Methane emission monitoring method and device, storage medium and electronic equipment
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING INSIGHTS VALUE TECHNOLOGY CO LTD
- Filing Date
- 2026-02-04
- Publication Date
- 2026-06-16
Smart Images
Figure CN122217880A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of satellite remote sensing technology, and in particular to a method, device, storage medium and electronic equipment for monitoring methane emissions. Background Technology
[0002] Methane is a highly influential anthropogenic greenhouse gas in the Earth's climate system, with a greenhouse effect potential significantly higher than carbon dioxide. It also promotes ground-based ozone formation, exacerbating air pollution and threatening human health. With the rapid development of satellite remote sensing technology, the spatial resolution, spectral resolution, and data coverage of satellite observations have been greatly improved, providing technical support for large-scale, high-resolution monitoring of methane emissions. Currently, the mainstream technical approach for methane emission monitoring using satellites is the "on-board acquisition, ground processing" model. The specific process is as follows: when the satellite passes overhead, it images the target area, acquiring raw spectral image data. This data is then transmitted to a ground receiving station via satellite downlink. The ground receiving station preprocesses the raw data and then uses techniques such as atmospheric radiative transfer mechanism models, data assimilation models, or neural network models to identify methane plumes and quantify emissions, ultimately generating methane emission monitoring data for enterprises and providing a basis for emission regulation.
[0003] While existing satellite methane monitoring technologies can achieve emission monitoring within a certain range, they suffer from a core technical problem: severely insufficient monitoring timeliness, failing to meet the real-time monitoring needs of key enterprises. Due to the massive amount of raw data collected by satellites, the ground reception and transmission processes are time-consuming, and the ground inversion process requires combining multiple meteorological data points and relies on complex physical models and algorithms. The entire data processing workflow often takes several days to complete. This lag means that monitoring results cannot reflect the real-time emission status of enterprises in a timely manner, hindering rapid response to sudden methane leaks and failing to provide regulatory authorities with timely and accurate enforcement evidence, severely restricting the efficient implementation of methane emission control measures. Summary of the Invention
[0004] In view of this, this application provides a method, apparatus, storage medium and electronic device for methane emission monitoring, which can realize real-time monitoring of methane emissions.
[0005] According to a first aspect of this application, a method for monitoring methane emissions is provided, the method being performed on a satellite equipped with an imaging spectrometer and an onboard computing platform, comprising:
[0006] During its operation in orbit, the satellite uses the imaging spectrometer to perform imaging operations on a preset target area of the enterprise to acquire methane spectral data; The on-board computing platform preprocesses the methane spectral data and uses a pre-built lightweight neural network model to make a preliminary prediction of methane emissions based on the preprocessed methane spectral data and geographical background information. The pre-stored correction coefficients are used to correct the initially predicted methane emissions, thus obtaining the methane emission monitoring results for the preset enterprise target area; The methane emission monitoring results are packaged and transmitted to the ground receiving station via satellite data transmission channel.
[0007] According to a second aspect of this application, a methane emission monitoring device is provided, the device being operated on a satellite carrying an imaging spectrometer and an on-board computing platform, comprising: The imaging module is used to perform imaging operations on a preset target area of an enterprise and acquire methane spectral data during the satellite's on-orbit operation using the imaging spectrometer. The prediction module is used to preprocess the methane spectral data through the on-board computing platform and use a pre-set lightweight neural network model to make a preliminary prediction of methane emissions based on the preprocessed methane spectral data and geographical background information. The correction module is used to call pre-stored correction coefficients to correct the preliminary predicted methane emissions and obtain the methane emission monitoring results for the preset enterprise target area. The downlink module is used to package the methane emission monitoring results into data and transmit them to the ground receiving station via the satellite data transmission channel.
[0008] According to a third aspect of this application, a storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the above-described methane emission monitoring method.
[0009] According to a fourth aspect of this application, an electronic device is provided, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, wherein the processor executes the program to implement the above-described methane emission monitoring method.
[0010] By employing the aforementioned technical solution, the methane emission monitoring method, device, storage medium, and electronic equipment provided in this application directly acquire methane spectral data by imaging a preset key enterprise target area using an imaging spectrometer during satellite operation. Subsequently, the methane spectral data is preprocessed through an onboard intelligent computing platform, and a pre-built lightweight neural network model combined with geographical background information is used to preliminarily predict methane emissions. This avoids the lengthy time required for ground-based reception of massive amounts of data and complex physical model inversion in traditional technologies. Furthermore, the lightweight model, through techniques such as depthwise separable convolution, adapts to the limited computing power onboard, enabling rapid inference and significantly shortening the data processing cycle. By calling pre-stored correction coefficients to correct the preliminary prediction results, the systematic bias of the on-board model can be effectively corrected, ensuring that real-time monitoring has both high accuracy and long-term stability. Finally, only the methane emission monitoring results are packaged and transmitted through the satellite data transmission channel, eliminating the need to transmit massive amounts of raw spectral data. This significantly reduces the amount of data transmission and bandwidth requirements, further compressing the overall process time. Ultimately, the processing delay of traditional satellite monitoring, which takes several days, is reduced to the hour level. This not only reflects the real-time emission status of enterprises in a timely manner, but also provides strong support for rapid response to sudden methane leaks and precise enforcement by regulatory authorities, effectively promoting the implementation of methane emission control measures.
[0011] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description
[0012] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A schematic flowchart of a methane emission monitoring method provided in an embodiment of this application is shown; Figure 2 A schematic flowchart of a methane emission monitoring method according to another embodiment of this application is shown; Figure 3 A schematic diagram of a methane emission monitoring device provided in an embodiment of this application is shown; Figure 4 A schematic diagram of a methane emission monitoring device provided in another embodiment of this application is shown. Detailed Implementation
[0013] The present application will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the embodiments of the present application can be combined with each other.
[0014] While existing satellite methane monitoring technologies can achieve emission monitoring within a certain range, they suffer from a core technical problem: severely insufficient monitoring timeliness, failing to meet the real-time monitoring needs of key enterprises. Due to the massive amount of raw data collected by satellites, the ground reception and transmission processes are time-consuming, and the ground inversion process requires combining multiple meteorological data points and relies on complex physical models and algorithms. The entire data processing workflow often takes several days to complete. This lag means that monitoring results cannot reflect the real-time emission status of enterprises in a timely manner, hindering rapid response to sudden methane leaks and failing to provide regulatory authorities with timely and accurate enforcement evidence, severely restricting the efficient implementation of methane emission control measures.
[0015] To address the aforementioned technical problems, embodiments of the present invention provide a method for monitoring methane emissions, such as... Figure 1 As shown, the method includes: Step 110: During the satellite's on-orbit operation, the imaging spectrometer is used to perform imaging operations on the preset target area of the enterprise to obtain methane spectral data.
[0016] Among them, "satellite" refers to a spacecraft carrying relevant monitoring equipment, operating in a predetermined orbit, and performing methane emission monitoring tasks; "on-orbit operation period" refers to the time period during which the satellite, after entering the preset orbit, carries out normal work according to the established procedures; "imaging spectrometer" refers to a specialized device that combines imaging and spectral detection functions and can capture spectral information of specific bands in the target area; "preset enterprise target area" refers to the relevant areas of enterprises that are pre-planned and determined and require key methane emission monitoring; "imaging operation" refers to the specific action of the imaging spectrometer to acquire spectral images of the target area; and "methane spectral data" refers to spectral data containing methane characteristic absorption spectral line information that can reflect the state of methane and related characteristics in the target area.
[0017] In the embodiments of this disclosure, when the satellite is operating normally in its predetermined orbit, it can use a suitable spectral detection device to carry out spectral imaging operations on the pre-planned key enterprise monitoring area. By sensing the optical signals and decomposing the spectra of the area through the spectral detection device, characteristic spectral information related to methane is captured, and finally spectral data that can characterize the methane-related properties of the area is obtained, providing basic data support for subsequent methane emission analysis.
[0018] This technical approach centers on real-time on-orbit satellite operations, using dedicated spectral detection equipment to directly collect data from key monitoring areas. This approach not only eliminates reliance on ground-based data transmission, ensuring timely data acquisition, but also accurately captures the characteristic spectral information of methane, ensuring the relevance and effectiveness of the data.
[0019] Step 120: Preprocess the methane spectral data using the on-board computing platform, and use a pre-built lightweight neural network model to make a preliminary prediction of methane emissions based on the preprocessed methane spectral data and geographical background information.
[0020] Among them, the on-board computing platform refers to the dedicated computing unit deployed on the satellite for performing tasks such as data processing and model calculation; preprocessing refers to the preliminary processing operations performed on the raw methane spectral data to optimize data quality for subsequent calculations; the pre-configured lightweight neural network model refers to the neural network model pre-configured in the on-board computing platform, which reduces resource consumption through simplified structural design; geographic background information refers to geographic information related to the monitoring target area, which can help characterize the regional environmental characteristics; and the preliminary prediction of methane emissions refers to the initial estimate of the amount of methane emissions in the target area based on the processed data and related models.
[0021] In this embodiment of the present disclosure, after the onboard computing platform of the satellite receives methane spectral data, it can first perform necessary preliminary optimization processing on the data, and then call the pre-configured lightweight neural network model to fuse the processed methane spectral data with the corresponding geographical background information. Through the calculation and reasoning of the model, the initial estimation of methane emissions in the target area is completed.
[0022] This technology enables data preprocessing and model inference to be completed directly on the onboard computing platform, avoiding the delays caused by ground processing in traditional technologies. The lightweight neural network model can be adapted to the limited computing resources on the satellite, ensuring the efficient operation of the methane emission monitoring mission. At the same time, the integration of geographical background information can provide more environmental references for emission estimation, effectively improving the accuracy and reliability of the preliminary prediction results.
[0023] Step 130: Call the pre-stored correction coefficients to correct the preliminary predicted methane emissions and obtain the methane emission monitoring results for the preset enterprise target area.
[0024] Among them, the pre-stored correction coefficients refer to the dedicated coefficients stored in advance in the on-board computing platform to correct the deviation of the prediction results; correction refers to the operation of using the correction coefficients to correct and adjust the preliminary predicted methane emissions in order to reduce the deviation; the methane emission monitoring results refer to the quantitative results that reflect the actual methane emissions in the target area after correction processing.
[0025] In this embodiment of the disclosure, the pre-stored correction coefficients can be retrieved from a designated storage location on the on-board computing platform. These correction coefficients are then combined with the initial estimated methane emissions obtained through data processing and model calculations. By using reasonable correction and adjustment methods, the preliminary prediction results are optimized and improved, and finally, the quantitative results of methane emissions for the pre-set enterprise monitoring area are obtained.
[0026] By calling pre-stored correction coefficients to specifically correct the preliminary methane emission forecast, systematic biases that may exist in the initial forecasting process can be effectively offset, significantly improving the accuracy and reliability of methane emission monitoring results.
[0027] Step 140: Package the methane emission monitoring results into data and transmit them to the ground receiving station via satellite data transmission channel.
[0028] Among them, data packaging refers to the operation of organizing and integrating monitoring results and related information to form data units that conform to transmission standards; satellite data transmission channel refers to a dedicated communication channel on a satellite used to transmit data to the ground; and ground receiving station refers to a dedicated facility deployed on the ground to receive data transmitted by the satellite through the data transmission channel.
[0029] In this embodiment of the disclosure, the methane emission monitoring results obtained after correction can be standardized, organized and integrated to form a data unit adapted to satellite transmission requirements. Then, the data unit can be stably transmitted to a ground facility dedicated to receiving satellite data through a dedicated data transmission channel carried by the satellite, thus completing the transmission of monitoring data from satellite to ground.
[0030] This technical step ensures the standardization and integrity of transmitted data by packaging the monitoring results. The data is then transmitted via a dedicated satellite data transmission channel, guaranteeing the stability and efficiency of data transmission. Furthermore, by transmitting only the processed final monitoring results, the amount of data transmitted can be significantly reduced, the bandwidth requirements can be lowered, and the data transmission speed can be accelerated. This enables the ground to obtain accurate monitoring data on methane emissions in the target area in a timely manner, providing timely and reliable data support for subsequent regulatory decisions.
[0031] In summary, the methane emission monitoring method provided in this application directly acquires methane spectral data by imaging a pre-defined target area of key enterprises using an imaging spectrometer during satellite operation. Subsequently, the onboard intelligent computing platform preprocesses the methane spectral data and uses a pre-built lightweight neural network model combined with geographical background information to preliminarily predict methane emissions. This avoids the lengthy time required for ground-based reception of massive amounts of data and complex physical model inversion in traditional technologies. Furthermore, the lightweight model, through techniques such as depthwise separable convolution, adapts to the limited computing power onboard, enabling rapid inference and significantly shortening the data processing cycle. Finally, by calling pre-stored correction systems… Correcting the preliminary prediction results with data pairs can effectively correct systematic biases in the on-board model, ensuring that real-time monitoring is both highly accurate and stable over the long term. Finally, only the methane emission monitoring results are packaged and transmitted via the satellite data transmission channel, eliminating the need to transmit massive amounts of raw spectral data. This significantly reduces data transmission volume and bandwidth requirements, further compressing the overall process time. Ultimately, the processing delay of traditional satellite monitoring, which takes several days, is reduced to the hour level. This not only reflects the real-time emission status of enterprises in a timely manner but also provides strong support for rapid response to sudden methane leaks and precise enforcement by regulatory authorities, effectively promoting the implementation of methane emission control measures.
[0032] Furthermore, as a refinement and extension of the specific implementation methods of the above embodiments, and to fully illustrate the implementation methods of this embodiment, this embodiment also provides another method for monitoring methane emissions, such as... Figure 2 As shown, the method includes: Step 210: During the satellite's on-orbit operation, the imaging spectrometer is used to perform imaging operations on the preset target area of the enterprise to obtain methane spectral data.
[0033] For the specific implementation process of the embodiments disclosed herein, please refer to the relevant description in step 110 of the embodiments, which will not be repeated here.
[0034] Step 220: The methane spectral data is denoised and radiometrically calibrated using the data preprocessing unit of the onboard computing platform.
[0035] Among them, the data preprocessing unit is a functional module in the on-board intelligent computing platform that is specifically responsible for optimizing the raw acquired data in the early stage; noise reduction processing is the operation of suppressing or removing irrelevant interference signals in methane spectral data to optimize data purity; and radiometric calibration processing is the operation of calibrating and correcting the radiometric amount of methane spectral data to ensure that the data can truly reflect the actual radiometric characteristics of the target area.
[0036] After receiving methane spectral data from the imaging spectrometer, the preprocessing unit of the onboard intelligent computing platform can sequentially suppress irrelevant interference signals in the methane spectral data and then standardize and calibrate the radiance of the methane spectral data. Through these two consecutive optimization operations, invalid interference components in the methane spectral data can be removed, and the radiative characteristic parameters in the methane spectral data can be standardized to form high-quality processed data that meets the requirements of subsequent model inference operations.
[0037] The data preprocessing unit performs targeted noise reduction and radiometric calibration, which effectively removes invalid information such as environmental interference and equipment noise from methane spectral data. At the same time, it calibrates the radiometric deviation of methane spectral data, ensuring that the processed methane spectral data has high purity and authenticity. This provides reliable data support for subsequent lightweight neural network models to integrate geographical background information for emission prediction.
[0038] Step 230: Retrieve the geographic background information corresponding to the preset enterprise target area from the storage module of the on-board computing platform.
[0039] The on-board computing platform's storage module refers to a dedicated component within the platform specifically used for storing various types of data, information, and related configurations. Geographic background information refers to geographic and environmental information associated with the preset target area of the enterprise. This information includes at least the location coordinates, terrain data, and real-time wind field vector data of the target area. Location coordinates are coordinate data that characterizes the spatial location of the target area; terrain data are data reflecting the terrain features of the target area; and real-time wind field vector data are vector data that characterizes the real-time wind direction and intensity of the target area.
[0040] In this embodiment of the disclosure, when the on-board computing platform performs methane emission-related processing, it can extract geographical background information corresponding to the preset enterprise target area from its own storage module. This information at least covers the spatial location coordinates, terrain feature data, and real-time wind direction and intensity data of the area, providing necessary environmental reference for subsequent emission prediction based on methane spectral data.
[0041] Retrieving geographic background information of the target enterprise area directly from the local storage module of the onboard computing platform avoids delays caused by data transmission and ensures the efficiency of real-time processing. The retrieved location coordinates, terrain data, and real-time wind field vector data can comprehensively reflect the geographic and environmental characteristics of the target area, providing key support for subsequent model inference by integrating multi-source information. This helps to effectively distinguish between real methane plumes and environmental interference, significantly improving the accuracy and robustness of methane emission prediction results. At the same time, it adapts to the resource-constrained operating environment of the satellite and promotes the efficient and smooth operation of the entire monitoring process.
[0042] Step 240: The preprocessed methane spectral data and geographical background information are used as multi-channel inputs and loaded into a lightweight neural network model for inference to obtain a preliminary prediction of methane emissions.
[0043] Among them, the lightweight neural network model refers to a neural network model that is optimized for the on-board operating environment, has fewer parameters and lower computational complexity, and has been trained by a special task for methane emission prediction, so that it can adapt to limited resources and complete specific prediction functions. The lightweight neural network model includes depthwise separable convolutional layers, which are a special convolutional structure in neural networks that can reduce the amount of computation and parameters while ensuring the computational effect.
[0044] For embodiments of this disclosure, step 240 may include the following steps: Step 240-1: Call the AI accelerator of the on-board computing platform to load parameters for the lightweight neural network model.
[0045] Among them, the AI accelerator refers to a dedicated hardware component in the on-board computing platform that is specifically designed to improve the computing speed and optimize the computing efficiency of artificial intelligence models; parameter loading refers to the operation of importing various parameters required for the operation of a lightweight neural network model into the corresponding computing hardware, so that the model has the conditions for operation.
[0046] In this embodiment of the disclosure, when the on-board computing platform performs calculations related to methane emission prediction, it can activate its own AI accelerator and import the running parameters of a lightweight neural network model that has been trained for the methane emission prediction task and contains depthwise separable convolutional layers into the AI accelerator to complete the adaptation of the model and the hardware, thus preparing for subsequent calculations of methane emission inference based on relevant data.
[0047] The onboard computing platform's AI accelerator is used to load parameters for lightweight neural network models. Leveraging the dedicated computing power of the AI accelerator significantly improves parameter loading efficiency and subsequent model computation speed. The design of depthwise separable convolutional layers further adapts to the limited computing resources onboard, effectively resolving the contradiction between limited onboard computing power and model computation requirements. This provides an efficient hardware and model foundation for the subsequent rapid and accurate prediction of methane emissions, contributing to the achievement of onboard real-time monitoring objectives.
[0048] Step 240-2: Input the preprocessed methane spectral data and geographical background information into the lightweight neural network model with loaded parameters. Extract methane plume features through a deep separable convolutional layer. Based on the methane plume features, locate the methane emission source location in the preset enterprise target area and make a preliminary prediction of the methane emission amount at the methane emission source location.
[0049] Among them, the preprocessed methane spectral data refers to the methane spectral data that meets the requirements of model calculation after undergoing preliminary optimization operations such as noise reduction and radiometric calibration to remove invalid interference and calibrate radiometric characteristics; the methane plume characteristics refer to the relevant characteristic information that can characterize methane emissions, covering key characteristics related to methane emissions such as spectral characteristics and morphological distribution; the methane emission source location refers to the specific spatial location of methane emissions within the pre-set target area of the enterprise; and the preliminary predicted methane emissions refer to the initial estimate of the amount of methane emissions obtained based on model calculations for a specific emission source location.
[0050] In this embodiment of the disclosure, the pre-optimized methane spectral data and the geographical background information corresponding to the preset enterprise target area can be input into a lightweight neural network model that has completed parameter loading. The lightweight neural network model performs calculations on the input multi-source data through the built-in deep separable convolutional layer, extracts the plume features that can characterize methane emissions, and then determines the specific spatial location of methane emissions within the preset enterprise target area based on these features, while completing the initial estimation of the amount of methane emissions at the spatial location.
[0051] By fusing preprocessed methane spectral data with geographic background information and inputting it into a lightweight neural network model, and leveraging deep separable convolutional layers to efficiently extract methane plume features, this approach not only adapts to the limited computing resources on the satellite and enables rapid data processing, but also improves the accuracy of plume feature extraction through the complementarity of multi-source information. Based on the extracted features, the emission source location and preliminary emission prediction are simultaneously completed, ensuring that the prediction results accurately correspond to specific emission locations. This effectively reduces errors caused by environmental interference, significantly improves the accuracy and relevance of preliminary methane emission prediction, and meets the efficiency requirements of real-time processing on the satellite.
[0052] Step 250: Call the pre-stored correction coefficients to correct the preliminary predicted methane emissions and obtain the methane emission monitoring results for the preset enterprise target area.
[0053] For the embodiments of this disclosure, step 250 may include the following steps: querying whether the pre-stored correction coefficients are within the validity period through the correction coefficient management unit of the on-board computing platform; if the correction coefficients are valid, directly calling the correction coefficients; if the correction coefficients are invalid, waiting for the ground control center to inject the latest correction coefficients through the uplink; calculating the preliminary predicted methane emissions with the latest correction coefficients within the validity period to obtain the methane emission monitoring results for the preset enterprise target area.
[0054] By verifying the validity of correction coefficients through the correction coefficient management unit, it can be ensured that the coefficients used for correction are always suitable, effectively avoiding deviations caused by invalid correction coefficients. If a correction coefficient becomes invalid, the latest ground correction coefficients are obtained in a timely manner, ensuring the continuity and timeliness of the correction process. Furthermore, by matching the valid correction coefficients with the preliminary prediction results, the systematic deviations in the preliminary predictions are significantly reduced, so that the final methane emission monitoring results have high accuracy and reliability, and can truly reflect the actual methane emission situation in the preset enterprise target area, providing accurate data support for relevant regulatory decisions.
[0055] In specific application scenarios, the ground control center can periodically collect preliminary methane emission predictions from satellites for key target areas of pre-defined enterprises within the ground monitoring network. Simultaneously, it can retrieve true methane emission data collected by corresponding calibration stations in the ground monitoring network for that area. Correlation analysis is performed on the two sets of data to clarify their relationship characteristics. Based on these characteristics, statistical regression and machine learning algorithms are used to calculate and ultimately generate the latest correction coefficients that can correct systematic biases in the on-board model. Calibration stations are dedicated monitoring stations deployed on the ground specifically for acquiring accurate methane emission data. They are a core component of the ground monitoring network, equipped with high-precision monitoring equipment, and capable of continuously capturing real quantitative data on methane emissions in their respective areas, providing a benchmark for correcting prediction biases in the on-board model.
[0056] Accordingly, the implementation steps may further include: receiving the latest correction coefficient sent by the ground control center via the uplink, wherein the latest correction coefficient is calculated by the ground control center through correlation analysis of the methane emissions of the preset enterprise target area (preliminary prediction) transmitted by the satellite and the true methane emission data of the preset enterprise target area (synchronously acquired by the ground monitoring network), using a statistical regression machine learning algorithm based on the correlation features confirmed by the correlation analysis; verifying the integrity of the received latest correction coefficient; storing the verified latest correction coefficient in the radiation-resistant non-volatile memory of the on-board computing platform, and updating the validity period of the correction coefficient. The dynamic update of the validity period of the correction coefficient ensures that the coefficients used in subsequent correction processes are always in a valid state. The radiation-resistant non-volatile memory refers to a dedicated storage component with radiation resistance that can retain stored data even after power failure; the validity period of the correction coefficient refers to the specific time range within which the correction coefficient can maintain an effective correction effect.
[0057] Step 260: Package the methane emission monitoring results into data and transmit them to the ground receiving station via satellite data transmission channel.
[0058] For embodiments of this disclosure, the steps may include: associating and binding the methane emission monitoring results with the corresponding target area location information and monitoring time information through the data management unit of the on-board computing platform; compressing the associated data and generating structured data packets according to the unified standard for space data transmission; activating the satellite high-speed data transmission channel to transmit the data packets to the ground receiving station and providing real-time feedback on the data transmission status.
[0059] By linking and binding methane emission monitoring results with target area location and time information through the data management unit, the traceability and correspondence of the data can be ensured. Compression processing can significantly reduce the data volume. Combined with a data packet format that conforms to a unified standard, it can effectively adapt to the transmission requirements of the satellite data transmission channel, significantly reducing the demand for transmission bandwidth. The activation of the high-speed satellite data transmission channel can ensure the high efficiency of data transmission, and real-time feedback on transmission status can facilitate timely monitoring of data transmission and avoid delays caused by transmission anomalies.
[0060] In specific application scenarios, as a preferred approach, the implementation steps may also include: real-time inspection of the operating status of the on-board computing platform and the working parameters of the imaging spectrometer; if an equipment abnormality is detected, a fault handling procedure is initiated, the current monitoring task is suspended, and an abnormal alarm message is sent to the ground receiving station.
[0061] By monitoring the operational status of the onboard computing platform and the working parameters of the imaging spectrometer in real time, potential equipment faults or anomalies can be detected in a timely manner. This can help avoid risks such as data distortion and mission interruption caused by equipment malfunctions. Initiating fault handling procedures and suspending the current mission can prevent the anomaly from escalating. Sending anomaly alarm information to the ground receiving station allows the ground to promptly grasp the satellite equipment status and take targeted measures. This can effectively ensure the stability, continuity, and reliability of the methane emission monitoring mission, and ensure the validity of the monitoring data and the smooth progress of subsequent regulatory work.
[0062] In summary, the technical solution in this application, by utilizing an onboard intelligent computing platform to preprocess methane spectral data, and employing a pre-built lightweight neural network model combined with geographical background information to initially predict methane emissions, and by calling pre-stored correction coefficients to correct the preliminary prediction results, can reduce the processing delay of traditional satellite methane monitoring's "onboard acquisition, ground processing" mode from several days to hours. This solves the core problem of insufficient timeliness in existing technologies, enabling timely reflection of the real-time emission status of key enterprise target areas, providing strong support for rapid response to sudden methane leaks and precise enforcement by regulatory authorities. Simultaneously, the fusion of the lightweight neural network model and geographical background information effectively reduces onboard computing resource consumption, improves the accuracy of methane plume identification and emission prediction, and, combined with a dynamic correction mechanism based on a ground monitoring network, further ensures the long-term stability and accuracy of monitoring results. Furthermore, by only transmitting methane emission monitoring results rather than massive amounts of raw data, the amount of data transmitted can be significantly reduced, lowering the bandwidth requirements of the satellite data transmission channel, and comprehensively improving the efficiency and practicality of methane emission monitoring for key enterprises.
[0063] Furthermore, as Figure 1 and Figure 2 The specific implementation of the method shown in this embodiment provides a methane emission monitoring device, such as... Figure 3 As shown, the device includes: an imaging module 31, a prediction module 32, a correction module 33, and a downlink module 34.
[0064] Imaging module 31 can be used to perform imaging operations on a preset target area of an enterprise and obtain methane spectral data during satellite operation in orbit; The prediction module 32 can be used to preprocess methane spectral data through the on-board computing platform and use a pre-built lightweight neural network model to make preliminary predictions of methane emissions based on the preprocessed methane spectral data and geographical background information. The correction module 33 can be used to call pre-stored correction coefficients to correct the preliminary predicted methane emissions and obtain the methane emission monitoring results for the preset enterprise target area; The downlink module 34 can be used to package the methane emission monitoring results into data and transmit them to the ground receiving station via the satellite data transmission channel.
[0065] In some embodiments of this application, the prediction module 32 can be specifically used to perform noise reduction and radiometric calibration on methane spectral data through the data preprocessing unit of the on-board computing platform; retrieve the geographic background information corresponding to the preset enterprise target area from the storage module of the on-board computing platform, the geographic background information including at least the location coordinates, terrain data and real-time wind field vector data of the preset enterprise target area; and load the preprocessed methane spectral data and geographic background information as multi-channel inputs into a lightweight neural network model for inference to obtain the preliminary predicted methane emissions.
[0066] In some embodiments of this application, the lightweight neural network model includes a deep separable convolutional layer and a prediction module 32, which can be used to call the AI accelerator of the on-board computing platform to load parameters into the lightweight neural network model; input the preprocessed methane spectral data and geographical background information into the parameter-loaded lightweight neural network model; extract methane plume features through the deep separable convolutional layer; locate the methane emission source location of the preset enterprise target area based on the methane plume features; and preliminarily predict the methane emission amount at the methane emission source location.
[0067] In some embodiments of this application, the correction module 33 can be used to query whether the pre-stored correction coefficients are within the validity period through the correction coefficient management unit of the on-board computing platform; if the correction coefficients are valid, the correction coefficients are directly called; if the correction coefficients are invalid, the system waits for the ground control center to inject the latest correction coefficients through the uplink; and the system calculates the preliminarily predicted methane emissions with the latest correction coefficients within the validity period to obtain the methane emission monitoring results for the preset enterprise target area.
[0068] In some embodiments of this application, such as Figure 4 As shown, the device also includes: a receiving module 35, a verification module 36, and a storage module 37.
[0069] The receiving module 35 can be used to receive the latest correction coefficient sent by the ground control center via the uplink. The latest correction coefficient is calculated by the ground control center through correlation analysis of the methane emission data of the preset enterprise target area transmitted by the satellite and the true methane emission data of the preset enterprise target area obtained synchronously by the ground monitoring network, and using the statistical regression machine learning algorithm based on the correlation characteristics confirmed by the correlation analysis. Verification module 36 can be used to perform integrity verification on the latest received correction coefficients; Storage module 37 can be used to store the latest calibration coefficients that have passed verification into the radiation-resistant non-volatile memory of the on-board computing platform and update the validity period of the calibration coefficients.
[0070] In some embodiments of this application, the correction module 33 can be specifically used to associate and bind the methane emission monitoring results with the corresponding target area location information and monitoring time information through the data management unit of the on-board computing platform; compress the associated data and generate structured data packets according to the unified standard for space data transmission; start the satellite high-speed data transmission channel to transmit the data packets to the ground receiving station and provide real-time feedback on the data transmission status.
[0071] In some embodiments of this application, such as Figure 4 As shown, the device also includes: an inspection module 38 and a sending module 39; The inspection module 38 can be used to inspect the operating status of the on-board computing platform and the working parameters of the imaging spectrometer in real time. The transmitting module 39 can be used to initiate a fault handling procedure, suspend the current monitoring task, and send abnormal alarm information to the ground receiving station if an abnormality is detected in the equipment.
[0072] It should be noted that other corresponding descriptions of the functional units involved in the methane emission monitoring device provided in this embodiment can be found in [reference needed]. Figure 1 and Figure 2 The corresponding description in [the document] will not be repeated here.
[0073] Based on the above, Figure 1 and Figure 2 Accordingly, this embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the above-described method. Figure 1 and Figure 2 The method for monitoring methane emissions is shown.
[0074] Based on this understanding, the technical solution of this application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as CD-ROM, USB flash drive, mobile hard drive, etc.) and includes several instructions to cause an electronic device (such as personal computer, server, or network device, etc.) to execute the methods of various implementation scenarios of this application.
[0075] Based on the above, Figure 1 and Figure 2 The method shown, and Figure 3 , 4 To achieve the above objectives, the present application also provides an electronic device, specifically a personal computer, tablet computer, server, or other network device, as shown in the virtual device embodiment. This device includes a storage medium and a processor; the storage medium stores a computer program; the processor executes the computer program to achieve the above-described objectives. Figure 1 and Figure 2 The method for monitoring methane emissions is shown.
[0076] Optionally, the aforementioned physical devices may also include a user interface, a network interface, a camera, radio frequency (RF) circuitry, sensors, audio circuitry, a Wi-Fi module, etc. The user interface may include a display screen, input units such as a keyboard, etc., and optional user interfaces may also include USB interfaces, card reader interfaces, etc. The network interface may optionally include standard wired interfaces, wireless interfaces (such as Wi-Fi interfaces), etc.
[0077] Those skilled in the art will understand that the physical device structure provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or combine certain components, or have different component arrangements.
[0078] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of the aforementioned physical device, supporting the operation of information processing programs and other software and / or programs. The network communication module is used to enable communication between the various components within the storage medium, as well as communication with other hardware and software in the information processing physical device.
[0079] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platform, or it can be implemented by hardware.
[0080] This invention utilizes an onboard intelligent computing platform to preprocess methane spectral data, employs a pre-built lightweight neural network model combined with geographic background information to initially predict methane emissions, and uses pre-stored correction coefficients to correct the initial prediction results. This reduces the processing delay of traditional satellite methane monitoring's "onboard acquisition, ground processing" model (which involves data acquisition from the satellite and processing on the ground) to the hourly level, addressing the core issue of insufficient timeliness in existing technologies. It can promptly reflect the real-time emission status of key enterprise target areas, providing strong support for rapid response to sudden methane leaks and precise enforcement by regulatory authorities. Simultaneously, the fusion of the lightweight neural network model and geographic background information effectively reduces onboard computing resource consumption and improves the accuracy of methane plume identification and emission prediction. Combined with a dynamic correction mechanism based on a ground monitoring network, it further ensures the long-term stability and accuracy of monitoring results. Furthermore, by only transmitting methane emission monitoring results instead of massive amounts of raw data, it significantly reduces data transmission volume and lowers the bandwidth requirements of the satellite data transmission channel, comprehensively improving the efficiency and practicality of methane emission monitoring for key enterprises.
[0081] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of a preferred embodiment, and the modules or processes shown in the drawings are not necessarily essential for implementing this application. Those skilled in the art will understand that the modules in the apparatus of the embodiment can be distributed within the apparatus of the embodiment as described, or can be modified to be located in one or more apparatuses different from this embodiment. The modules of the above-described embodiment can be combined into one module, or further divided into multiple sub-modules.
[0082] The serial numbers in this application are for descriptive purposes only and do not represent the superiority or inferiority of any particular implementation scenario. The above disclosures are merely a few specific implementation scenarios of this application; however, this application is not limited thereto, and any variations conceived by those skilled in the art should fall within the protection scope of this application.
Claims
1. A method for monitoring methane emissions, characterized in that, The method is performed on a satellite equipped with an imaging spectrometer and an onboard computing platform, and includes: During its operation in orbit, the satellite uses the imaging spectrometer to perform imaging operations on a preset target area of the enterprise to acquire methane spectral data; The on-board computing platform preprocesses the methane spectral data and uses a pre-built lightweight neural network model to make a preliminary prediction of methane emissions based on the preprocessed methane spectral data and geographical background information. The pre-stored correction coefficients are used to correct the initially predicted methane emissions, thus obtaining the methane emission monitoring results for the preset enterprise target area; The methane emission monitoring results are packaged and transmitted to the ground receiving station via satellite data transmission channel.
2. The method according to claim 1, characterized in that, The process involves preprocessing the methane spectral data using the onboard computing platform and then using a pre-built lightweight neural network model to preliminarily predict methane emissions based on the preprocessed methane spectral data and geographical background information, including: The methane spectral data is subjected to noise reduction and radiometric calibration by the data preprocessing unit of the on-board computing platform. The geographic background information corresponding to the preset enterprise target area is retrieved from the storage module of the on-board computing platform. The geographic background information includes at least the location coordinates, terrain data, and real-time wind field vector data of the preset enterprise target area. The preprocessed methane spectral data and the aforementioned geographical background information are used as multi-channel inputs and loaded into a lightweight neural network model for inference to obtain a preliminary prediction of methane emissions.
3. The method according to claim 2, characterized in that, The lightweight neural network model includes depthwise separable convolutional layers. The preprocessed methane spectral data and the geographical background information are used as multi-channel inputs and loaded into the lightweight neural network model for inference to obtain a preliminary predicted methane emission, including: The AI accelerator of the on-board computing platform is invoked to load parameters into the lightweight neural network model; The preprocessed methane spectral data and the geographical background information are input into a lightweight neural network model with loaded parameters. Methane plume features are extracted through the depthwise separable convolutional layer. Based on the methane plume features, the methane emission source location of the preset enterprise target area is located, and the methane emission amount at the methane emission source location is preliminarily predicted.
4. The method according to claim 1, characterized in that, The process of calling pre-stored correction coefficients to correct the initially predicted methane emissions yields the methane emission monitoring results for the preset enterprise target area, including: The on-board computing platform's correction coefficient management unit can be used to query whether the pre-stored correction coefficients are within their validity period. If the correction coefficient is valid, it is directly invoked; if the correction coefficient is invalid, the system waits for the ground control center to inject the latest correction coefficient via the uplink. The methane emissions are calculated by combining the preliminary predicted methane emissions with the latest correction coefficient that is within its validity period, to obtain the methane emission monitoring results for the preset enterprise target area.
5. The method according to claim 4, characterized in that, The method further includes: The latest correction coefficient is received from the ground control center via the uplink. The latest correction coefficient is calculated by the ground control center through correlation analysis of the methane emission data of the target area of the preset enterprise transmitted by the satellite and the true methane emission data of the target area of the preset enterprise synchronously obtained by the ground monitoring network, and using a statistical regression machine learning algorithm based on the correlation characteristics confirmed by the correlation analysis. Perform integrity verification on the latest received correction coefficients; The latest calibration coefficients that have passed verification are stored in the radiation-resistant non-volatile memory of the on-board computing platform, and the validity period of the calibration coefficients is updated.
6. The method according to claim 1, characterized in that, The step of packaging the methane emission monitoring results into data and transmitting it to the ground receiving station via satellite data transmission channel includes: The data management unit of the on-board computing platform associates and binds the methane emission monitoring results with the corresponding target area location information and monitoring time information. The associated data is compressed and structured data packets are generated according to the unified standard for spatial data transmission. The satellite high-speed data transmission channel is activated to transmit the data packets to the ground receiving station and provide real-time feedback on the data transmission status.
7. The method according to any one of claims 1 to 6, characterized in that, The method further includes: The operating status of the on-board computing platform and the working parameters of the imaging spectrometer are checked in real time. If an equipment malfunction is detected, a fault handling procedure is initiated, the current monitoring task is suspended, and an abnormal alarm message is sent to the ground receiving station.
8. A methane emission monitoring device, characterized in that, The device operates on a satellite carrying an imaging spectrometer and an onboard computing platform, and includes: The imaging module is used to perform imaging operations on a preset target area of an enterprise and acquire methane spectral data during the satellite's on-orbit operation using the imaging spectrometer. The prediction module is used to preprocess the methane spectral data through the on-board computing platform and use a pre-set lightweight neural network model to make a preliminary prediction of methane emissions based on the preprocessed methane spectral data and geographical background information. The correction module is used to call pre-stored correction coefficients to correct the preliminary predicted methane emissions and obtain the methane emission monitoring results for the preset enterprise target area. The downlink module is used to package the methane emission monitoring results into data and transmit them to the ground receiving station via the satellite data transmission channel.
9. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.
10. An electronic device comprising a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 7.