Design method for reducing energy consumption of starting up drying furnace

Abstract

The present application relates to the technical field of drying furnace, and particularly relates to a design method for reducing energy consumption of starting of a drying furnace, which comprises the following steps: obtaining an ambient temperature of the drying furnace; inputting the ambient temperature into a pre-trained temperature rising time length prediction model to obtain a predicted temperature rising time length; calculating an optimal starting time point of the drying furnace based on the predicted temperature rising time length and a preset target process temperature reaching time; starting the drying furnace at the optimal starting time point; compiling the temperature rising time length prediction model into an executable program and embedding the executable program into a manufacturing execution system; outputting the calculated optimal starting time point through a man-machine interaction interface of the manufacturing execution system to prompt an operator to start the drying furnace according to the optimal starting time point. The present application can accurately predict the temperature rising time length according to the current ambient temperature and then determine the optimal starting time point, so as to reduce the energy consumption of the starting stage of the drying furnace.

Description

Technical Field

[0001] This application relates to the field of drying oven technology, and in particular to a design method for reducing the start-up energy consumption of a drying oven. Background Technology

[0002] The drying oven in the painting workshop is a core thermal equipment in the process chain. Its functional integrity and control precision directly determine the final physicochemical properties of the coating material. After wet film processes such as electrophoresis, adhesive application, and spraying, the drying oven, through heating, holding, and cooling processes, allows the resin of the coating or adhesive to fully cross-link and cure, thereby achieving the key engineering indicators of coating performance curing. However, the drying oven is also a major energy consumer in the painting workshop, and its operating costs directly affect the overall cost of automobile production. Taking a production line with a capacity of 240,000 vehicles per year as an example, the drying oven consumes 18,088.64 kWh of electricity and 11,898.56 m³ of natural gas per day. Because the drying oven must reach the process temperature inside the furnace before the workpiece reaches the furnace inlet, it needs to be started some time in advance.

[0003] Currently, the start-up and shutdown of drying ovens are mainly done manually. Operators rely on personal experience to determine the start-up timing, making it difficult to adjust the equipment's operating status in a timely manner according to changes in external ambient temperature, production conditions, and other factors. This results in difficulties in meeting the required temperature arrival time for the process and hinders precise control of the start-up time. Figure 1 As shown in the figure, the comparison data of the heating time and waiting time when a certain drying oven is started shows that the waiting time sometimes even exceeds the heating time, resulting in a large amount of energy waste.

[0004] Analysis revealed that the aforementioned problem stems from the lack of a mechanism in existing technologies capable of dynamically predicting the heating time of the drying oven based on real-time ambient temperature. Operators cannot accurately determine how far in advance the drying oven needs to be started under current environmental conditions, leading to premature startup based on experience and unnecessary energy waste during the start-up phase. Therefore, providing a method to precisely control the start-up timing of the drying oven and reduce energy consumption during startup is a crucial technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a design method for reducing the energy consumption of drying ovens during startup, in order to solve the problem that in the prior art, drying ovens use manual start-up and shutdown, which cannot accurately control the startup timing according to the ambient temperature, resulting in serious energy waste during waiting.

[0006] This invention provides the following solution:

[0007] A design method for reducing the start-up energy consumption of a drying oven includes:

[0008] Obtain the ambient temperature of the drying oven;

[0009] The ambient temperature is input into a pre-trained temperature rise duration prediction model to obtain the predicted temperature rise duration;

[0010] Based on the predicted heating time and the preset target process temperature arrival time, the optimal start-up time of the drying furnace is calculated.

[0011] The drying oven is started at the optimal start-up time.

[0012] Furthermore, the process of establishing the heating duration prediction model is as follows:

[0013] Acquire the historical startup data of the drying oven, which includes historical ambient temperature and its corresponding historical heating duration;

[0014] Based on the historical startup data, a regression algorithm is used to fit the data and generate the heating duration prediction model.

[0015] Furthermore, the step of generating the temperature rise duration prediction model by fitting the historical startup data using a regression algorithm includes:

[0016] A scatter plot is drawn with the historical ambient temperature as the horizontal axis and the historical temperature rise duration as the vertical axis.

[0017] Based on the scatter plot, a regression curve is fitted, and the regression curve is used as the prediction model for the heating duration.

[0018] Furthermore, the step of inputting the ambient temperature into a pre-trained temperature rise duration prediction model to obtain the predicted temperature rise duration includes:

[0019] The ambient temperature is input into the heating time prediction model corresponding to each of the multiple process sections to obtain multiple predicted heating times.

[0020] The calculation of the optimal start-up time of the drying oven based on the predicted heating time and the preset target process temperature arrival time includes:

[0021] The maximum value among the multiple predicted heating times is used as the predicted heating time for calculating the optimal start-up time.

[0022] Furthermore, obtaining the ambient temperature of the drying oven includes:

[0023] The initial temperature of the drying oven before startup is collected in real time by a temperature sensor installed inside or at the entrance of the drying oven, and is used as the ambient temperature.

[0024] Furthermore, in response to the start-up of the drying oven, the real-time temperature of the drying oven is continuously acquired at fixed time intervals;

[0025] The real-time temperature is input into the heating duration prediction model to update the predicted heating duration.

[0026] Determine whether the real-time temperature has reached the target temperature. If not, repeat the above steps until the target temperature is reached.

[0027] Furthermore, the fixed time interval is 5 minutes.

[0028] Furthermore, the calculation of the optimal start-up time of the drying oven based on the predicted heating time and the preset target process temperature arrival time includes:

[0029] The optimal start-up time is obtained by subtracting the predicted heating time from the time it takes for the target process temperature to be reached.

[0030] Furthermore, the heating time prediction model is compiled into an executable program, and the executable program is embedded into the manufacturing execution system;

[0031] The calculated optimal start-up time is output through the human-machine interface of the manufacturing execution system to prompt the operator to start the drying oven at the optimal start-up time.

[0032] Furthermore, the heating duration prediction model is a regression model, which is obtained by fitting historical startup data using the least squares method.

[0033] The above solution achieves the following beneficial technical effects:

[0034] This invention collects historical temperature data at the start-up of the drying furnace and the time data for reaching the process temperature, fits a regression curve to establish a predictive model, and can accurately predict the heating time based on the current ambient temperature, thereby calculating the optimal start-up time. Operators or the control system can start the drying furnace at this time, effectively avoiding energy consumption caused by premature start-up. Practical verification shows that after adopting this invention, the waiting time for a single start-up of a certain drying furnace can be reduced by 21.38 minutes, and the cost per unit can be reduced by 1.26 yuan. Long-term operation can bring considerable economic benefits to enterprises.

[0035] In particular, this invention utilizes a regression model to dynamically calculate the start-up time based on real-time ambient temperature, completely overcoming the uncertainty and lag inherent in manual experience-based operations. By compiling the model into an executable program and embedding it into the manufacturing execution system, automated and intelligent decision-making regarding the start-up time of the drying oven can be achieved without manual intervention. Simultaneously, the predicted duration is dynamically updated at fixed time intervals during the heating process, realizing closed-loop monitoring and further improving the robustness and reliability of control, providing a technological foundation for unmanned and intelligent operation in coating workshops.

[0036] In particular, this invention is not dependent on any specific programming language, machine learning framework, or hardware platform, and possesses strong versatility. This invention discloses in detail the complete implementation path from historical data collection, regression model establishment (including the specific formula for the least squares method), modular program design, to manufacturing execution system integration; those skilled in the art can directly implement it based on the specification. Furthermore, through a modular path definition method (such as equipment identifier_process section type_model file path), this invention can be easily extended to multiple drying ovens of different models and process sections, making it suitable for new production lines and energy-saving retrofits of old production lines, and has broad prospects for widespread application. Attached Figure Description

[0037] Figure 1 A flowchart illustrating a design method for reducing the start-up energy consumption of a drying oven according to an embodiment of the present invention is shown.

[0038] Figure 2 A schematic diagram of the fitting curve according to an embodiment of the present invention is shown.

[0039] Figure 3 A business process diagram according to an embodiment of the present invention is shown.

[0040] Figure 4 A diagram of the E-MES management interface according to an embodiment of the present invention is shown. Detailed Implementation

[0041] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. 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.

[0042] It should be noted that all uses of "first" and "second" in the embodiments of the present invention are for the purpose of distinguishing two entities or parameters with the same name but different names. It is clear that "first" and "second" are only for the convenience of expression and should not be construed as limiting the embodiments of the present invention. Subsequent embodiments will not explain this in detail.

[0043] like Figure 1 As shown, this embodiment of the invention provides a design method for reducing the start-up energy consumption of a drying oven, including:

[0044] Step S1: Obtain the ambient temperature of the drying oven.

[0045] In this embodiment of the invention, ambient temperature refers to the initial temperature inside the furnace of the drying oven before startup, or the ambient temperature of the workshop where the drying oven is located. This temperature can be acquired in real time by a temperature sensor (e.g., thermocouple, resistance temperature detector) installed inside or at the entrance of the drying oven. This temperature value serves as the basis for subsequent prediction of the heating time.

[0046] Step S2: Input the ambient temperature into the pre-trained heating duration prediction model to obtain the predicted heating duration.

[0047] Specifically, the warming duration prediction model is a mathematical model pre-trained using machine learning or regression algorithms. This model reflects the mapping relationship between ambient temperature and warming duration. For example... Figure 2 As shown, the model can be a regression curve or a nonlinear function in the form of a neural network. The ambient temperature value obtained in step 1 is substituted into the model as an input parameter. After calculation, the model outputs the corresponding predicted heating time (e.g., the time required to rise from the current temperature to the set process temperature, in minutes).

[0048] Step S3: Based on the predicted heating time and the preset target process temperature arrival time, calculate the optimal start-up time of the drying oven.

[0049] In this embodiment of the invention, the time to reach the target process temperature is predetermined by the production plan. For example, it is required that the furnace temperature of the drying oven reach 200°C at 8:00 AM. Assuming that the predicted heating time calculated in step 2 is 30 minutes, the optimal start-up time is the target process temperature arrival time minus the predicted heating time, i.e., 7:30 AM. This calculation can be performed by the arithmetic logic unit inside the controller.

[0050] Step S4: Start the drying oven at the optimal start-up time.

[0051] In this embodiment of the invention, when the calculated optimal start-up time (e.g., 7:30) is reached, the controller automatically sends a start-up command to the burners, fans, and other actuators of the drying oven; alternatively, the time point is displayed on the human-machine interface as a text or voice prompt, allowing the operator to manually start the oven. Through these methods, timely start-up of the drying oven can be achieved.

[0052] The present invention, through the above-described scheme, can dynamically adjust the start-up timing according to the current ambient temperature, avoiding energy waste caused by starting too early and production waiting caused by starting too late, thereby effectively reducing energy consumption during the start-up phase of the drying oven.

[0053] In some embodiments of the present invention, based on the above technical solution, the step of establishing the heating duration prediction model is also included.

[0054] Specifically, the method further includes:

[0055] Acquire the historical startup data of the drying oven, which includes historical ambient temperature and its corresponding historical heating duration;

[0056] Based on the historical startup data, a regression algorithm is used to fit the data and generate the heating duration prediction model.

[0057] In practical applications, each time the drying furnace completes a startup process, the system automatically records the initial ambient temperature at the time of startup and the total time elapsed from startup to reaching the target process temperature (i.e., the heating time). It should be noted that the historical heating time is defined as the total time elapsed from the moment the drying furnace is powered on until the moment its internal temperature first reaches the preset target process temperature. After multiple startups (e.g., dozens or hundreds of times), these data points constitute a historical startup dataset. Then, regression algorithms (e.g., linear regression, multinomial regression, etc.) are used to fit this data to obtain a mathematical model that describes the functional relationship between ambient temperature and heating time. This model is the heating time prediction model. After the model is trained, it can be stored locally or on a server for subsequent real-time prediction.

[0058] In some embodiments of the present invention, the fitting process of establishing the heating duration prediction model described above may specifically include:

[0059] A scatter plot is drawn with the historical ambient temperature as the horizontal axis and the historical temperature rise duration as the vertical axis.

[0060] Based on the scatter plot, a regression curve is fitted, and the regression curve is used as the prediction model for the heating duration.

[0061] like Figure 2 As shown, each historical startup record is considered a coordinate point, with the horizontal axis representing the historical ambient temperature (°C) and the vertical axis representing the historical temperature rise time (min). All points form a scatter plot. Using the least squares method or a similar fitting algorithm, an optimal trend line (i.e., regression curve) is found among these scatter points. This curve can be a straight line or a smooth curve. This regression curve is the temperature rise time prediction model. When a new ambient temperature value is input, simply find the point corresponding to that value on the horizontal axis, draw a vertical line upwards intersecting the regression curve, and then draw a horizontal line to the left intersecting the vertical axis. Reading the vertical coordinate value of the intersection point yields the predicted temperature rise time. This regression curve-based model is intuitive and easy to implement.

[0062] In some embodiments of the present invention, for drying ovens having multiple process sections (e.g., electrophoresis section, adhesive drying section, topcoat section, etc.), steps 2 and 3 above can be optimized as follows:

[0063] The drying oven includes multiple process sections;

[0064] The step of inputting the ambient temperature into a pre-trained heating time prediction model to obtain the predicted heating time includes: inputting the ambient temperature into the heating time prediction model corresponding to each of the multiple process segments to obtain multiple predicted heating times;

[0065] The step of calculating the optimal start-up time of the drying furnace based on the predicted heating time and the preset target process temperature arrival time includes: taking the maximum value among the multiple predicted heating times as the predicted heating time for calculating the optimal start-up time.

[0066] Because different process sections have different target temperatures (e.g., 160℃ for electrophoresis and 180℃ for topcoat), and their heating power and thermal inertia also differ, the heating time for each section often varies. To ensure that all process sections are ready before the workpiece arrives, the longest heating time must be taken as the start-up lead time for the entire drying oven. Specifically, an independent heating time prediction model is trained for each process section. Under the same ambient temperature, these models are called to obtain the predicted heating times T1, T2, ..., Tn for each section, and then T_max = max(T1, T2, ..., Tn). Finally, the optimal start-up time is calculated based on T_max. This ensures that no process section will delay production due to slow heating.

[0067] In some embodiments of the present invention, the method of obtaining ambient temperature is specifically defined to ensure the reliability and real-time performance of the data:

[0068] The process of obtaining the ambient temperature of the drying oven includes: using a temperature sensor installed inside or at the entrance of the drying oven to collect the initial temperature of the drying oven before it is started in real time, which is then used as the ambient temperature.

[0069] Specifically, the temperature sensor can be a type K thermocouple or a PT100 resistance temperature detector (RTD), with a measurement accuracy of no less than ±0.5℃. The sensor is installed inside the feed inlet of the drying oven, approximately 10cm from the inner wall of the furnace, to avoid interference from radiant heat. When the drying oven is in standby mode and heating is not activated, a stable reading from the sensor is taken, which is the ambient temperature. This step is completed automatically before each execution of the start-up control logic.

[0070] In some embodiments of the present invention, in order to achieve closed-loop monitoring of the startup process and avoid control failure caused by deviation between the actual heating rate and the predicted value, the method further includes:

[0071] In response to the start-up of the drying oven, the real-time temperature of the drying oven is continuously acquired at fixed time intervals;

[0072] The real-time temperature is input into the heating duration prediction model to update the predicted heating duration.

[0073] Determine whether the real-time temperature has reached the target temperature. If not, repeat the above steps until the target temperature is reached.

[0074] During the actual start-up and heating process of the drying oven, the actual heating rate may deviate from the model prediction due to environmental changes or equipment performance fluctuations. Therefore, after the drying oven starts, the system reads the current real-time temperature again via the temperature sensor at preset time intervals (e.g., every 5 minutes). Using the real-time temperature as new input, the heating time prediction model is called again to recalculate the remaining time required to rise from the current temperature to the target temperature (i.e., the updated predicted heating time). Then, it is determined whether the current real-time temperature has reached the target temperature. If not, the process continues to wait for the next time interval and repeats the above process; if the target temperature has been reached, monitoring ends. This dynamic refresh mechanism provides continuously updated forecast information for operators or upper-level systems, improving control robustness.

[0075] To balance prediction accuracy and system overhead, this embodiment provides a preferred time interval:

[0076] The fixed time interval is 5 minutes. In other embodiments, the time interval can also be set to 1 minute, 10 minutes or other values ​​according to the thermal inertia and process requirements of the drying oven, but 5 minutes is the optimal choice.

[0077] In some embodiments of the present invention, a specific method for calculating the optimal startup time is provided:

[0078] The step of calculating the optimal start-up time of the drying furnace based on the predicted heating time and the preset target process temperature arrival time includes: subtracting the predicted heating time from the target process temperature arrival time to obtain the optimal start-up time.

[0079] Assuming the predicted heating time is Δt (in minutes) obtained through model prediction, and the target process temperature arrival time given by the production scheduling system is T_target (e.g., accurate to the minute), then the optimal start time T_start = T_target – Δt. For example, if T_target = 08:00 and Δt = 25 minutes, then T_start = 07:35. This subtraction operation can be implemented internally by a system timer or real-time clock module. It is important to note that if the calculation result is not an integer number of minutes, it is usually rounded down to ensure the process temperature is reached on time.

[0080] In some embodiments of the present invention, to facilitate the use of this method by operators in production lines with low levels of automation, the calculation results can also be presented to personnel in an intuitive manner:

[0081] The heating time prediction model is compiled into an executable program, and the executable program is embedded into the manufacturing execution system;

[0082] The calculated optimal start-up time is output through the human-machine interface of the manufacturing execution system to prompt the operator to start the drying oven at the optimal start-up time.

[0083] Specifically, the following detailed explanation uses a specific factory intelligent transformation project as an example. In this project, the energy-saving optimization program is deployed on a designated server within the factory using a standardized interface and is centrally managed through the Enterprise Manufacturing Execution System (E-MES). The program files include a main control module, a model invocation module, a path definition module, and a log module. The main control module (e.g., main.txt) is responsible for the overall process control of the program: importing the asynchronous context manager, prediction functions, data processing tools, application programming interface module, tool module, and server module; defining health check endpoints to verify interface functionality; selecting different prediction functions based on the model type (classification or regression); setting the start and stop conditions for the log; and binding the lifecycle manager to the application. The model invocation module (e.g., xg.txt) is responsible for executing the specific prediction logic: importing machine learning libraries, defining regression model prediction functions, loading the trained model, and performing predictions. The path definition module (e.g., common.txt) stores the path information for each drying oven and its corresponding model. For example, it defines the storage path for the model files of two electrophoresis drying ovens, one adhesive drying oven, and two topcoat drying ovens, with the format "equipment identifier_process section type_model file path". If the number of prediction models needs to be expanded later, simply add path definitions to this file according to the specified format. The log module (e.g., log.txt) is responsible for setting the log format, specifying the log file path, creating the log directory (if it does not exist), and configuring basic log settings. Through the above modular design, this method can achieve pluggable and easily scalable deployment in the manufacturing execution system.

[0084] In some embodiments of the present invention, a specific model construction method is provided:

[0085] The prediction model for the heating duration is a regression model, which is obtained by fitting historical startup data using the least squares method.

[0086] Specifically, collect at least 20 sets of historical startup data (ambient temperature xi, heating time yi). Assume the model is a univariate linear regression equation: y = ax + b. According to the least squares formula:

[0087] a = (n∑xiyi - ∑xi∑yi) / (n∑xi² - (∑xi)²)

[0088] b = (∑yi - a∑xi) / n

[0089] The coefficients a and b are calculated to determine the regression equation. For non-linearly distributed data, a quadratic polynomial regression can be used: y = ax² + bx + c, and the coefficients are solved using the least squares method. After deployment, this regression model can calculate the predicted warming duration y each time the ambient temperature x is input.

[0090] In some embodiments of the present invention, in order to more intuitively illustrate the application effect of the method in actual production, a specific implementation case is described below.

[0091] In an intelligent transformation project of an automotive painting production line, the above method was deployed as a software program on the factory server and integrated with the E-MES system. After the program started running, as follows... Figure 4 As shown, in the energy management workbench interface of the E-MES system, the system calculates an independent preheating start-up time for each drying oven (including electrophoresis drying oven, adhesive drying oven, and topcoat drying oven) defined in the model. Once the preset optimal start-up time is reached, the workbench begins real-time monitoring of the equipment, dynamically displaying key parameters including "real-time temperature," "equipment start-up status," and "estimated heating time." The "estimated heating time" is not a fixed value but is dynamically calculated by calling the integrated heating time prediction model in the background, combined with real-time operating conditions. This value is automatically acquired and refreshed by the energy workbench every 5 minutes, providing continuously updated estimated information for production scheduling. This dynamic monitoring process continues until the actual equipment temperature reaches the preset target process temperature. Actual verification shows that using this method reduces the waiting time for a single start-up of this drying oven by 21.38 minutes and lowers the cost per unit by 1.26 yuan, achieving significant energy savings.

[0092] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. For those skilled in the art, other variations or modifications can be made based on the above description. It is impossible to exhaustively list all the implementation methods here. All obvious variations or modifications derived from the technical solutions of the present invention are still within the protection scope of the present invention.

Claims

1. A design method for reducing the start-up energy consumption of a drying oven, characterized in that, include: Obtain the ambient temperature of the drying oven; The ambient temperature is input into a pre-trained temperature rise duration prediction model to obtain the predicted temperature rise duration; Based on the predicted heating time and the preset target process temperature arrival time, the optimal start-up time of the drying furnace is calculated. The drying oven is started at the optimal start-up time.

2. The design method for reducing the start-up energy consumption of a drying oven according to claim 1, characterized in that, The process of establishing the heating duration prediction model is as follows: Acquire the historical startup data of the drying oven, which includes historical ambient temperature and its corresponding historical heating duration; Based on the historical startup data, a regression algorithm is used to fit the data and generate the heating duration prediction model.

3. The design method for reducing the start-up energy consumption of a drying oven according to claim 2, characterized in that, The step of generating the temperature rise duration prediction model based on the historical startup data and fitting it using a regression algorithm includes: A scatter plot is drawn with the historical ambient temperature as the horizontal axis and the historical temperature rise duration as the vertical axis. Based on the scatter plot, a regression curve is fitted, and the regression curve is used as the prediction model for the heating duration.

4. The design method for reducing the start-up energy consumption of a drying oven according to claim 1, characterized in that, The step of inputting the ambient temperature into a pre-trained temperature rise duration prediction model to obtain the predicted temperature rise duration includes: The ambient temperature is input into the heating time prediction model corresponding to each of the multiple process sections to obtain multiple predicted heating times. The calculation of the optimal start-up time of the drying oven based on the predicted heating time and the preset target process temperature arrival time includes: The maximum value among the multiple predicted heating times is used as the predicted heating time for calculating the optimal start-up time.

5. The design method for reducing the start-up energy consumption of a drying oven according to claim 1, characterized in that, The process of obtaining the ambient temperature of the drying oven includes: The initial temperature of the drying oven before startup is collected in real time by a temperature sensor installed inside or at the entrance of the drying oven, and is used as the ambient temperature.

6. The design method for reducing the start-up energy consumption of a drying oven according to claim 1, characterized in that, In response to the start-up of the drying oven, the real-time temperature of the drying oven is continuously acquired at fixed time intervals; The real-time temperature is input into the heating duration prediction model to update the predicted heating duration. Determine whether the real-time temperature has reached the target temperature. If not, repeat the above steps until the target temperature is reached.

7. The design method for reducing the start-up energy consumption of a drying oven according to claim 6, characterized in that, The fixed time interval is 5 minutes.

8. The design method for reducing the start-up energy consumption of a drying oven according to claim 1, characterized in that, The calculation of the optimal start-up time of the drying oven based on the predicted heating time and the preset target process temperature arrival time includes: The optimal start-up time is obtained by subtracting the predicted heating time from the time it takes for the target process temperature to be reached.

9. The design method for reducing the start-up energy consumption of a drying oven according to claim 1, characterized in that, The heating time prediction model is compiled into an executable program, and the executable program is embedded into the manufacturing execution system; The calculated optimal start-up time is output through the human-machine interface of the manufacturing execution system to prompt the operator to start the drying oven at the optimal start-up time.

10. The design method for reducing the start-up energy consumption of a drying oven according to claim 1, characterized in that, The prediction model for the heating duration is a regression model, which is obtained by fitting historical startup data using the least squares method.

Images