the main path is used to input the melting parameters and the auxiliary path is used to input the melting heat requirements. The hidden layer is used to perform feature fusion on the melting parameters and melting heat requirements, and calculates nonlinear residual terms based on the fused features. The nonlinear residual terms are used to represent the defect compensation of the calculation results of the melting heat calculation model.

[0104] In this embodiment, a dual-input gated recurrent unit (GRU) neural network structure is constructed. The melting heat demand calculated based on the physical model is used as an auxiliary path input, which is then fused with the main path input of the real-time acquired melting parameters in the hidden layer. This design not only preserves the interpretability and basic accuracy of the physical model, but also utilizes the powerful nonlinear fitting capability of neural networks to effectively identify and quantify the systematic biases caused by traditional heat calculation models neglecting complex dynamic factors (such as heat loss fluctuations, combustion efficiency changes, and kiln aging). By calculating and outputting nonlinear residual terms for defect compensation, dynamic correction of the initial heat demand is achieved, significantly improving the accuracy of heat demand prediction and its adaptability to actual operating conditions.

[0105] The correction module 540 corrects the melting heat requirement based on the nonlinear residual term.

[0106] The detection module 550 detects the controllable current process parameters in the photovoltaic glass melting process.

[0107] Process module 560 calculates the minimum heat of melting required for photovoltaic glass to meet the target transmittance requirement and the corresponding target process parameters, based on the transmittance and preset target transmittance, current process parameters and corresponding critical values, and melting heat requirements.

[0108] In this embodiment, real-time transmittance detection data and preset target values ​​are directly incorporated into the optimization decision-making process. Taking into account the current process status, safety threshold constraints, and precise heat requirements corrected by a neural network, the minimum melting heat requirement and corresponding target process parameters are calculated to meet the high transmittance quality requirements. This breaks the traditional control model where quality assurance and energy saving are separated, achieving a closed-loop optimization goal of "achieving optimal quality with the lowest energy consumption," making it particularly suitable for the production of high-transmittance ultra-thin photovoltaic glass with stringent optical performance requirements.

[0109] The melting module 570 controls the energy supply for the photovoltaic glass melting process according to the target process parameters in order to execute the photovoltaic glass melting process.

[0110] In this embodiment, by accurately predicting and dynamically correcting the melting heat, and based on this, solving for the minimum energy consumption process scheme that satisfies quality constraints, the "overheating" phenomenon existing in traditional experience-based control can be avoided, reducing unnecessary energy waste. Executing the melting process according to the calculated optimal target process parameters ensures that energy consumption per unit product is minimized while guaranteeing product quality, significantly reducing production costs and aligning with the development direction of green manufacturing.

[0111] According to the technical solution of this embodiment, by deeply integrating the mechanism model and deep learning technology, an energy intelligent management and control system integrating precise perception, intelligent diagnosis and collaborative optimization is constructed, which effectively solves the problems of low energy efficiency and difficult quality control in the production of high-transmittance ultra-thin photovoltaic glass, and has significant economic and social benefits.

[0112] The basic principles of this application have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this application are merely examples and not limitations, and should not be considered as essential features of each embodiment of this application. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the application to the necessity of employing the aforementioned specific details for implementation.

[0113] The block diagrams of devices, apparatuses, devices, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0114] It should also be noted that in the apparatus, equipment, and methods of this application, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions of this application.

[0115] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this application. Therefore, this application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0116] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this application to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. A method for intelligent energy management in the production process of high-transmittance ultrathin photovoltaic glass, including: Real-time acquisition of melting parameters and light transmittance during the photovoltaic glass melting process, including: Real-time acquisition of the photovoltaic glass Liquid temperature change rate Raw material mass flow rate Specific heat capacity and the electro-fusion power using electrodes , Based on the preset melting heat calculation model and the melting parameters, the melting heat requirement of the photovoltaic glass is calculated, including: According to the formula ; Calculate the required melting heat ,in, The preset baseline energy conversion efficiency coefficient, This indicates the production year of the electrode. This indicates that the electrode is subject to the preset production years. The attenuation coefficient of the effect This indicates the service life of the electrode. This indicates that the electrode is subject to the temperature of the photovoltaic glass as preset. The attenuation coefficient of the effect The preset reference temperature, Indicates multiplication operation; The melting heat requirement of the photovoltaic glass is calculated based on the preset melting heat calculation model and the melting parameters. The melting parameters and the melting heat requirement are input into a preset neural network. The neural network is a dual-input gated recurrent unit structure. The main path is used to input the melting parameters, the auxiliary path is used to input the melting heat requirement, and the hidden layer is used to perform feature fusion on the melting parameters and the melting heat requirement, and calculate a nonlinear residual term based on the fused features. The nonlinear residual term is used to represent the defect compensation of the calculation result of the melting heat calculation model. The melting heat requirement is corrected based on the nonlinear residual term. Detect the controllable current process parameters in the photovoltaic glass melting process; Based on the light transmittance and the preset target light transmittance, the current process parameters and the corresponding critical values, and the melting heat requirement, calculate the minimum melting heat requirement and the corresponding target process parameters for the photovoltaic glass to meet the target light transmittance requirement; The energy supply for the photovoltaic glass melting process is controlled according to the target process parameters to execute the photovoltaic glass melting process.

2. The intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process according to claim 1, wherein, The calculation of the melting heat requirement of the photovoltaic glass based on the preset melting heat calculation model and the melting parameters further includes: The external temperature of the furnace containing the photovoltaic glass is detected; The nonlinear heat loss of the photovoltaic glass melting process is calculated based on the external temperature of the furnace, the material of the furnace, and the ventilation conditions. The heat requirement for melting the photovoltaic glass is adjusted based on the nonlinear heat loss in the photovoltaic glass melting process. The energy intelligent management method for the high-transmittance ultrathin photovoltaic glass production process further includes the following steps, prior to inputting the melting parameters and melting heat requirements into a preset neural network: The neural network is trained, and the value of a preset loss function is calculated during the training process; Based on the gradient change of the loss function value, the parameters of the main path and the parameters of the auxiliary path are adjusted; Based on the gradient change of the value of the loss function, the baseline power conversion efficiency coefficient is adjusted. The electrode is subject to the number of years of production. The attenuation coefficient of the influence The electrode is affected by the temperature of the photovoltaic glass. The attenuation coefficient of the influence or the reference temperature Adjustments will be made.

3. The intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process according to claim 1, wherein, Feature fusion of the melting parameters and the melting heat requirement includes: According to the preset time window The first feature vector is obtained by independently linearly projecting the melting parameters of the main path. ; According to the time window The second feature vector is obtained by independently linearly projecting the melting heat requirement of the auxiliary path. ; According to the time window The corresponding first feature vector Second eigenvector Calculate the time window Corresponding dynamic fusion intensity ,in For the sigmoid function, In the time window Previous time window The first feature vector generated is represented by []. , These are preset parameters; According to the time window Corresponding dynamic fusion intensity For the time window The corresponding first feature vector Second eigenvector Perform feature fusion.

4. The intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process according to claim 1, wherein, Before controlling the energy supply for the photovoltaic glass melting process according to the target process parameters to execute the photovoltaic glass melting process, the intelligent energy management method for the high-transmittance ultrathin photovoltaic glass production process further includes: Calculate the difference between the target process parameter and the current process parameter; Based on the difference between the target process parameters and the current process parameters, an S-curve interpolation method is used to generate a sequence of transition process parameters within a preset transition time. During the transition time, the photovoltaic glass melting process is performed according to the transition process parameter sequence. After the transition time, the energy supply for the photovoltaic glass melting process is controlled according to the target process parameters to execute the photovoltaic glass melting process. The energy intelligent management method for the high-transmittance ultrathin photovoltaic glass manufacturing process further includes the following, prior to generating the transition process parameter sequence within a preset transition time: The thickness and conveying speed of the photovoltaic glass are detected; Based on the thickness of the photovoltaic glass and the conveying speed, the speed-to-thickness ratio of the photovoltaic glass is calculated to represent the tensile strain rate per unit thickness of the photovoltaic glass. According to the temperature of the photovoltaic glass Calculate the temperature change rate of the photovoltaic glass; Based on the velocity-to-thickness ratio and temperature change rate of the photovoltaic glass, the maximum stress change rate of the photovoltaic glass is calculated, which is used to represent the stress growth rate of the photovoltaic glass per unit time. The transition time is set according to the maximum stress change rate of the photovoltaic glass.

5. The intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process according to claim 1, wherein, The intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process also includes: The vibration frequency of the furnace is detected by vibration sensors deployed on the furnace. Anomalies on the melting furnace are detected using infrared equipment; The malfunction of the furnace is determined based on the vibration frequency and abnormal points of the furnace. When the fault of the furnace is determined to be a structural resonant frequency shift, the energy supply of the furnace will be reduced by a preset first limit, and a preset backup pump will be activated to supply energy. When it is determined that the fault of the melting furnace is a defect in the transmission component, the energy supply of the melting furnace is reduced according to a preset second limit, and a preset diesel generator is activated to supply energy. The second limit is higher than the first limit.

6. The intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process according to claim 5, wherein, Based on the vibration frequency and abnormal points of the furnace, the malfunction of the furnace is determined, including: The vibration frequency of the melting furnace is subjected to Fourier transform to extract the time-frequency characteristics of the melting furnace; The dynamic trajectory of the dominant frequency component from the time-frequency characteristics of the melting furnace; Calculate the resonance trend level between the dynamic trajectory of the dominant frequency component in the time-frequency characteristics of the melting furnace and the natural frequency of the melting furnace; Calculate the area of ​​the abnormal region of the melting furnace based on the abnormal points on the furnace. And calculate the growth rate of the abnormal area of ​​the furnace, wherein This indicates the number of the anomalies. The number of infrared bands for the abnormal points mentioned by the user. Represents the step function. For the first through the infrared device The band in the first Coordinates of anomaly points The radiation value detected at that location, , For the first The mean and standard deviation of each band. The unit area of ​​the anomaly point; When the resonance trend level is higher than a preset first threshold and the abnormal area growth rate is lower than a preset second threshold, the fault of the furnace is determined to be that the furnace has a structural resonance frequency shift. When the resonance trend level is lower than the first threshold and the abnormal area growth rate is higher than the second threshold, the fault of the furnace is determined to be a defect in the transmission component of the furnace.

7. The intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process according to claim 6, wherein, When the fault of the melting furnace is determined to be a structural resonant frequency shift, the energy supply to the melting furnace is reduced by a preset first limit, and a preset backup pump is activated to supply energy, including: Based on the dynamic trajectory of the dominant frequency component in the time-frequency characteristics of the furnace, the natural frequency of the furnace, and the resonance trend level, the main resonance frequency range of the furnace is calculated. The operating frequency of the standby pump is compared with the main resonant frequency range of the furnace. When the operating frequency of the standby pump is within the main resonant frequency range of the furnace, the standby pump shall be prohibited from receiving power. When the operating frequency of the standby pump is outside the main resonant frequency range of the furnace, the standby pump is controlled according to the current... Provide power, of which , This indicates the preset DC bias current. The amplitude of the pulse current of the standby pump is [value missing]. The function is of width rectangular pulse, Indicates the current time. The standby pump is the first The pulse start time of each pulse current.

8. The intelligent energy management method for the high-transmittance ultrathin photovoltaic glass manufacturing process according to claim 6, wherein, When the malfunction of the melting furnace is determined to be a defect in the transmission components, the energy supply to the melting furnace is reduced according to a preset second limit, and a preset diesel generator is activated to supply energy, including: Based on the area of ​​the abnormal zone of the furnace Calculate the fault repair time of the defect in the transmission component. ; Based on the melting heat requirement, calculate the heat gap caused by the defect in the transmission component. ; Calculate the starting index of the diesel generator. ,in , The preset weighting coefficients, The duration of the defect in the transmission component. The natural base, The unit time power generation cost of the diesel generator; The starting index of the diesel generator When the energy level exceeds a preset threshold, the diesel generator is activated to supply energy.

9. An intelligent energy management system for the high-transmittance ultra-thin photovoltaic glass manufacturing process, including: The data acquisition module collects melting parameters and light transmittance in real time during the photovoltaic glass melting process, including: Real-time acquisition of the photovoltaic glass Liquid temperature change rate Raw material mass flow rate Specific heat capacity and the electro-fusion power using electrodes , Based on the preset melting heat calculation model and the melting parameters, the melting heat requirement of the photovoltaic glass is calculated, including: According to the formula ; Calculate the required melting heat ,in, The preset baseline energy conversion efficiency coefficient, This indicates the production year of the electrode. This indicates that the electrode is subject to the preset production years. The attenuation coefficient of the influence This indicates the service life of the electrode. This indicates that the electrode is subject to the temperature of the photovoltaic glass as preset. The attenuation coefficient of the effect The preset reference temperature, Indicates multiplication operation; The demand module calculates the melting heat requirement of the photovoltaic glass based on the preset melting heat calculation model and the melting parameters. The residual module inputs the melting parameters and the melting heat requirement into a preset neural network. The neural network is a dual-input gated recurrent unit structure, in which the main path is used to input the melting parameters and the auxiliary path is used to input the melting heat requirement. The hidden layer is used to perform feature fusion on the melting parameters and the melting heat requirement, and calculates a nonlinear residual term based on the fused features. The nonlinear residual term is used to represent the defect compensation of the calculation result of the melting heat calculation model. The correction module corrects the melting heat requirement based on the nonlinear residual term; The detection module detects the controllable current process parameters in the photovoltaic glass melting process; The process module calculates the minimum heat of melting required for the photovoltaic glass to meet the target light transmittance requirement and the corresponding target process parameters, based on the light transmittance and the preset target light transmittance, the current process parameters and the corresponding critical values, and the melting heat requirement. The melting module controls the energy supply for the photovoltaic glass melting process according to the target process parameters in order to execute the photovoltaic glass melting process.

Images