Methods and related equipment for infrared temperature data correction of thermal drainage UAVs

Abstract

This application provides a method and related equipment for correcting infrared temperature data from a thermal drainage unmanned aerial vehicle (UAV), belonging to the field of marine surveying and remote sensing data processing technology. The method includes: collecting infrared temperature distribution data of a UAV in a target thermal drainage area; sampling the UAV infrared temperature distribution data according to location points to obtain a first temperature value corresponding to each location point; determining the corresponding tidal and seasonal information based on the acquisition time of the UAV infrared temperature distribution data; constructing an input vector based on the location point, its first temperature value, and the tidal and seasonal information; inputting the input vector into a temperature correction model to correct the first temperature value, obtaining a second temperature value for the location point; and updating the UAV infrared temperature distribution data based on the second temperature value of each location point. This application can improve the measurement accuracy of temperature distribution data in thermal drainage areas.

Description

Technical Field

[0001] This application relates to the field of marine surveying and remote sensing data processing technology, and in particular to a method and related equipment for correcting infrared temperature data of a thermally dissipated unmanned aerial vehicle (UAV). Background Technology

[0002] Some factories (such as nuclear power plants) discharge warm wastewater with characteristics such as continuous discharge, high heat load, and strong hydrodynamic control over the diffusion process. The surface temperature field of the receiving water area exhibits significant nonlinearity and heterogeneity both spatially and temporally. High-precision and high-efficiency monitoring of the temperature diffusion range and intensity of coastal warm wastewater is one of the core requirements of environmental regulation. UAV infrared thermal imaging technology, with its high spatiotemporal resolution and maneuverability, has become an important means of remote sensing monitoring of warm wastewater. However, in practical applications, the retrieved surface water temperature can be affected by the sensor's own accuracy and environmental interference, leading to measurement deviations in the UAV infrared temperature data.

[0003] In related technologies, UAV infrared temperature correction methods typically rely on synchronous mobile survey data and employ globally uniform linear regression or simple empirical models for correction. These methods struggle to characterize the nonlinear variations of the thermal discharge temperature field and measurement errors across different spatiotemporal scales, and also lack generalization ability. If such correction models are applied to the verification of infrared temperature data from thermal discharge UAVs in other locations or marine environments, the correction accuracy is also low, resulting in low measurement accuracy of temperature distribution data in the thermal discharge area. Summary of the Invention

[0004] The main objective of this application is to propose an infrared temperature data correction method and related equipment for a thermal drainage UAV, aiming to improve the measurement accuracy of temperature distribution data in thermal drainage areas.

[0005] To achieve the above objectives, one aspect of this application proposes a method for correcting infrared temperature data of a thermal drainage unmanned aerial vehicle (UAV), comprising the following steps: Collect infrared temperature distribution data of the target thermal drainage area using drones; The infrared temperature distribution data of the UAV is sampled according to location points to obtain the first temperature value corresponding to each location point; Based on the acquisition time of the UAV infrared temperature distribution data, the corresponding tidal and seasonal information are determined; An input vector is constructed based on the location point, its first temperature value, the tidal information, and the seasonal information. The input vector is input into the temperature correction model to correct the first temperature value, thereby obtaining the second temperature value of the location point, and then the infrared temperature distribution data of the UAV is updated according to the second temperature value of each location point.

[0006] In some embodiments, the temperature correction model is trained through the following steps: Simultaneously collect UAV infrared temperature distribution data and mobile measured temperature data in the thermal drainage area; The infrared temperature distribution data of the UAV and the measured temperature data during mobile navigation are spatiotemporally matched according to the measured points during mobile navigation to obtain a first sample data set. Add corresponding tidal and seasonal information to each first sample data in the first sample data set to obtain a second sample data set of the warm drainage area, and then determine the training dataset based on the second sample data set. The machine learning-based temperature correction model is trained using the training dataset to obtain a well-trained temperature correction model.

[0007] In some embodiments, the step of performing spatiotemporal matching of the UAV infrared temperature distribution data and the measured temperature data during transit according to the measured points to obtain a first sample data set includes the following steps: The infrared temperature distribution data of the UAV and the measured temperature data during mobile navigation are spatiotemporally matched according to the measured points during mobile navigation to obtain a third sample data set. The abnormal sample data in the third sample data set is identified by the residual between the temperature value observed by the UAV infrared observation and the measured temperature value, and the abnormal sample data is removed to obtain the first sample data set.

[0008] In some embodiments, adding corresponding tidal and seasonal information to each first sample data in the first sample data set to obtain a second sample data set for the thermal drainage area includes the following steps: Based on the acquisition time of the UAV infrared temperature distribution data of the warm drainage area, tidal information and seasonal information are determined; The measured point location and UAV infrared observation temperature value of each first sample data in the first sample data set are normalized, and the tidal information and the seasonal information are individually encoded. The tidal information and the seasonal information after the individually encoded tidal information and the seasonal information are added to each first sample data to obtain the second sample data set.

[0009] In some embodiments, training the machine learning-based temperature correction model using the training dataset to obtain a trained temperature correction model includes the following steps: Initialize a temperature correction model based on a multilayer perceptron neural network; The hyperparameter combination of the temperature correction model is optimized based on the training dataset to obtain the optimized hyperparameter combination; the hyperparameter combination includes network structure and training control parameters; The structure of the temperature correction model is updated to the optimized network structure, and then the neuron parameters of the temperature correction model are trained according to the training dataset and the training control parameters to obtain the trained temperature correction model.

[0010] In some embodiments, optimizing the hyperparameter combination of the temperature correction model based on the training dataset to obtain the optimized hyperparameter combination includes the following steps: A network search is performed on the search space of each hyperparameter to obtain multiple combinations of hyperparameters; For each hyperparameter combination, the temperature correction model is trained based on the hyperparameter combination and the training dataset, and the score of the hyperparameter combination is obtained by cross-validation evaluation of the currently trained temperature correction model. The hyperparameter combination with the highest score is selected as the optimized hyperparameter combination.

[0011] To achieve the above objectives, another aspect of this application proposes an infrared temperature data correction system for a thermal drainage unmanned aerial vehicle, comprising: The first module is used to collect infrared temperature distribution data of the UAV in the target thermal drainage area; The second module is used to sample the infrared temperature distribution data of the UAV according to the location points to obtain the first temperature value corresponding to each location point; The third module is used to determine the corresponding tidal and seasonal information based on the acquisition time of the UAV infrared temperature distribution data; The fourth module is used to construct an input vector based on the location point, its first temperature value, the tidal information, and the seasonal information. The fifth module is used to input the input vector into the temperature correction model to correct the first temperature value, obtain the second temperature value of the location point, and then update the UAV infrared temperature distribution data according to the second temperature value of each location point.

[0012] To achieve the above objectives, another aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described method.

[0013] To achieve the above objectives, another aspect of the embodiments of this application proposes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0014] To achieve the above objectives, another aspect of the embodiments of this application proposes a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0015] The embodiments of this application include at least the following beneficial effects: This application provides a method, system, electronic device, storage medium, and program product for correcting infrared temperature data of a UAV used for warm drainage. This solution first collects infrared temperature distribution data of the UAV in the target warm drainage area. Then, it samples the UAV infrared temperature distribution data according to location points to obtain a first temperature value corresponding to each location point. Next, based on the acquisition time of the UAV infrared temperature distribution data, it determines the corresponding tidal and seasonal information. An input vector is constructed based on the location point, its first temperature value, and the tidal and seasonal information. This input vector is then input into a temperature correction model to correct the first temperature value, obtaining a second temperature value for the location point. Finally, the UAV infrared temperature distribution data is updated based on the second temperature value of each location point. When mapping and correcting the UAV infrared observation temperature value, the temperature correction model of this solution considers not only a single temperature value but also geographical location, tidal information, and seasonal information, achieving spatiotemporal nonlinear correction of the UAV infrared observation temperature value, improving correction accuracy, and thus improving the measurement accuracy of temperature distribution data in the warm drainage area. Attached Figure Description

[0016] Figure 1 This is a flowchart of the infrared temperature data correction method for a thermal drainage UAV provided in the embodiments of this application; Figure 2 This is a schematic diagram of the network architecture of the temperature correction model provided in the embodiments of this application; Figure 3 This is a scatter plot of the predicted and measured values ​​on the test set provided in the embodiments of this application; Figure 4 This is a comparison diagram of the correction effect before and after, using the example of a rapid summer tide provided in this application embodiment; Figure 5 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit it. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with those of this application; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this application as detailed in the appended claims.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0019] Nuclear power plant thermal wastewater is characterized by continuous discharge, high heat load, and strong hydrodynamic control over its diffusion process. The surface temperature field of the receiving water area exhibits significant nonlinearity and heterogeneity both spatially and temporally. High-precision and high-efficiency monitoring of the temperature diffusion range and intensity of thermal wastewater is one of the core requirements for environmental regulation of nuclear power plants. Unmanned aerial vehicle (UAV) infrared thermal imaging technology, with its high spatiotemporal resolution and maneuverability, has become an important means of remote sensing monitoring of coastal thermal wastewater. However, in practical applications, the retrieved surface temperature is not only affected by the accuracy of the sensor itself but is also easily interfered with by a combination of environmental factors such as atmospheric conditions, water surface state, tides, and seasonal changes. Under different geographical locations, tidal states, and seasonal backgrounds, the flow velocity structure, mixing intensity, and air-sea heat exchange conditions of the water surface exhibit systematic differences, leading to significant spatial correlation and environmental dependence in measurement bias. Related UAV infrared temperature correction methods typically rely on synchronous mobile field measurement data, employing globally uniform linear regression or simple empirical models for correction. On the one hand, these methods are difficult to characterize the nonlinear variation of the temperature field and measurement error of the warm drainage area at different spatiotemporal scales; on the other hand, the related technologies fail to consider that the accuracy of UAV infrared temperature measurement is also affected by factors such as tides and seasons, and fail to incorporate key environmental driving factors such as geographical location and tidal state as input variables into the modeling process. This results in insufficient adaptability and generalization ability of the model under different working conditions, limited correction accuracy and stability, and consequently, low measurement accuracy of temperature distribution data in the warm drainage area.

[0020] In view of this, this application provides a method and related equipment for correcting infrared temperature data of a thermal drainage UAV. When mapping and correcting the infrared observation temperature value of the UAV, the temperature correction model of this scheme considers not only a single temperature value, but also geographical location, tidal information and seasonal information, so as to realize the spatiotemporal nonlinear correction of the infrared observation temperature value of the UAV, improve the correction accuracy, and thus improve the measurement accuracy of temperature distribution data in the thermal drainage area.

[0021] The infrared temperature data correction method for thermal drainage UAVs provided in this application relates to the field of marine surveying and remote sensing data processing technology. This method can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, or vehicle-mounted terminal, but is not limited to these. The server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network. The software can be an application implementing the infrared temperature data correction method for thermal drainage UAVs, but is not limited to the above forms.

[0022] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0023] Figure 1 This is an optional flowchart of the infrared temperature data correction method for a thermal drainage UAV provided in the embodiments of this application. Figure 1 The method may include, but is not limited to, steps S101 to S105.

[0024] S101, collects infrared temperature distribution data of the UAV in the target thermal drainage area; S102, sample the infrared temperature distribution data of the UAV according to the location points to obtain the first temperature value corresponding to each location point; S103, determine the corresponding tidal and seasonal information based on the acquisition time of the UAV infrared temperature distribution data; S104, construct the input vector based on the location point, its first temperature value, tidal information, and seasonal information; S105, input the input vector into the temperature correction model to correct the first temperature value, obtain the second temperature value of the location point, and then update the UAV infrared temperature distribution data according to the second temperature value of each location point.

[0025] In step S101 of some embodiments, infrared temperature distribution data of the UAV to be corrected in the target thermal drainage area is acquired. The UAV infrared temperature distribution data can be represented as an infrared temperature distribution image of the UAV. After acquiring the UAV infrared temperature distribution data to be corrected, filtering and noise reduction processing can be performed on the UAV infrared temperature distribution data to improve the quality of the original data.

[0026] In step S102 of some embodiments, the UAV infrared temperature distribution data is sampled according to location points to obtain a first temperature value corresponding to each location point. For example, longitude and latitude can be sampled at certain intervals to construct a location matrix on the target thermal drainage area, where the rows of the location matrix represent the sampled longitude and the columns represent the sampled latitude. The corresponding UAV infrared observation temperature value (i.e., the first temperature value) is extracted from the UAV infrared temperature distribution data according to the elements of the location matrix to obtain the first temperature value corresponding to each location point. If the UAV infrared temperature distribution data is represented as a UAV infrared temperature distribution image, it can also be that each pixel in the image is traversed to extract the UAV infrared observation temperature value of that pixel. and latitude and longitude coordinates This yields the first temperature value corresponding to each location point.

[0027] In step S103 of some embodiments, tidal information is calculated and seasonal information is mapped based on the time when the UAV collects infrared temperature distribution data, to obtain the corresponding tidal information and seasonal information. The tidal information in this embodiment may include, but is not limited to, tidal states. Japanese style Tidal pattern refers to the basic classification of tidal phenomena based on the number of tidal rises and falls and the difference in tidal range within a lunar day (approximately 24 hours and 50 minutes), including but not limited to semi-diurnal tides, diurnal tides, and mixed tides. Tidal state refers to the specific stage or instantaneous state of the tide during a rise and fall cycle, including but not limited to high tide, low tide, slack tide, cessation of tide, rising tide, and ebb tide.

[0028] In step S104 of some embodiments, an input vector is constructed based on the location point, its first temperature value, tidal information, and seasonal information. Further, the location point and the first temperature value can be normalized, and the tidal information and seasonal information can be one-hot encoded. The processing uses the same preprocessing parameters as the training time. , (encoding dictionary).

[0029] In step S105 of some embodiments, the input vector is input into the temperature correction model to correct the first temperature value, thereby obtaining a high-precision temperature value for the location point. (i.e., the second temperature value), and then update the UAV infrared temperature distribution data based on the second temperature value at each location point. Specifically, temperature interpolation is performed between location points based on the second temperature value of each location point to obtain new UAV infrared temperature distribution data, which is then represented as a UAV infrared temperature distribution image. This image can clearly and accurately identify the high-temperature zone and temperature rise range of the target thermal drainage area, providing a direct basis for environmental thermal impact assessment. For example, a comparison of the effects before and after temperature correction is shown below. Figure 4 As shown, where, Figure 4 (a) in the image is the original UAV infrared temperature image during a high tide in summer; Figure 4 (b) in the image shows the corrected temperature image during the summer high tide.

[0030] In some embodiments, the temperature correction model in step S105 can be obtained by, but is not limited to, the following steps: S201, synchronously collects UAV infrared temperature distribution data and mobile measured temperature data in the warm drainage area; S202, perform spatiotemporal matching of UAV infrared temperature distribution data and mobile measured temperature data according to mobile measured points to obtain the first sample data set; S203, add corresponding tidal and seasonal information to each first sample data in the first sample data set to obtain the second sample data set of the warm drainage area, and then determine the training dataset based on the second sample data set; S204. Train the machine learning-based temperature correction model using the training dataset to obtain the trained temperature correction model.

[0031] In step S201 of some embodiments, after acquiring the UAV infrared temperature distribution data for training, the UAV infrared temperature data for training can also be filtered and denoised to improve the quality of the original data.

[0032] For example, when collecting data, observations are conducted under weather conditions in the survey area that are cloudless, free of haze, with high visibility, and stable atmospheric conditions, where atmospheric visibility is no less than 30 km and sea surface wind speed is less than 4 m / s. In typical waters where nuclear power plant thermal discharge diffuses (e.g., within a few kilometers downstream of the discharge outlet), a UAV flight path and a survey vessel are planned for simultaneous measurement. The UAV, equipped with an infrared thermal imager, conducts aerial surveys of the target area, acquiring infrared temperature images covering the entire survey area. Simultaneously, the survey vessel, equipped with a temperature, salinity, depth, and turbidity meter, navigates along a pre-defined cross-section, simultaneously collecting measured surface temperature data at several discrete points. The data collection process is optimized for the diffusion characteristics of nuclear power plant thermal discharge. The measurement route is laid out along the main diffusion direction downstream of the nuclear power plant discharge outlet, its lateral temperature gradient region, and background waters far from the discharge outlet, ensuring that the collected sample points cover the high-temperature zone, transition zone, and background zone of the thermal discharge, thereby enhancing the representativeness of the samples in the temperature gradient change region.

[0033] Typical scenarios for monitoring thermal discharge from nuclear power plants were selected, covering key hydrodynamic stages such as spring tides, neap tides, and ebb and flow tides, as well as aerial and mobile UAV measurements conducted in summer and winter respectively. This ensured that the training dataset could encompass various typical thermal discharge diffusion conditions, preventing the temperature correction model from depending solely on a single tidal condition and improving the applicability of the constructed temperature correction model in practical regulatory applications. After acquiring the UAV infrared temperature distribution data, median filtering was further applied to the data, replacing the center pixel value with the median value within the pixel's neighborhood to suppress salt-and-pepper noise in the infrared image.

[0034] In step S202 of some embodiments, after synchronously collecting UAV infrared temperature distribution data and mobile measured temperature data, the corresponding temperature value is extracted from the UAV infrared temperature distribution data according to the mobile measured point, and the temperature value is matched with the temperature measured at the mobile measured point to form first sample data. The first sample data includes the geographical location of the mobile measured point, the UAV infrared observed temperature value, and the mobile measured temperature value. The first sample data of all mobile measured points form a first sample data set. The UAV infrared observed temperature value and geographical location in the sample are used as input data in subsequent training, and the mobile measured temperature value is used as label data in subsequent training.

[0035] In step S203 of some embodiments, tidal information representing instantaneous hydrodynamic conditions of the tidal current, tidal pattern information representing tidal energy levels, and seasonal information representing background heat exchange conditions are added to each first sample data in the first sample data set at the corresponding time, resulting in a second sample data set for the warm discharge area. This information is used as input data in subsequent training. In this embodiment, instead of simply superimposing multiple environmental parameters, a correlation analysis of UAV infrared measurement errors is performed on all candidate influencing factors based on the formation process of infrared temperature measurement errors of the nuclear power plant's warm discharge diffuser and UAVs. Several closely related key driving factors are then selected from all candidate influencing factors. It is understood that the key driving factors selected in this way may include, but are not limited to, tidal information, tidal pattern information, and seasonal information. Furthermore, these key driving factors are introduced as independent features into the temperature correction model input, enabling the temperature correction model to automatically learn the influence weights of different driving factors on infrared temperature measurement errors and their nonlinear relationships during training. This achieves adaptive temperature correction for different spatial locations, different tidal processes, and different seasonal background conditions, avoiding systematic bias caused by using uniform correction parameters. The training dataset may include at least one second sample data set for a warm discharge area.

[0036] In step S204 of some embodiments, the temperature correction model can learn from relevant machine learning models, specifically, it can employ a neural network model. More specifically, the temperature correction model can employ a multilayer perceptron neural network model. When training the temperature correction model, grid search, random search, or Bayesian optimization methods can be used to optimize the model's hyperparameters to obtain optimal performance.

[0037] In this embodiment, environmental information highly correlated with warm wastewater diffusion and infrared thermometry errors, such as geographical location, tidal state, tidal type, and season, is incorporated into the temperature correction model. This allows the model to learn the systematic variation patterns of temperature measurement errors under different spatial locations and environmental conditions, thereby characterizing the complex nonlinear relationships and spatial heterogeneity of the warm wastewater temperature field. This enables adaptive and high-precision correction for warm wastewater monitoring scenarios. Based on the corrected temperature data, a high-precision spatial distribution map of warm wastewater temperature can be generated, intuitively and quantitatively reflecting the range and intensity of warm wastewater diffusion, providing clear and reliable decision support information for environmental supervision.

[0038] In some embodiments, step S202 may include, but is not limited to, the following steps: S301, the infrared temperature distribution data of the UAV and the measured temperature data during the mobile flight are spatiotemporally matched according to the measured points during the mobile flight to obtain a third sample data set; S302, based on the residual between the infrared temperature value observed by the UAV and the measured temperature value, identify abnormal sample data in the third sample data set, and remove abnormal sample data to obtain the first sample data set.

[0039] In step S301 of some embodiments, for each mobile measurement point, based on its precise timestamp and latitude / longitude coordinates, within a time threshold (less than 10 minutes), the corresponding geographical location data is retrieved from the UAV infrared temperature distribution data. Thus, the mobile measurement temperature (i.e., the true temperature) of each mobile measurement point is determined. ) and the corresponding UAV infrared observation temperature value Matching is performed to obtain third sample data, and the third sample data corresponding to all the mobile survey points are combined to form a third sample dataset.

[0040] In step S302 of some embodiments, outlier detection and removal from cloud shadows, ship wakes, or sensor noise in the third sample dataset is performed to improve the reliability of the training data. The outlier removal method may involve calculating the residual between the UAV infrared observation value and the measured temperature value during navigation in each third sample dataset; determining an anomaly threshold based on the residual distribution of all third sample datasets; and removing third sample datasets whose absolute residual value is greater than the anomaly threshold. The anomaly threshold can be determined based on the ratio between the average and standard deviation of the residuals. The specific implementation steps are as follows: Calculate the residual: for the third sample data set... Based on a sample of data, calculate the residual between the UAV infrared temperature observation value and the mobile measured temperature value. .

[0041] Determine the anomaly detection threshold: Calculate the average of the residuals of all sample data. and standard deviation In this embodiment, it can be based on statistical methods. The criteria set the anomaly detection threshold as follows: That is, the residual is considered to fall within [ , Samples outside the specified range are outliers.

[0042] Remove outlier samples: Iterate through all sample data in the third sample dataset and remove all samples that meet the criteria. The remaining sample data constitutes a more reliable cleaned sample set, namely the first sample data set, which is used for subsequent model training.

[0043] In some embodiments, step S203, which involves adding corresponding tidal and seasonal information to each first sample data in the first sample data set to obtain a second sample data set for the thermal drainage area, may include, but is not limited to, the following steps: S401, based on the acquisition time of UAV infrared temperature distribution data in the warm drainage area, determine tidal and seasonal information; S402, normalize the measured point location and UAV infrared observation temperature value of each first sample data in the first sample data set, and perform unique thermal encoding on the tidal information and seasonal information. Add the tidal information and seasonal information after unique thermal encoding to each first sample data to obtain the second sample data set.

[0044] In this embodiment, the first sample data set includes multiple first sample data sets after removing outliers, and each first sample data set includes a mobile survey point (i.e., a geographical location). The data includes the corresponding UAV infrared observation temperature value and the mobile measured temperature value, and then each first sample data is compared with the tidal state determined according to the sampling time. Trendy Seasonal Information Pairing is performed to form initial samples. All successfully matched point pairs constitute the initial sample set.

[0045] To meet the requirements of model training, eliminate dimensional differences, and accelerate model convergence, the input features in the initial samples can be preprocessed to obtain a second sample dataset. Preprocessing includes normalizing numerical features (such as temperature values ​​and coordinates) and performing one-hot encoding on categorical features (such as tide patterns and seasonal classifications).

[0046] Numerical feature normalization: for , , These three continuous numerical features are linearly transformed to the [0,1] interval using the min-max normalization method. The transformation formula is as follows: ,in and These are the minimum and maximum values ​​of the feature in the training dataset, respectively.

[0047] One-hot encoding of categorical features: For the three categorical features of tide state, tide type, and season, one-hot encoding is used to convert them into binary vectors. For example, if there are 4 types of tide state, they are encoded as [1,0,0,0], [0,1,0,0], [0,0,1,0], and [0,0,0,1].

[0048] It should be noted that each training sample in the training dataset consists of an input feature vector X and an output label y, as follows: The input feature vector X must include at least the following dimensions: UAV infrared temperature readings: .

[0049] Geographic location information This is used to characterize the spatial diffusion characteristics of thermal wastewater from nuclear power plants in receiving water bodies. Because the temperature gradient, mixing mechanism, and infrared radiation characteristics of thermal wastewater from nuclear power plants differ significantly between the near-field and far-field regions, relying solely on temperature values ​​is insufficient to accurately reflect actual water temperature changes. Introducing geographic location information can effectively characterize spatial heterogeneity.

[0050] Tidal information (rapid rise, slow rise, rapid fall, slow fall): reflects changes in water velocity and direction over short timescales, and has a decisive influence on the stretching, diffusion, and backflow of warm-discharge plumes. Under different tidal conditions, the infrared thermometry error at the same location exhibits significant differences. Incorporating tidal information helps distinguish the variation patterns of thermometry errors under different tidal phases, improving the model's adaptability to rapidly changing tidal conditions.

[0051] Tidal pattern information (spring tide, mid-tide, neap tide): reflects long-term tidal energy levels and background hydrodynamic conditions, affecting the overall dilution capacity of warm water discharge and the structure of the background temperature field, and is an important factor causing trans-diurnal and trans-period temperature deviations. By introducing tidal pattern information, the correction model can adaptively adjust the correction relationship for different tidal energy conditions, avoiding systematic deviations caused by tidal pattern changes.

[0052] Seasonal information (winter, summer): Seasonal changes systematically alter solar radiation, atmospheric stability, and air-sea heat exchange fluxes, thus affecting the infrared radiation retrieval process. Using seasonal information as model input helps improve the model's generalization ability when applied across seasons.

[0053] Output label y: the corresponding measured temperature value during the underway survey. .

[0054] Furthermore, all samples are randomly shuffled and divided into a training set (training dataset) and an independent test set in an 8:2 ratio. The test set is completely isolated throughout the model training and tuning process and is used only for the final evaluation of the model's generalization performance.

[0055] In some embodiments, step S204 may include, but is not limited to, the following steps: S501, Initialize the temperature correction model based on a multilayer perceptron neural network; S502, optimize the hyperparameter combination of the temperature correction model based on the training dataset to obtain the optimized hyperparameter combination; the hyperparameter combination includes the network structure and training control parameters; S503 updates the structure of the temperature correction model to the optimized network structure, and then trains the neuron parameters of the temperature correction model based on the training dataset and training control parameters to obtain the trained temperature correction model.

[0056] In step S501 of some embodiments, please refer to Figure 2 The temperature correction model's network architecture consists of an input layer, two hidden layers, and an output layer. The number of nodes in the input layer equals the total dimension of the preprocessed features. The two hidden layers have 100 and 50 neurons respectively, both using the ReLU activation function. The output layer has one neuron, using a linear activation function, to output the predicted corrected temperature value. This temperature correction model employing a multilayer perceptron neural network can more accurately extract the spatiotemporal nonlinear features from the input, thus more accurately calibrating the temperature value.

[0057] In some embodiments, step S502 may include, but is not limited to, the following steps: S601 performs a network search on the search space of each hyperparameter to obtain multiple hyperparameter combinations; S602, for each hyperparameter combination, the temperature correction model is trained based on the hyperparameter combination and the training dataset, and the score of the hyperparameter combination is obtained by cross-validation of the currently trained temperature correction model. S603 selects the hyperparameter combination with the highest score as the hyperparameter combination for optimization.

[0058] In this embodiment, to obtain optimal performance and avoid overfitting, grid search and cross-validation are used to optimize the model hyperparameters. The hyperparameter search space is defined as follows: network structure [(100, 50), (200, 100)]; L2 regularization coefficient [0.001, 0.01]; learning rate [0.001, 0.01]; maximum training epochs [1000, 2000]. During training, 5-fold cross-validation can be used for evaluation, with mean squared error as the main evaluation metric. The optimizer is Adam. An early stopping mechanism is also enabled, automatically terminating training when the validation loss no longer improves within a certain number of consecutive epochs to prevent overfitting. An exhaustive search strategy is used to select the hyperparameter combination that performs optimally on the cross-validation set to train the final model.

[0059] For example, through the above optimization process, the final model hyperparameter combination that achieves the best correction performance can be determined as follows: network structure (100, 50), L2 regularization coefficient 0.001, learning rate 0.01, and maximum number of iterations 1000. Retraining on the complete training set with this parameter combination yields the final temperature correction model.

[0060] In some embodiments, the final model is evaluated using an independent test set that was not involved in the training. The mean absolute error and coefficient of determination R² between the predicted and actual temperatures are calculated. For example, see [link to relevant documentation]. Figure 3 After correction using the temperature correction model of this embodiment, the MAE is 0.152°C and the R² is 0.998, which is significantly better than the traditional global linear regression method (MAE is 0.937°C and R² is 0.947). The above results show that the method of this embodiment improves the correction accuracy (MAE) by approximately 84% and significantly enhances the model fit (R²) compared to the traditional linear regression method. To comprehensively analyze the performance characteristics of the model under different environmental conditions, all sample data were grouped and statistically analyzed according to tide type, tide state, and season, and the performance indicators of each group are shown in Table 1.

[0061] Table 1 Performance indicators of the temperature correction model across all samples, grouped by different tide types, tidal states, and seasons.

[0062] This application embodiment also provides an infrared temperature data correction system for a thermal drainage UAV, including: The first module is used to collect infrared temperature distribution data of the UAV in the target thermal drainage area; The second module is used to sample the infrared temperature distribution data of the UAV according to the location points to obtain the first temperature value corresponding to each location point; The third module is used to determine the corresponding tidal and seasonal information based on the acquisition time of the UAV infrared temperature distribution data; The fourth module is used to construct an input vector based on the location point, its first temperature value, tidal information, and seasonal information; The fifth module is used to input the input vector into the temperature correction model to correct the first temperature value, obtain the second temperature value of the location point, and then update the UAV infrared temperature distribution data according to the second temperature value of each location point.

[0063] It is understood that the methods described in the above method embodiments are applicable to this system embodiment. The specific functions implemented in this system embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0064] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described method. This electronic device can be any smart terminal, including tablet computers and computers.

[0065] It is understood that the content of the above method embodiments is applicable to this device embodiment. The specific functions implemented by this device embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0066] Please see Figure 5 , Figure 5 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes: The processor 901 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 902 can be implemented as a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 902 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 902 and is called and executed by the processor 901. The input / output interface 903 is used to implement information input and output; The communication interface 904 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 905 transmits information between various components of the device (e.g., processor 901, memory 902, input / output interface 903, and communication interface 904); The processor 901, memory 902, input / output interface 903, and communication interface 904 are connected to each other within the device via bus 905.

[0067] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0068] It is understood that the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0069] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0070] It is understood that the content of the above method embodiments is applicable to the embodiments of this program product. The specific functions implemented by the embodiments of this program product are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0071] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0072] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0073] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0074] The system embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0075] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0076] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or modules is not necessarily limited to those steps or modules explicitly listed, but may include other steps or modules not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0077] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0078] In the embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0079] The modules described above as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0080] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0081] If the integrated module is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0082] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for correcting infrared temperature data of a thermal drainage UAV, characterized in that, Includes the following steps: Collect infrared temperature distribution data of the target thermal drainage area using drones; The infrared temperature distribution data of the UAV is sampled according to location points to obtain the first temperature value corresponding to each location point; Based on the acquisition time of the UAV infrared temperature distribution data, the corresponding tidal and seasonal information are determined; An input vector is constructed based on the location point, its first temperature value, the tidal information, and the seasonal information. The input vector is input into the temperature correction model to correct the first temperature value, thereby obtaining the second temperature value of the location point, and then the infrared temperature distribution data of the UAV is updated according to the second temperature value of each location point.

2. The infrared temperature data correction method for a thermal drainage UAV according to claim 1, characterized in that, The temperature correction model is trained through the following steps: Simultaneously collect UAV infrared temperature distribution data and mobile measured temperature data in the thermal drainage area; The infrared temperature distribution data of the UAV and the measured temperature data during mobile navigation are spatiotemporally matched according to the measured points during mobile navigation to obtain a first sample data set. Add corresponding tidal and seasonal information to each first sample data in the first sample data set to obtain a second sample data set of the warm drainage area, and then determine the training dataset based on the second sample data set. The machine learning-based temperature correction model is trained using the training dataset to obtain a well-trained temperature correction model.

3. The infrared temperature data correction method for a thermal drainage UAV according to claim 2, characterized in that, The step of performing spatiotemporal matching between the UAV infrared temperature distribution data and the mobile measured temperature data according to the mobile measured points to obtain the first sample data set includes the following steps: The infrared temperature distribution data of the UAV and the measured temperature data during mobile navigation are spatiotemporally matched according to the measured points during mobile navigation to obtain a third sample data set. The abnormal sample data in the third sample data set is identified by the residual between the temperature value observed by the UAV infrared observation and the measured temperature value, and the abnormal sample data is removed to obtain the first sample data set.

4. The infrared temperature data correction method for a thermal drainage UAV according to claim 2, characterized in that, The step of adding corresponding tidal and seasonal information to each first sample data in the first sample data set to obtain a second sample data set for the warm drainage area includes the following steps: Based on the acquisition time of the UAV infrared temperature distribution data of the warm drainage area, tidal information and seasonal information are determined; The measured point location and UAV infrared observation temperature value of each first sample data in the first sample data set are normalized, and the tidal information and the seasonal information are individually encoded. The tidal information and the seasonal information after the individually encoded tidal information and the seasonal information are added to each first sample data to obtain the second sample data set.

5. The infrared temperature data correction method for a thermal drainage UAV according to claim 2, characterized in that, The step of training the machine learning-based temperature correction model using the training dataset to obtain the trained temperature correction model includes the following steps: Initialize a temperature correction model based on a multilayer perceptron neural network; The hyperparameter combination of the temperature correction model is optimized based on the training dataset to obtain the optimized hyperparameter combination; the hyperparameter combination includes network structure and training control parameters; The structure of the temperature correction model is updated to the optimized network structure, and then the neuron parameters of the temperature correction model are trained according to the training dataset and the training control parameters to obtain the trained temperature correction model.

6. The infrared temperature data correction method for a thermal drainage UAV according to claim 5, characterized in that, The optimization of the hyperparameter combination of the temperature correction model based on the training dataset to obtain the optimized hyperparameter combination includes the following steps: A network search is performed on the search space of each hyperparameter to obtain multiple combinations of hyperparameters; For each hyperparameter combination, the temperature correction model is trained based on the hyperparameter combination and the training dataset, and the score of the hyperparameter combination is obtained by cross-validation evaluation of the currently trained temperature correction model. The hyperparameter combination with the highest score is selected as the optimized hyperparameter combination.

7. A thermal drainage UAV infrared temperature data correction system, characterized in that, include: The first module is used to collect infrared temperature distribution data of the UAV in the target thermal drainage area; The second module is used to sample the infrared temperature distribution data of the UAV according to the location points to obtain the first temperature value corresponding to each location point; The third module is used to determine the corresponding tidal and seasonal information based on the acquisition time of the UAV infrared temperature distribution data; The fourth module is used to construct an input vector based on the location point, its first temperature value, the tidal information, and the seasonal information. The fifth module is used to input the input vector into the temperature correction model to correct the first temperature value, obtain the second temperature value of the location point, and then update the UAV infrared temperature distribution data according to the second temperature value of each location point.

8. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method according to any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 6.

Images