An asphalt panel temperature calibration model training and calibration method, system and device

Abstract

The present application provides a kind of asphalt panel temperature calibration model training method, first training sample set is used to obtain first calibration model and second calibration model respectively by training, then the first calibration model and second calibration model obtained by training are cascaded to obtain initial temperature calibration model, finally, the initial temperature calibration model is trained to obtain temperature calibration model. The temperature calibration model obtained by training in the way of hierarchical training is more accurate, and is more suitable for the temperature calibration of impermeable panel. Meanwhile, the present application also provides a method for temperature calibration using the trained temperature calibration model. Through this temperature calibration method, different collection devices, different collection environments, panels with different compositions and structures can be calibrated for different depths of temperature to obtain the temperature distribution of the longitudinal depth of the panel, improve the efficiency of temperature measurement, and ensure the construction quality and safety of asphalt concrete panel.

Description

Technical Field

[0001] This invention belongs to the field of water conservancy and hydropower engineering construction technology, specifically relating to a method, system and device for training and calibrating an asphalt panel temperature calibration model. Background Technology

[0002] Temperature control is a core control aspect in the construction of asphalt concrete anti-seepage panels for water conservancy and hydropower projects, directly affecting the panel construction quality and the stability of its anti-seepage function. Excessive temperature can easily cause shrinkage and cracking of the panel later on, while insufficient temperature will lead to insufficient aggregate bonding density; both significantly reduce the anti-seepage effect and shorten its service life. Therefore, accurate temperature measurement during key construction processes such as asphalt concrete anti-seepage panel paving and compaction is a necessary prerequisite for ensuring project quality. Currently, the mainstream temperature measurement methods are mainly divided into two categories: (1) Manual insertion measurement: The measurement is completed by manually inserting the thermometer into the panel to a specified depth. Its core advantage is that it can directly obtain temperature data at a specified depth inside the panel, and the measurement accuracy is relatively high; however, it has significant shortcomings, such as the inability to achieve real-time and continuous measurement, low measurement efficiency, difficulty in covering the entire area of ​​the panel, and easy to miss local areas with abnormal temperatures; at the same time, the ambient temperature at the construction site is usually between 120℃ and 180℃, and manual operation poses a very high risk of burns.

[0003] (2) IoT Infrared Sensing Measurement: With the development of Internet of Things (IoT) technology, some projects have attempted to install infrared temperature sensors at a certain height above the panel on paving and compaction machinery. The sensors collect infrared radiation signals from the panel surface in real time, realizing non-contact real-time automated measurement. This method can cover the entire area of ​​the panel, greatly improving the efficiency of temperature measurement; however, the technology has significant limitations. It can only obtain the surface temperature of the panel and cannot obtain the temperature distribution of the panel in the vertical direction. Moreover, the obtained surface temperature is inaccurate and insufficient to truly reflect the temperature status of key control points in the panel construction quality. Ultimately, this restricts the large-scale promotion and application of this technology in the construction of asphalt concrete anti-seepage panels.

[0004] Existing temperature measurement technologies for asphalt concrete anti-seepage panels all have core technical pain points: one type is difficult to adapt to the needs of efficient construction due to insufficient real-time performance and continuity and high safety risks; the other type is difficult to support precise quality control due to insufficient measurement accuracy and inability to obtain the temperature distribution of the panel in the vertical direction. Neither can meet the actual needs of projects for comprehensive and high-precision monitoring of construction temperature.

[0005] In view of this, the present invention is hereby proposed. Summary of the Invention

[0006] To address the aforementioned technical problems in the prior art, this invention provides a method, system, and apparatus for training and calibrating an asphalt panel temperature calibration model, which solves the technical problem that "the prior art cannot efficiently and accurately calibrate the temperature of asphalt concrete anti-seepage panels at different vertical depths."

[0007] To achieve the above objectives, the technical solution of the present invention is as follows: A method for training an asphalt panel temperature calibration model includes: Constructing the initial first calibration model and the initial second calibration model, specifically including: A first training sample set is obtained, wherein each first training sample in the first training sample set includes a first parameter matrix, a first measured temperature, and a first true temperature; the first parameter matrix indicates information related to temperature acquisition. Obtain a second training sample set, wherein each second training sample in the second training sample set includes a second parameter matrix and a second true temperature; the second parameter matrix indicates information related to temperature calibration. The first calibration model is trained using the first parameter matrix, the first measured temperature, and the first actual temperature to obtain the first calibration model; The second parameter matrix, the second true temperature, and the first true temperature are used to train an initial second calibration model to obtain the second calibration model. The first calibration model and the second calibration model are cascaded; A temperature calibration model is obtained by training a cascaded first calibration model and a second calibration model using the first parameter matrix, the second parameter matrix, the first measured temperature, and the second actual temperature.

[0008] Furthermore, the first parameter matrix includes first acquisition device parameters and first environmental parameters; the first acquisition device parameters include at least device height, and the first environmental parameters include at least ambient temperature; the second parameter matrix includes at least the aggregate type, aggregate ratio, asphalt content, mineral powder content, and target calibration depth of the panel.

[0009] Further, the step of training an initial first calibration model using the first parameter matrix, the first measured temperature, and the first real temperature to obtain the first calibration model specifically includes: inputting the first parameter matrix and the first measured temperature into the initial first calibration model to obtain a first model output; determining whether to stop model training based on the distance between the first model output and the first real temperature; stopping model training when the distance between the first model output and the first real temperature is less than a preset value, thereby obtaining the first calibration model; Further, the step of training an initial second calibration model using the second parameter matrix, the second true temperature, and the first true temperature to obtain a second calibration model specifically includes: inputting the second parameter matrix and the first true temperature into the initial second calibration model to obtain a second model output; determining whether to stop model training based on the distance between the second model output and the second true temperature; stopping model training when the distance between the second model output and the second true temperature is less than a preset value, thereby obtaining the second calibration model.

[0010] Furthermore, the first calibration model and the second calibration model are cascaded; the cascaded first calibration model and the second calibration model are trained using the first parameter matrix, the second parameter matrix, the first measured temperature, and the second actual temperature to obtain a temperature calibration model, specifically including: The first calibration model and the second calibration model are cascaded to obtain the initial temperature calibration model; The first parameter matrix, the second parameter matrix, and the first measured temperature are input into the initial temperature calibration model to obtain the third model output. Based on the distance between the third model output and the second real temperature, it is determined whether to stop model training. When the distance between the third model output and the second real temperature is less than a preset value, model training is stopped, thereby obtaining the temperature calibration model.

[0011] Furthermore, this invention also proposes a method for calibrating an asphalt panel temperature calibration model. The method utilizes a temperature model constructed using the aforementioned asphalt panel temperature calibration model training method to calibrate the temperature of the asphalt panel, including: The system acquires a third parameter matrix, a fourth parameter matrix, and a second measured temperature; the third parameter matrix indicates information related to temperature acquisition, and the fourth parameter matrix indicates information related to temperature calibration. By combining the third parameter matrix, the fourth parameter matrix, and the second measured temperature and input temperature calibration model, the temperature distribution at different depths of the panel can be obtained.

[0012] Furthermore, the method also includes: acquiring a digital model of the panel; and visualizing the temperature field in the digital model according to the temperature distribution.

[0013] Furthermore, the present invention also proposes a temperature calibration model training system for waterproof panels, the system comprising: The building module is used to build the initial first calibration model and the initial second calibration model; The first acquisition module is used to acquire a first training sample set, wherein each training sample in the first training sample set includes a first parameter matrix, a first measured temperature, and a first real temperature; the first parameter matrix indicates information related to temperature acquisition. The second acquisition module is used to acquire a second training sample set, wherein each training sample in the second training sample set includes a second parameter matrix and a second true temperature; the second parameter matrix indicates information related to temperature calibration. The first training module is used to train an initial first calibration model using the first parameter matrix, the first measured temperature, and the first real temperature to obtain the first calibration model. The second training module is used to train an initial second calibration model using the second parameter matrix, the second true temperature, and the first true temperature to obtain the second calibration model. A cascade module is used to cascade the first calibration model and the second calibration model; The third training module is used to obtain the first calibration model and the second calibration model that are cascaded together with the first parameter matrix, the second parameter matrix, the first measured temperature and the second real temperature to obtain the temperature calibration model.

[0014] Furthermore, the present invention also proposes an apparatus for calibrating the temperature of an asphalt panel using the temperature calibration model described above, the apparatus comprising: The acquisition module is used to acquire a third parameter matrix, a fourth parameter matrix, and a second measured temperature; the third parameter matrix indicates information related to temperature acquisition, and the fourth parameter matrix indicates information related to temperature calibration. The calibration module is used to calibrate the third parameter matrix, the fourth parameter matrix, and the second measured temperature and input temperature calibration model to obtain the temperature distribution at different depths of the panel; The visualization module is used to acquire a digital model of the panel and, based on the temperature distribution, to visualize the temperature field within the digital model.

[0015] Furthermore, the present invention also proposes an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method described above.

[0016] Furthermore, the present invention also proposes a computer-readable storage medium for storing computer instructions, which, when executed by a processor, implement the method described above.

[0017] Furthermore, the present invention also proposes a computer program product comprising computer instructions that, when executed by a processor, implement the method described above.

[0018] Compared with the prior art, the beneficial effects of the present invention include: 1. The temperature calibration model trained by this invention for calibrating the construction temperature of asphalt concrete anti-seepage panels consists of a first calibration model and a second calibration model cascaded together. Through this model, the temperature distribution at each depth of the anti-seepage panel can be calibrated with high precision using only the temperature measured by the acquisition equipment. It can realize real-time automated measurement of the entire area of ​​the asphalt concrete anti-seepage panel, with a wider temperature measurement coverage and higher efficiency. At the same time, the entire construction process is non-contact temperature measurement, and the workers do not need to come into close contact with the high-temperature asphalt panel, which greatly reduces the risk of burns caused by manual temperature measurement and reduces the incidence of safety accidents.

[0019] 2. The temperature calibration model trained by this invention for calibrating the construction temperature of asphalt concrete anti-seepage panels has a third parameter matrix and a fourth parameter matrix in its input parameters. The third parameter matrix indicates information related to temperature acquisition, and the fourth parameter matrix indicates information related to temperature calibration. Therefore, this temperature calibration model can be adapted to different acquisition devices, different acquisition environments, and panels with different compositions and structures, and has high application value.

[0020] 3. This invention uses a hierarchical training method, first training the first calibration model, then training the second calibration model, and finally training the cascaded first and second calibration models, so that the temperature calibration model obtained by training is more accurate and more suitable for the temperature calibration of waterproof panels. Attached Figure Description

[0021] Figure 1 A schematic diagram of the temperature calibration model training method provided in an embodiment of the present invention; Figure 2 A framework diagram of the temperature calibration model training method provided in this embodiment of the invention; Figure 3 A schematic diagram of the method for calibrating the construction temperature of asphalt concrete anti-seepage panels provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of a temperature calibration model training device provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of a device for calibrating the construction temperature of asphalt concrete anti-seepage panels provided in an embodiment of the present invention. Detailed Implementation

[0022] The technical solution of the present invention will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are not all embodiments of the present invention. All other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present invention.

[0023] It should be noted that, unless otherwise specifically stated, the relative arrangement and numerical expressions of the components and steps described in these embodiments should not be construed as limiting the scope of the invention.

[0024] The following description of exemplary embodiments is merely illustrative and is not intended to limit the invention or its application or use in any way. Techniques, methods, and apparatus known to those skilled in the art may not be discussed in detail herein, but where applicable, such techniques, methods, and apparatus should be considered part of this specification.

[0025] See Figure 1 and Figure 2 , Figure 1 This invention proposes a training method for an asphalt panel temperature calibration model. Figure 2 This is a framework diagram of the temperature calibration model training method provided in an embodiment of the present invention. Through hierarchical training, the trained temperature calibration model can accurately obtain the temperature components at different depths of the waterproof panel based on the third and fourth parameter matrices. Specifically, it includes the following steps: S1: Construct the initial first calibration model and the initial second calibration model. The initial first calibration model and the initial second calibration model can use machine learning models such as multiple regression, logistic regression, and decision trees, or they can use neural network models or deep learning models. The initial first calibration model and the initial second calibration model can use the same model or different models.

[0026] In one embodiment, the initial first calibration model adopts a logistic regression model, and the initial second calibration model adopts a large language model constructed from deep learning and neural networks.

[0027] S2: Obtain the first training sample set. Each first training sample in the first training sample set includes a first parameter matrix, a first measured temperature, and a first actual temperature. The first parameter matrix indicates information related to temperature acquisition, specifically including first acquisition device parameters and first environmental parameters. The acquisition device consists of an infrared temperature sensor installed under the paving / compacting machinery. It quickly obtains the surface measured temperature of the seepage-proof panel through scanning and uses it as the first measured temperature. The parameters of the first data acquisition device should include at least the device height. The first environmental parameter includes at least the ambient temperature. Equipment height The vertical distance between the infrared temperature sensor and the surface of the waterproof panel is measured and recorded using a laser rangefinder or meter stick during installation; this value is fixed. Ambient temperature... The ambient temperature at the time of data collection was acquired in real time by a high-precision environmental sensor installed at the construction site and uploaded to the system platform. First Real Temperature The temperature calibrated on the surface of the waterproof panel by a standard contact thermometer.

[0028] S3: Obtain the second training sample set. Each second training sample in the second training sample set includes a second parameter matrix and a second true temperature. The second parameter matrix indicates information related to temperature calibration and includes at least the aggregate type of the panel. Aggregate proportion Asphalt content Mineral powder content and target calibration depth Second true temperature Temperatures calibrated at different depths inside the waterproof panel using a standard contact thermometer, at each calibration depth Corresponding to a second real temperature It is understandable that the second training sample set contains training data from different panels at different depths, as well as training data from the same panel at different depths.

[0029] S4: Train an initial first calibration model using the first parameter matrix, the first measured temperature, and the first real temperature to obtain the first calibration model. Specifically, train the initial first calibration model using the first parameter matrix, the first measured temperature, and the first real temperature corresponding to each training sample in the first training sample set to obtain the first calibration model. This involves inputting the first parameter matrix and the first measured temperature into the initial first calibration model to obtain the first model output; determining whether to stop model training based on the distance between the first model output and the first real temperature. When the distance between the first model output and the first real temperature is less than a preset value, stop model training to obtain the first calibration model.

[0030] According to the blackbody radiation law:

[0031] in, For radiative exitance, The Stepan-Boltzmann constant is... Absolute temperature; In one embodiment, the initial first calibration model uses a logistic regression (LR) model to measure the first temperature. Equipment height Ambient temperature Inputting the data into the logistic regression model yields the first model output. ,Right now: ; Calculate the output of the first model With the first true temperature The distance between them, Distance1, when When the time is reached, model training is stopped, thus obtaining the first calibrated model. The distance Distance1 is calculated using methods such as normative distance and similarity distance.

[0032] S5: Train the initial second calibration model using the second parameter matrix, the second true temperature, and the first true temperature to obtain the second calibration model. Specifically, input the second parameter matrix and the first true temperature into the initial second calibration model to obtain the second model output; determine whether to stop model training based on the distance between the second model output and the second true temperature. When the distance between the second model output and the second true temperature is less than a preset value, stop model training to obtain the second calibration model.

[0033] In one embodiment, the initial second calibration model adopts a large language model (LLM), and the second parameter matrix is... , , , , and first true temperature Input into a large language model, and obtain the output of the second model. : ; Calculate the output of the first model With the second true temperature The distance between them, Distance2, when At this point, model training is stopped, thus obtaining the first calibrated model. The distance Distance2 is calculated using methods such as normative distance and similarity distance. , , , and These represent aggregate type, aggregate ratio, asphalt content, mineral powder content, and target calibration depth, respectively.

[0034] S6: Cascade the first and second calibration models. By cascading the first and second calibration models, an initial temperature calibration model is obtained. Cascading the first and second calibration models results in the output of the first calibration model being the first model output. The first actual temperature is used instead of the input temperature of the second calibration model; that is, the first measured temperature is used instead of the input temperature of the first calibration model. Equipment height Ambient temperature The input for the second calibration model is , , , and ,as well as .

[0035] S7: The first and second calibration models are trained using the first parameter matrix, the second parameter matrix, the first measured temperature, and the second actual temperature to obtain the temperature calibration model. Specifically, the first parameter matrix, the second parameter matrix, and the first measured temperature are input into the initial temperature calibration model to obtain the output of the third model; based on the distance between the output of the third model and the second actual temperature, it is determined whether to stop model training. When the distance between the output of the third model and the second actual temperature is less than a preset value, model training is stopped, thereby obtaining the temperature calibration model.

[0036] In one embodiment, the first calibration model is the logistic regression model (LR) trained in step S4, and the second calibration model is the large language model (LLM) trained in step S5. The first and second calibration models are cascaded, with the output of the first calibration model (LR) used as an input to the second calibration model (LLM). The cascaded first and second calibration models are trained using the first parameter matrix, the second parameter matrix, the first measured temperature, and the second actual temperature to obtain the temperature calibration model, specifically: The first measured temperature Equipment height Ambient temperature The input is fed into the trained logistic regression model (i.e., the first calibration model) to obtain the output of the third model. ,Right now: ; The second parameter matrix , , , , The output of the fourth model is fed into the trained large language model (i.e., the second calibration model) to obtain the output of the third model. : ; Calculate the output of the third model With the second true temperature Distance3 between them, when When the time is reached, model training is stopped, thus obtaining the temperature calibration model. Distance3 is calculated using methods such as paradigm distance and similarity distance.

[0037] This invention employs a tiered training approach, using the same training sample set. First, a first calibration model is trained, then a second calibration model is trained, and finally, the cascaded first and second calibration models are trained to obtain the final temperature calibration model. This results in a more accurate temperature calibration model that is more suitable for the temperature calibration of waterproof panels.

[0038] Meanwhile, this invention uses training samples from different depths of the same panel to train a temperature calibration model, which allows the temperature calibration model to calibrate the temperature at different depths on the same panel, thereby obtaining the temperature distribution of the panel along its longitudinal depth. Furthermore, this invention uses training samples from different depths of different panels to train a temperature calibration model, which allows the temperature calibration model to be adapted to different panels, different acquisition devices, different acquisition environments, and panels with different compositions and structures, thus having high application value.

[0039] See Figure 3 After obtaining the temperature calibration model through the above method, this invention also proposes a calibration method for the temperature calibration model of asphalt panels. The temperature model constructed using the aforementioned asphalt panel temperature calibration model training method is used to calibrate the temperature of the asphalt panel, specifically including: S21: Acquire the third parameter matrix, the fourth parameter matrix, and the second measured temperature; the third parameter matrix indicates information related to temperature acquisition, and the fourth parameter matrix indicates information related to temperature calibration.

[0040] The system collects the third parameter matrix, the fourth parameter matrix, and the second measured temperature from the actual application scenario. The third parameter matrix includes the parameters of the second acquisition device and the second environmental parameters. The parameters of the second acquisition device include at least the device height. The second environmental parameter includes at least the ambient temperature. Second temperature measurement The panel surface temperature is collected by the acquisition device. The fourth parameter matrix includes the aggregate type of the panel. Aggregate proportion Asphalt content Mineral powder content and target calibration depth .

[0041] S22: Input the third parameter matrix, the fourth parameter matrix, and the second measured temperature into the temperature calibration model to obtain the temperature distribution at different depths of the panel. The third parameter matrix... , The fourth parameter matrix , , , and and the second measured temperature Input the temperature calibration model to calibrate the temperature distribution at different depths on the calibration panel.

[0042] After calibrating the temperature distribution at different depths on the panel, this invention also proposes to visualize the temperature distribution, specifically as follows: S23: Obtain the digital model of the panel. Construct a digital model of the panel based on its components and structure.

[0043] S24: Based on the temperature distribution, visualize the temperature field in the digital model. Use gradient colors to display the temperature distribution in the digital model according to the temperature magnitude.

[0044] By displaying the temperature distribution in a digital model, users can quickly understand the temperature distribution inside and on the surface of the panel, further enhancing its application value.

[0045] See Figure 4 , Figure 4 This invention proposes a training system for an asphalt panel temperature calibration model, the system comprising: The building module is used to build the initial first calibration model and the initial second calibration model; The first acquisition module is used to acquire a first training sample set. Each training sample in the first training sample set includes a first parameter matrix, a first measured temperature, and a first true temperature. The first parameter matrix indicates information related to temperature acquisition. The first parameter matrix includes first acquisition device parameters and first environmental parameters. The first acquisition device parameters include at least the device height, and the first environmental parameters include at least the ambient temperature. The first measured temperature is the surface temperature of the waterproof panel acquired by the acquisition device, and the first true temperature is the temperature calibrated on the surface of the waterproof panel by a standard contact thermometer.

[0046] The second acquisition module is used to acquire a second training sample set. Each training sample in the second training sample set includes a second parameter matrix and a second true temperature. The second parameter matrix indicates information related to temperature calibration. The second parameter matrix includes at least the aggregate type, aggregate ratio, asphalt content, mineral powder content, and target calibration depth of the panel. The second true temperature is the temperature calibrated by a standard contact thermometer at different depths inside the impermeable panel.

[0047] The first training module is used to train an initial first calibration model using a first parameter matrix, a first measured temperature, and a first real temperature to obtain the first calibration model. Specifically, the first parameter matrix and the first measured temperature are input into the initial first calibration model to obtain the first model output. Based on the distance between the first model output and the first real temperature, it is determined whether to stop model training. When the distance between the first model output and the first real temperature is less than a preset value, model training is stopped, thereby obtaining the first calibration model.

[0048] The second training module is used to train an initial second calibration model using a second parameter matrix, a second true temperature, and a first true temperature to obtain a second calibration model. Specifically, the second parameter matrix and the first true temperature are input into the initial second calibration model to obtain the output of the second model. Based on the distance between the output of the second model and the second true temperature, it is determined whether to stop model training. When the distance between the output of the second model and the second true temperature is less than a preset value, model training is stopped, thereby obtaining the second calibration model.

[0049] The cascading module is used to cascade the first calibration model and the second calibration model; specifically, it uses the output of the first calibration model as an input to the second calibration model.

[0050] The third training module is used to obtain the first parameter matrix, the second parameter matrix, the first measured temperature, and the second real temperature to train the first calibration model and the second calibration model in a cascade, thereby obtaining the temperature calibration model. Specifically, the first parameter matrix, the second parameter matrix, and the first measured temperature are input into the initial temperature calibration model to obtain the output of the third model. Based on the distance between the output of the third model and the second real temperature, it is determined whether to stop the model training. When the distance between the output of the third model and the second real temperature is less than a preset value, the model training is stopped, thereby obtaining the temperature calibration model.

[0051] See Figure 5 After obtaining the temperature calibration model, this invention also proposes an apparatus for applying the calibration method of the trained asphalt panel temperature calibration model, the apparatus comprising: The data acquisition module is used to acquire a third parameter matrix, a fourth parameter matrix, and a second measured temperature. The third parameter matrix indicates information related to temperature acquisition in the actual application scenario, and the fourth parameter matrix indicates information related to temperature calibration in the actual application scenario. The third parameter matrix includes second acquisition device parameters and second environmental parameters. The second acquisition device parameters include at least the device height, and the second environmental parameters include at least the ambient temperature. The second measured temperature is the panel surface temperature acquired by the acquisition device. The fourth parameter matrix includes the panel's aggregate type, aggregate ratio, asphalt content, mineral powder content, and target calibration depth.

[0052] The calibration module is used to calibrate the model using the third parameter matrix, the fourth parameter matrix, the second measured temperature, and the input temperature to obtain the temperature distribution at different depths of the panel. The third parameter matrix, the fourth parameter matrix, and the second measured temperature are input into the temperature calibration model to calibrate the temperature distribution of the panel at different depths.

[0053] The visualization module is used to acquire a digital model of the panel and visualize the temperature field within that model based on the temperature distribution. The digital model of the panel is constructed based on its components and structure; the temperature distribution is then displayed in the digital model using gradient colors according to the temperature magnitude.

[0054] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., high-density digital video discs (DVDs)), or semiconductor media (e.g., solid-state drives (SSDs)).

[0055] The memory in this embodiment of the invention can be volatile memory or non-volatile memory, or may include both. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which serves as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous linked dynamic random access memory (SLDRAM), and direct rambus RAM (DRRAM). It should be noted that the memory used in the methods described in this invention is intended to include, but is not limited to, these and any other suitable types of memory.

[0056] In implementation, each step of the above method can be completed by integrated logic circuits in the processor's hardware or by instructions in software. The steps of the method disclosed in the embodiments of this application can be directly implemented by a hardware processor, or by a combination of hardware and software modules in the processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method. To avoid repetition, detailed descriptions are omitted here.

[0057] It should be noted that the processor in the embodiments of this application can be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method embodiments can be completed by the integrated logic circuitry in the processor's hardware or by instructions in software form. The processor can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied as being executed by a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory, and the processor reads the information in the memory and, in conjunction with its hardware, completes the steps of the above methods.

[0058] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the foregoing application concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions claimed in this application.

Claims

1. A method for training an asphalt panel temperature calibration model, characterized in that, The method includes: Constructing the initial first calibration model and the initial second calibration model, specifically including: A first training sample set is obtained, wherein each first training sample in the first training sample set includes a first parameter matrix, a first measured temperature, and a first true temperature; the first parameter matrix indicates information related to temperature acquisition; the first measured temperature is the surface measured temperature of the impermeable panel; the first true temperature is the temperature calibrated on the surface of the impermeable panel by a standard contact thermometer; the first parameter matrix includes first acquisition device parameters and first environmental parameters; the first acquisition device parameters include at least the device height, and the first environmental parameters include at least the ambient temperature. A second training sample set is obtained, and each second training sample in the second training sample set includes a second parameter matrix and a second true temperature; the second parameter matrix indicates information related to temperature calibration; the second parameter matrix includes at least the aggregate type, aggregate ratio, asphalt content, mineral powder content and target calibration depth of the panel; the second true temperature is the temperature calibrated by a standard contact thermometer at different depths inside the seepage-proof panel; The first calibration model is trained using the first parameter matrix, the first measured temperature, and the first actual temperature to obtain the first calibration model; The second parameter matrix, the second true temperature, and the first true temperature are used to train an initial second calibration model to obtain the second calibration model. The first calibration model and the second calibration model are cascaded; A temperature calibration model is obtained by training a cascaded first calibration model and a second calibration model using the first parameter matrix, the second parameter matrix, the first measured temperature, and the second actual temperature.

2. The asphalt panel temperature calibration model training method according to claim 1, characterized in that, The step of training an initial first calibration model using the first parameter matrix, the first measured temperature, and the first real temperature to obtain a first calibration model specifically includes: inputting the first parameter matrix and the first measured temperature into the initial first calibration model to obtain a first model output; determining whether to stop model training based on the distance between the first model output and the first real temperature; stopping model training when the distance between the first model output and the first real temperature is less than a preset value, thereby obtaining the first calibration model. The step of training an initial second calibration model using the second parameter matrix, the second real temperature, and the first real temperature to obtain a second calibration model specifically includes: inputting the second parameter matrix and the first real temperature into the initial second calibration model to obtain a second model output; determining whether to stop model training based on the distance between the second model output and the second real temperature; stopping model training when the distance between the second model output and the second real temperature is less than a preset value, thereby obtaining the second calibration model.

3. The asphalt panel temperature calibration model training method according to claim 1, comprising cascading the first calibration model and the second calibration model; training the cascaded first calibration model and the second calibration model using the first parameter matrix, the second parameter matrix, the first measured temperature, and the second actual temperature to obtain the temperature calibration model, specifically including: The first calibration model and the second calibration model are cascaded to obtain the initial temperature calibration model; The first parameter matrix, the second parameter matrix, and the first measured temperature are input into the initial temperature calibration model to obtain the third model output. Based on the distance between the third model output and the second real temperature, it is determined whether to stop model training. When the distance between the third model output and the second real temperature is less than a preset value, model training is stopped, thereby obtaining the temperature calibration model.

4. A method for calibrating the temperature of an asphalt panel using the temperature calibration model according to any one of claims 1-3, characterized in that, The method includes: The system acquires a third parameter matrix, a fourth parameter matrix, and a second measured temperature; the third parameter matrix indicates information related to temperature acquisition, and the fourth parameter matrix indicates information related to temperature calibration. The third parameter matrix, the fourth parameter matrix, and the second measured temperature are input into the temperature calibration model to obtain the temperature distribution at different depths of the panel.

5. The calibration method for the asphalt panel temperature calibration model according to claim 4, characterized in that: The method further includes: Obtain the digital model of the panel; Based on the temperature distribution, the temperature field is visualized in the digital model.

6. A training system for an asphalt panel temperature calibration model, characterized in that, The system includes: The building module is used to build the initial first calibration model and the initial second calibration model; The first acquisition module is used to acquire a first training sample set, wherein each training sample in the first training sample set includes a first parameter matrix, a first measured temperature, and a first true temperature; the first parameter matrix indicates information related to temperature acquisition; the first measured temperature is the surface measured temperature of the impermeable panel acquired by the acquisition device; and the first true temperature is the temperature calibrated on the surface of the impermeable panel by a standard contact thermometer. The first parameter matrix includes first acquisition device parameters and first environmental parameters; the first acquisition device parameters include at least device height, and the first environmental parameters include at least ambient temperature; The second acquisition module is used to acquire a second training sample set, wherein each training sample in the second training sample set includes a second parameter matrix and a second true temperature; the second parameter matrix indicates information related to temperature calibration; the second parameter matrix includes at least the aggregate type, aggregate ratio, asphalt content, mineral powder content and target calibration depth of the panel; the second true temperature is the temperature calibrated by a standard contact thermometer at different depths inside the seepage-proof panel; The first training module is used to train an initial first calibration model using the first parameter matrix, the first measured temperature, and the first real temperature to obtain the first calibration model. The second training module is used to train an initial second calibration model using the second parameter matrix, the second true temperature, and the first true temperature to obtain the second calibration model. A cascade module is used to cascade the first calibration model and the second calibration model; The third training module is used to train a cascaded first calibration model and a second calibration model using the first parameter matrix, the second parameter matrix, the first measured temperature, and the second real temperature to obtain a temperature calibration model.

7. A device for calibrating the construction temperature of asphalt panels using the temperature calibration model according to any one of claims 1-3, characterized in that, The device includes: The acquisition module is used to acquire a third parameter matrix, a fourth parameter matrix, and a second measured temperature; the third parameter matrix indicates information related to temperature acquisition, and the fourth parameter matrix indicates information related to temperature calibration. The calibration module is used to input the third parameter matrix, the fourth parameter matrix and the second measured temperature into the temperature calibration model to obtain the temperature distribution at different depths of the panel; The visualization module is used to acquire a digital model of the panel and, based on the temperature distribution, to visualize the temperature field within the digital model.

Images