Method for testing concentration of carbon dioxide

Abstract

The invention discloses a method for testing carbon dioxide concentration, and relates to the technical field of environmental monitoring, and the method comprises the following steps: collecting multi-wavelength spectral data and environmental data, and carrying out data packaging and preliminary analysis; the data packet is encrypted and sent to a cloud server through wireless transmission, and the cloud server decrypts the data and integrates the data into a data lake; constructing a CO2 concentration correction model based on the integrated data set, and adjusting a CO2 concentration measurement value; performing intelligent early warning by calculating a dynamic threshold value in combination with the adjusted CO2 concentration measurement value and historical data analysis; the multi-wavelength spectrum and environmental data are integrated, and the advanced infrared spectrometer and the sensor are combined, so that high-precision CO2 monitoring is realized. Data are optimized through deep learning, concentration changes are dynamically monitored through an intelligent early warning system, a prediction trend is analyzed through a time sequence, data safety and system intelligence are enhanced, and the method is crucial to environmental protection and health management and assists in coping with climate changes and maintaining ecological balance.

Description

Technical Field

[0001] The invention relates to the technical field of environmental monitoring, in particular to a method for testing carbon dioxide concentration. Background Art

[0002] Carbon dioxide (CO2), as an important component of greenhouse gases, plays a vital role in global climate change research, environmental monitoring, and industrial production. Traditional CO2 concentration measurement methods mainly rely on chemical titration, electrochemical sensors, non-dispersive infrared (NDIR) sensors, and Fourier transform infrared spectroscopy (FTIR) technologies. In recent years, with the development of quantum cascade lasers (QCLs) technology, high-precision CO2 measurement technology based on laser absorption spectroscopy has gradually become a research hotspot. QCLs can provide high-power, narrow-linewidth infrared light sources, which are particularly suitable for accurate detection of CO2 absorption peaks in atmospheric environments, significantly improving the sensitivity and selectivity of measurements.

[0003] Although the existing technology has made significant progress in the field of CO2 concentration measurement, there are still some shortcomings. On the one hand, traditional measurement methods often need to be carried out in a laboratory environment, which makes it difficult to achieve real-time, online monitoring, especially in remote areas or complex environments. Application is limited. On the other hand, there is a lack of effective methods to ensure the accuracy and security of data during data processing and transmission, especially in the wireless transmission link, the application of data encryption and compression technology is not yet mature, resulting in low data transmission efficiency and susceptibility to interference. In addition, due to the lack of effective data fusion and analysis methods, data from different monitoring points are difficult to manage and analyze uniformly, limiting the overall effectiveness of the CO2 concentration monitoring system. More importantly, the existing technology is relatively extensive in the correction of environmental factors, and fails to make full use of advanced algorithms such as deep learning to fine-tune the CO2 concentration measurement value, thereby affecting the accuracy and reliability of the measurement results. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a method for testing carbon dioxide concentration to solve the problems of insufficient real-time monitoring, secure data transmission, unified data analysis and refined correction of environmental factors.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In the first aspect, an embodiment of the present invention provides a method for testing carbon dioxide concentration, which includes collecting multi-wavelength spectral data and environmental data, packaging and preliminary analysis of the data; encrypting the data packets and sending them to a cloud server using wireless transmission, the cloud server decrypting the data and integrating them into a data lake; fusing the data in the data lake on the cloud platform and integrating them into a data set; constructing a CO2 concentration correction model based on the integrated data set, and adjusting the CO2 concentration measurement value; combining the adjusted CO2 concentration measurement value and historical data analysis, and performing intelligent early warning by calculating dynamic thresholds; storing the early warning records in the cloud platform database, and using time series analysis to predict the long-term trend of CO2 concentration.

[0008] As a preferred embodiment of the method for testing carbon dioxide concentration of the present invention, the steps of collecting multi-wavelength spectral data and environmental data are as follows:

[0009] At selected monitoring points, infrared spectrometers equipped with quantum cascade lasers are installed to collect multi-wavelength spectral data;

[0010] Connect the built-in temperature and humidity sensor to the spectrometer to record the ambient temperature and humidity of the current environment;

[0011] The infrared spectrometer scans continuously, emitting infrared light and entering the monitoring environment through the optical path. The infrared light interacts with gas molecules, and different gases absorb light of different wavelengths, forming a unique spectrum.

[0012] The detector of the spectrometer receives the transmitted and reflected spectral signals and converts the optical signals into electrical signals, thereby obtaining the spectral data of the ambient gas;

[0013] The temperature and humidity sensor simultaneously records the temperature and humidity of the current environment.

[0014] As a preferred solution of the method for testing carbon dioxide concentration of the present invention, data packaging and preliminary analysis are performed, and the specific steps are as follows:

[0015] The microcontroller executes a preliminary spectral data analysis algorithm to predict the gas concentration using the intensity of the transmitted light and the intensity of the incident light as well as the known optical path length and absorption coefficient. The expression is:

[0016]

[0017] Among them, I0 is the intensity of the incident light, I is the intensity of the transmitted light, C is the predicted CO2 concentration, L is the length of the light path through the gas, and α is the absorption coefficient of the gas type and wavelength, which is related to the gas type and wavelength;

[0018] The built-in embedded microcontroller receives the spectral data and the temperature and humidity data, and packages the spectral data and the temperature and humidity data by organizing and formatting them, adding header and tail information, data compression, and data packet encapsulation.

[0019] As a preferred solution of the method for testing carbon dioxide concentration of the present invention, the data packet is encrypted and sent to the cloud server by wireless transmission, and the cloud server decrypts the data and integrates it into a data lake. The specific steps are as follows:

[0020] In the device, the end microcontroller uses the AES encryption algorithm to encrypt the environmental data packet, and uses the Diffie-Hellman protocol for key negotiation to generate an encrypted data stream;

[0021] Using Long Range, the encrypted data packets are converted into signals suitable for wireless transmission;

[0022] The wireless module on the device sends the modulated signal into the air and searches for the nearest gateway through the LoRaWAN network;

[0023] After the nearest LoRa gateway receives the signal, it demodulates and repackages the data, and after packaging, it forwards it to the cloud server for data transmission over the Internet;

[0024] The cloud server receives the data from the gateway and decrypts the information containing CO2 concentration, temperature and humidity using the same encryption algorithm and key as the device side;

[0025] The server integrates the decrypted data with data from other monitoring points into a unified data lake, and classifies and stores CO2 concentration data and temperature and humidity data on the cloud platform.

[0026] As a preferred solution of the method for testing carbon dioxide concentration of the present invention, the data in the data lake is merged and integrated into a data set on the cloud platform, and the specific steps are as follows:

[0027] Clean, verify, standardize and normalize the CO2 concentration data and temperature and humidity data in the data lake;

[0028] Based on the common time frame requirements of all monitoring point data in the data lake, a unified benchmark time interval is established as the time benchmark for all data in the data lake;

[0029] Adopt the nearest neighbor matching strategy to accurately find the data record closest to each benchmark time, and then use the linear interpolation method to fill in the missing monitoring point data near the benchmark time to align the time of data in different geographical locations in the data lake;

[0030] The address information of the monitoring points is converted into longitude and latitude coordinates through geocoding, and then these coordinates are unified into a common projection coordinate system through projection transformation;

[0031] Use spatial indexing technology to accelerate the retrieval and processing of geospatial data and perform spatial alignment of data from different geographic locations in the data lake;

[0032] Through comprehensive time alignment and space alignment, it facilitates the deep integration of cross-regional monitoring data in the data lake;

[0033] Use ETL to integrate the fused data into data set D.

[0034] As a preferred embodiment of the method for testing carbon dioxide concentration of the present invention, the CO2 concentration correction model is constructed based on the integrated data set to adjust the CO2 concentration measurement value, and the specific steps are as follows:

[0035] According to the spectral data and temperature and humidity information, a CO2 concentration correction model based on deep neural network is constructed. By reducing the measurement error caused by environmental data, the CO2 concentration prediction result is adjusted. The expression is:

[0036]

[0037] in, is the CO2 concentration after correction, C is the CO2 concentration obtained by preliminary analysis, T is the ambient temperature, H is the ambient humidity, T r is the reference temperature, H r is the reference humidity, β is the temperature sensitivity coefficient, γ is the humidity sensitivity coefficient, η is the seasonal fluctuation amplitude, ω is the angular frequency, t is the time, and φ is the periodic adjustment coefficient;

[0038] As a preferred solution of the method for testing carbon dioxide concentration of the present invention, wherein: the adjusted CO2 concentration measurement value and historical data analysis are combined to calculate the dynamic threshold value for intelligent early warning, and the specific steps are as follows:

[0039] Use LSTM to identify the trend and periodic pattern of CO2 concentration over time, and determine the baseline level of CO2 concentration based on the change trend;

[0040] The baseline level of CO2 concentration is determined based on historical data, and the dynamic threshold is calculated based on the fluctuation of baseline and historical data. The expression is:

[0041]

[0042] Among them, τ t is the alert level set at time point t based on the volatility of historical data, m tis the moving average at time point t, and w is the dynamic threshold τ t Sensitivity to recent data changes, k is a positive constant factor, x i is the measured value of CO2 concentration at time point i;

[0043] Real-time monitoring of corrected carbon dioxide concentration and with the dynamic threshold τ t Make comparisons to control the warning mechanism and the severity level of the warning;

[0044] when When the CO2 concentration exceeds the normal range, it is considered that the CO2 concentration does not exceed the normal range and there is no need to trigger the early warning mechanism;

[0045] when When the CO2 concentration exceeds the normal range, the warning mechanism is triggered. Exceed τ t Determine the severity of the alarm and make corresponding strategies based on the severity level.

[0046] As a preferred solution of the method for testing carbon dioxide concentration of the present invention, wherein: the warning record is stored in the cloud platform database, and the long-term trend of CO2 concentration is predicted by time series analysis. The specific steps are as follows:

[0047] Store warning records, CO2 concentrations, and environmental parameters in the cloud platform database, use time series to analyze historical data, identify periodic patterns, seasonal changes, and potential nonlinear relationships, and predict long-term trends in CO2 concentrations. At the same time, integrate external data sources to enhance the comprehensiveness and accuracy of predictions.

[0048] Based on the results of trend forecasting, a long-term trend forecast report is generated.

[0049] In a second aspect, an embodiment of the present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the method for testing carbon dioxide concentration as described in the first aspect of the present invention is implemented.

[0050] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the method for testing carbon dioxide concentration as described in the first aspect of the present invention is implemented.

[0051] The beneficial effects of the present invention are as follows: by collecting multi-wavelength spectral data and environmental data, combining the infrared spectrometer of the quantum cascade laser with the temperature and humidity sensor, high-precision CO2 concentration measurement can be achieved. The preliminary analysis of the data, encrypted transmission to the cloud, and the use of deep learning technology for data fusion and correction have significantly improved the accuracy and efficiency of monitoring. The intelligent early warning system can effectively monitor changes in CO2 concentration and promptly warn of abnormal situations through dynamic threshold calculation. In addition, time series analysis is used to predict the long-term trend of CO2 concentration, providing strong data support for environmental protection and public health management. The integrated application of this series of technologies not only enhances the security and reliability of the data, but also greatly improves the intelligence level and response speed of the monitoring system, which is of great significance for responding to global climate change and maintaining ecological balance. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0053] Figure 1 This is a flow chart of the method for testing carbon dioxide concentration in Example 1.

[0054] Figure 2 This is a flow chart of the intelligent early warning in Example 1. DETAILED DESCRIPTION

[0055] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0056] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0057] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0058] Example 1, reference Figure 1 and Figure 2, which is the first embodiment of the present invention, provides a method for testing carbon dioxide concentration, comprising the following steps:

[0059] S1. Collect multi-wavelength spectral data and environmental data, and perform data packaging and preliminary analysis.

[0060] Furthermore, an infrared spectrometer equipped with a quantum cascade laser is installed at the selected monitoring point (the best installation location is selected according to the site conditions to ensure that the infrared spectrometer can fully contact the ambient air and avoid obstruction and reflection affecting the measurement accuracy). The QCL is capable of emitting high-power, narrow-linewidth infrared light, which is suitable for the precise detection of the CO2 absorption peak, with particular attention paid to the band near 4.26μm, which is the main absorption band of CO2.

[0061] Connect the built-in temperature and humidity sensor to the spectrometer to record the ambient temperature and humidity of the current environment;

[0062] The infrared spectrometer scans continuously, emitting infrared light and entering the monitoring environment through the optical path. The infrared light interacts with gas molecules, and different gases absorb light of different wavelengths, forming a unique spectrum.

[0063] The detector of the spectrometer receives the transmitted and reflected spectral signals and converts the optical signals into electrical signals, thereby obtaining the spectral data of the ambient gas;

[0064] The temperature and humidity sensor simultaneously records the temperature and humidity of the current environment;

[0065] The microcontroller executes a preliminary spectral data analysis algorithm to predict the gas concentration using the intensity of the transmitted light and the intensity of the incident light as well as the known optical path length and absorption coefficient. The expression is:

[0066]

[0067] Among them, I0 is the intensity of the incident light, I is the intensity of the transmitted light, C is the predicted CO2 concentration, L is the length of the light path through the gas, and α is the absorption coefficient of the gas type and wavelength, which is related to the gas type and wavelength;

[0068] The built-in embedded microcontroller receives the spectral data and temperature and humidity data, and packages the data by organizing and formatting the data, adding header and footer information, data compression and data packet encapsulation.

[0069] It should also be noted that the process of converting optical signals into electrical signals and obtaining spectral data of ambient gases is achieved through an infrared spectrometer. Specifically, when the infrared beam emitted by an infrared spectrometer equipped with a quantum cascade laser (QCL) passes through the monitoring environment, the CO2 molecules therein selectively absorb infrared light of a specific wavelength to form a unique absorption spectrum. After the infrared beam passes through the gas sample, the remaining light (transmitted light) is captured by the detector of the spectrometer together with the reflected light. The detector, such as a photodiode array or a pyroelectric detector, converts the received optical signal into an electrical signal that is proportional to the light intensity. This electrical signal is then amplified, digitized, and converted into a digital signal through an analog-to-digital converter (ADC) for easy processing by a microcontroller. By analyzing these digital signals, the microcontroller can reconstruct the spectral image of CO2, that is, the light intensity distribution of different wavelengths, and extract the concentration information of CO2 from it.

[0070] Data organization and formatting: The microcontroller organizes and formats the data. The data may need to be arranged according to a certain protocol or format, for example, placing spectral data in the first half of the data packet and environmental parameters in the second half, or defining data fields according to a specific protocol.

[0071] Add header and tail information: Add header information before the data, which usually contains metadata such as the data packet identifier, data type, data length, source address, destination address, etc. Add tail information at the end of the data packet, which may be a checksum or CRC (cyclic redundancy check) code for integrity checking.

[0072] Data compression: In order to reduce the bandwidth requirements and energy consumption of data transmission, data may be compressed. The compression algorithm can be a simple differential encoding (transmitting only the change in data) or a more complex algorithm such as LZ77, Huffman coding, etc.

[0073] Packet encapsulation: Finally, the data, header information, trailer information, and possible compression and encryption results are combined into a complete packet. This packet is now ready for transmission over the network or wireless interface.

[0074] S2, encrypts the data packet and sends it to the cloud server using wireless transmission. The cloud server decrypts the data and integrates it into a data lake.

[0075] Furthermore, in the device, the end microcontroller uses the AES encryption algorithm to encrypt the environmental data packet, and uses the Diffie-Hellman protocol for key negotiation to generate an encrypted data stream;

[0076] Using Long Range, the encrypted data packets are converted into signals suitable for wireless transmission;

[0077] The wireless module on the device sends the modulated signal into the air and searches for the nearest gateway through the LoRaWAN network;

[0078] After the nearest LoRa gateway receives the signal, it demodulates and repackages the data, and after packaging, it forwards it to the cloud server for data transmission over the Internet;

[0079] The cloud server receives the data from the gateway and decrypts the information containing CO2 concentration, temperature and humidity using the same encryption algorithm and key as the device side;

[0080] The server integrates the decrypted data with data from other monitoring points into a unified data lake, and classifies and stores CO2 concentration data, temperature and humidity data on the cloud platform;

[0081] It should also be noted that the specific process of finding the nearest gateway through the LoRaWAN network is as follows: the wireless module on the device side is equipped with LoRa (Long Range) modulation technology, which converts the encrypted data packets into signals suitable for long-distance wireless transmission. LoRa technology utilizes the principle of spread spectrum communication and can provide long-distance transmission under low power consumption conditions, which is particularly suitable for remote monitoring devices in IoT applications. When the device sends these LoRa signals, the signals propagate in the air at a specific frequency and spread spectrum factor. The design of the LoRaWAN network allows multiple gateways to listen to the same wireless signal, which means that when the signal is broadcast, multiple gateways may receive the signal at the same time. However, because the LoRaWAN network uses geolocation and signal strength indication (RSSI) technology, it is able to identify the gateway with the strongest signal or the nearest one.

[0082] Demodulation and repackaging of data are as follows: After the nearest LoRa gateway receives the LoRa signal, it first demodulates the signal to recover the original data. This process involves reversing LoRa's modulation technology, extracting the data bit stream, and performing error correction. Subsequently, the gateway converts the data from the LoRa PHY layer format to the MAC layer, removes unnecessary overhead, and repacks it into a format that complies with IP network transmission. The data is securely transmitted to the cloud server via the Internet, achieving seamless conversion from physical layer signals to network layer data packets, ensuring remote transmission of monitoring data.

[0083] The integration into a unified data lake is as follows: After the server receives the decrypted monitoring data, it will integrate it with the previously collected data from other monitoring points to form a unified data lake. This process involves standardization and format unification of data to ensure that all data can be understood and processed under a common architecture. The data lake acts as a large, flexible repository that can accommodate data in various formats and sources, including structured, semi-structured and unstructured data. Through integration, the data lake provides a unified data view, which facilitates subsequent data analysis and mining.

[0084] Classified storage is as follows: Classified storage refers to the organized management of data in the data lake to ensure that different types of data can be effectively retrieved and analyzed. In the context of the present invention, environmental monitoring information such as CO2 concentration data, temperature and humidity data will be assigned to corresponding categories for storage. This usually involves establishing data tags and metadata, and using directory and indexing technologies to improve the efficiency of data queries. Classified storage helps maintain a clear structure of data and simplify the data analysis process. It also facilitates compliance with data compliance and privacy protection regulations, ensuring proper isolation and protection of sensitive data.

[0085] S3. Integrate the data in the data lake into a data set on the cloud platform.

[0086] Furthermore, the CO2 concentration data and temperature and humidity data in the data lake are cleaned, verified, standardized and normalized;

[0087] Based on the common time frame requirements of all monitoring point data in the data lake, establish a unified benchmark time interval, such as integer times every minute, hour, or day, as the time benchmark for all data in the data lake;

[0088] Adopt the nearest neighbor matching strategy to accurately find the data record closest to each benchmark time. Even if a monitoring point has data before and after the benchmark time, the point closest to it will be selected as the matching object. Then use linear interpolation to fill in the missing monitoring point data near the benchmark time, and align the time of data from different geographical locations in the data lake.

[0089] The address information of the monitoring points is converted into longitude and latitude coordinates through geocoding, and then these coordinates are unified into a common projection coordinate system through projection transformation;

[0090] Use spatial indexing technology to accelerate the retrieval and processing of geospatial data and perform spatial alignment of data from different geographic locations in the data lake;

[0091] Through comprehensive time alignment and space alignment, it facilitates the deep integration of cross-regional monitoring data in the data lake;

[0092] Use ETL to integrate the merged data into data set D;

[0093] It should also be noted that data cleaning: In the data lake, each piece of data is first cleaned to remove any obvious erroneous data, such as outliers or missing values ​​that are beyond the normal range. For missing values, if the missing ratio is small, you can use interpolation methods such as time series interpolation or fill it based on the average value of neighboring points; if the missing ratio is large, consider discarding the entry or using a machine learning model to predict and fill it.

[0094] Data validation: Ensure that the data format is correct, such as whether the date and time stamps are consistent, whether the value type is correct, and whether the data units are consistent. In addition, the integrity of the data needs to be verified to ensure that no important fields are missing.

[0095] Standardization: Standardize the data to eliminate the impact of dimension so that data from different monitoring points can be compared on the same scale. Standardization can be achieved by subtracting the mean and dividing by the standard deviation;

[0096] Normalization: Further scale the data to a fixed range, such as between 0 and 1, expressed as:

[0097]

[0098] Among them, x max and x min are the minimum and maximum values ​​in the data set, respectively. x represents a specific value in the original data set, and x′ represents the value after normalization.

[0099] Use ETL to integrate the fused data into data set D as follows:

[0100] Data extraction: Extract processed data from different partitions or tables in the data lake using ETL (Extract, Transform, Load) tools.

[0101] Data Integration: Merge all extracted data into a middle tier, which is a data warehouse, to ensure data consistency and integrity.

[0102] Data loading: Load the integrated data into the final dataset, which is a specially optimized big data storage solution such as Hadoop HDFS, Amazon S3.

[0103] S4. Construct a CO2 concentration correction model based on the integrated data set and adjust the CO2 concentration measurement value.

[0104] Furthermore, according to the spectral data and temperature and humidity information, a CO2 concentration correction model based on a deep neural network is constructed to adjust the CO2 concentration prediction result by reducing the measurement error caused by environmental data. The model structure contains multiple convolutional layers, fully connected layers, and attention mechanisms to capture the local features and overall patterns of spectral data. At the same time, environmental data (temperature, humidity) is used for multimodal fusion. The expression is:

[0105]

[0106] in, is the CO2 concentration after correction, C is the CO2 concentration obtained by preliminary analysis, T is the ambient temperature, H is the ambient humidity, T r is the reference temperature, H r is the reference humidity, β is the temperature sensitivity coefficient, γ is the humidity sensitivity coefficient, η is the seasonal fluctuation amplitude, ω is the angular frequency, t is the time, and φ is the periodic adjustment coefficient;

[0107] when Close to C, it means that environmental factors have little influence on the measured value;

[0108] when Significantly greater or less than C, indicating that temperature, humidity or seasonal changes have a significant impact on the measured value;

[0109] Divide the dataset D into training set D t , validation set D v ;

[0110] The training set D t As the model input, the CO2 concentration correction model is trained to minimize the loss function, which is expressed as:

[0111]

[0112] Among them, L(θ) is the loss function, θ is the parameter set, which includes the temperature sensitivity coefficient, humidity sensitivity coefficient, seasonal fluctuation amplitude, etc., and x represents the input feature vector, which includes all relevant environmental variables affecting CO2 concentration measurement, such as temperature, humidity, etc.

[0113] Use the validation set D v Adjust hyperparameters and prevent overfitting;

[0114] It should also be noted that at the end of each epoch, the validation set D v Calculate the loss and indicators (such as accuracy, recall, etc.) on the validation set to evaluate the performance of the model;

[0115] Observe the performance on the validation set. If the model is vThe poor performance on may be due to overfitting or underfitting;

[0116] According to D v Adjust hyperparameters based on the performance. For example, if the model is overfitting, you can increase the regularization strength, reduce the network complexity, or add dropout layers. If it is underfitting, you can try increasing the network depth or width, reducing the regularization strength, or increasing the number of training rounds.

[0117] Repeat the training and evaluation process until D v Achieve satisfactory performance;

[0118] Monitoring D v If the performance no longer improves or starts to deteriorate within a certain number of consecutive epochs, training should be stopped to prevent the model from overfitting on the training set.

[0119] S5. Combine the adjusted CO2 concentration measurement value with historical data analysis and perform intelligent early warning by calculating dynamic thresholds.

[0120] Furthermore, LSTM is used to identify trends and periodic patterns in CO2 concentration over time, and a baseline level of CO2 concentration is determined based on the trend (this can be a long-term average or median, depending on the data distribution);

[0121] The baseline level of CO2 concentration is determined based on historical data. This can be a long-term average or median, depending on the data distribution. Based on the fluctuations of the baseline and historical data, the dynamic threshold is calculated. The dynamic threshold is automatically adjusted with seasonal changes and weather conditions to reflect the impact of different background conditions on CO2 concentration. The expression is:

[0122]

[0123] Among them, τ t is the alert level set at time point t based on the volatility of historical data, m t is the moving average at time point t, and w is the dynamic threshold τ t Sensitivity to recent data changes, k is a positive constant factor, x i is the measured value of CO2 concentration at time point i;

[0124] Real-time monitoring of corrected carbon dioxide concentration and with the dynamic threshold τ t Make comparisons to control the warning mechanism and the severity level of the warning;

[0125] when When the CO2 concentration exceeds the normal range, it is considered that the CO2 concentration does not exceed the normal range and there is no need to trigger the early warning mechanism;

[0126] when When the CO2 concentration exceeds the normal range, the warning mechanism is triggered. Exceed τ t Determine the severity of the alarm and make corresponding strategies based on the severity level.

[0127] For example, if Ratio τ t If the level is higher than 50%, a Level 1 warning is triggered. At this time, the environmental monitoring personnel should be notified immediately to check the sensor status, confirm whether it is a false alarm, and start monitoring trend changes and prepare further response plans.

[0128] if Ratio τ t If the level is 50% higher, a Level 2 warning is triggered and a comprehensive emergency response plan is immediately implemented, including emergency evacuation of potentially threatened areas, notification of local emergency management departments, initiation of pollution control and cleanup procedures, and issuance of emergency health warnings to the public;

[0129] It should also be noted that the alarm should contain the following information: the specific value of the CO2 concentration value that exceeds the limit;

[0130] Monitoring point location: geographical coordinates or description of the monitoring point where the exceedance occurred;

[0131] Recommended Action: Based on the extent of the exceedance, provide appropriate recommendations such as increasing ventilation, adjusting environmental control equipment, or taking other emergency measures.

[0132] S6. Store the warning records in the cloud platform database and use time series analysis to predict the long-term trend of CO2 concentration.

[0133] Furthermore, warning records, CO2 concentrations, and environmental parameters (such as temperature and humidity) are stored in the cloud platform database, and historical data are analyzed using time series to identify periodic patterns, seasonal changes, and potential nonlinear relationships, and to predict long-term trends in CO2 concentrations. At the same time, external data sources such as weather forecasts, industrial activity indexes, and population density are integrated to enhance the comprehensiveness and accuracy of the forecast;

[0134] Based on the results of trend forecasting, a long-term trend forecast report is generated.

[0135] The report should include an explanation of the projections and an analysis of the possible drivers of future changes in CO2 concentrations.

[0136] It should also be noted that the long-term trend of CO2 concentration is predicted by time series analysis. First, the time series of CO2 concentration is extracted from historical data, and statistical methods are applied to identify periodic patterns and seasonal changes, such as annual and daily changes. Then, LSTM is used to reveal nonlinear relationships and potential trends in the data. At the same time, multiple external data sources are integrated, including weather forecast data, industrial activity index, and population density information. These data are associated with CO2 concentration data through API or data import to enhance the comprehensiveness and accuracy of the prediction model. For example, meteorological data can provide environmental conditions such as temperature, wind speed, and precipitation, which affect CO2 emissions and diffusion; the industrial activity index reflects the intensity of economic activities and is an important driving factor of CO2 emissions; and population density data is associated with the local impact of human activities on CO2 concentration. By incorporating these external variables into the model, the trend of CO2 concentration can be predicted more accurately, providing strong support for environmental management decisions.

[0137] This embodiment also provides a computer device, which is suitable for the method of testing carbon dioxide concentration, including: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the method of testing carbon dioxide concentration proposed in the above embodiment.

[0138] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0139] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for testing the carbon dioxide concentration proposed in the above embodiment is implemented; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0140] In summary, the present invention achieves automation and intelligence in CO2 monitoring by integrating infrared spectroscopy technology and environmental sensing, adopts encrypted transmission and cloud integration to ensure data security and build a unified data lake, improves the measurement accuracy of CO2 concentration through data cleaning and deep learning models, combines historical data with the intelligent early warning system to dynamically adjust thresholds and respond to anomalies in a timely manner, stores early warning records, predicts long-term trends, and provides a scientific basis for environmental management, thus realizing automation and intelligence in CO2 monitoring and making significant contributions to environmental protection and climate research.

[0141] Example 2, referring to Table 1, is the second example of the present invention. To further verify the advancement of the present invention, experimental simulation data of a method for testing carbon dioxide concentration are provided.

[0142] The experiment was conducted at four different locations, each representing a typical environment (city center, suburbs, industrial area, forest edge). An infrared spectrometer equipped with a quantum cascade laser (QCL) was installed at each monitoring point to ensure that the equipment can fully contact the ambient air. The infrared spectrometer scans continuously, focusing on the band near 4.26μm to accurately detect the CO2 absorption peak. At the same time, the built-in temperature and humidity sensor records environmental parameters for later data correction.

[0143] First, an advanced infrared spectrometer system was deployed at four monitoring points in different environments to continuously collect CO2 concentration and related environmental data, and the data was packaged through preliminary algorithm analysis.

[0144] Secondly, the packaged data is securely encrypted and efficiently transmitted to the cloud through the LoRaWAN network. The cloud server decrypts and integrates the data to form a unified data lake.

[0145] Next, a deep learning model was used to clean, fuse and correct the information in the data lake, build an accurate CO2 concentration correction model, and determine dynamic thresholds based on historical data analysis to achieve intelligent early warning.

[0146] Finally, the warning records are stored in the cloud database, and time series analysis is used to predict the future trend of CO2 concentration, generating a detailed long-term forecast report, which provides strong support for environmental monitoring and decision-making.

[0147] Table 1 CO2 concentration prediction table

[0148]

[0149] By analyzing the data in the above table, it can be seen that the prediction accuracy of the present invention for monitoring CO2 concentration in different environments is higher than 95%, which reflects the high accuracy of the system. Compared with traditional monitoring methods, such as chemical sensors, whose accuracy is often limited by environmental factors, the present invention effectively overcomes the influence of environmental factors such as temperature and humidity on the measurement results through the correction of the deep neural network model, and significantly improves the accuracy of CO2 concentration measurement.

[0150] The method for measuring carbon dioxide concentration of the present invention can not only monitor the CO2 content in the atmosphere in real time to ensure the accuracy and timeliness of the data, but also predict the changing trend of CO2 concentration through an integrated intelligent analysis system to provide early warning of possible environmental risks. Especially for climate change research and industrial emission monitoring, this method provides more sophisticated and comprehensive data support. This reflects the innovation and practicality of the present invention in the field of environmental monitoring technology, and provides a scientific basis for the formulation of sustainable development and environmental protection strategies.

[0151] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for testing carbon dioxide concentration, Features: include, Collect multi-wavelength spectral data and environmental data, and perform data packaging and preliminary analysis; The data packets are encrypted and sent to the cloud server using wireless transmission. The cloud server decrypts the data and integrates it into a data lake. On the cloud platform, the data in the data lake is integrated and consolidated into data sets; Build CO based on the integrated dataset 2 Concentration correction model, adjusting CO 2 Concentration measurements; Combined with adjusted CO 2 Analysis of concentration measurements and historical data, and intelligent early warning by calculating dynamic thresholds; The warning records are stored in the cloud platform database, and the CO 2 Long-term trends in concentrations.

2. The method for testing carbon dioxide concentration as claimed in claim 1, Features: The specific steps of collecting multi-wavelength spectral data and environmental data are as follows: At selected monitoring points, infrared spectrometers equipped with quantum cascade lasers were installed; Connect the built-in temperature and humidity sensor to the spectrometer to record the ambient temperature and humidity of the current environment; The infrared spectrometer scans continuously, emitting infrared light and entering the monitoring environment through the optical path. The infrared light interacts with gas molecules, and different gases absorb light of different wavelengths, forming a unique spectrum. The detector of the spectrometer receives the transmitted and reflected spectral signals and converts the optical signals into electrical signals, thereby obtaining the spectral data of the ambient gas; The temperature and humidity sensor simultaneously records the temperature and humidity of the current environment.

3. The method for testing carbon dioxide concentration as claimed in claim 2, Features: Carry out data packaging and preliminary analysis. The specific steps are as follows: The microcontroller executes a preliminary spectral data analysis algorithm to predict the gas concentration using the intensity of the transmitted light and the intensity of the incident light as well as the known optical path length and absorption coefficient. The expression is: Among them, I 0 is the intensity of the incident light, I is the intensity of the transmitted light, and C is the predicted CO 2 The concentration of, L is the length of the light path through the gas, α is the absorption coefficient of the gas type and wavelength; The built-in embedded microcontroller receives the spectral data and the temperature and humidity data, and packages the spectral data and the temperature and humidity data by organizing and formatting them, adding header and tail information, data compression, and data packet encapsulation.

4. The method for testing carbon dioxide concentration as claimed in claim 3, Features: The data packet is encrypted and sent to the cloud server using wireless transmission. The cloud server decrypts the data and integrates it into a data lake. The specific steps are as follows: In the device, the end microcontroller uses the AES encryption algorithm to encrypt the environmental data packet, and uses the Diffie-Hellman protocol for key negotiation to generate an encrypted data stream; Using Long Range, the encrypted data packets are converted into signals suitable for wireless transmission; The wireless module on the device sends the modulated signal into the air and searches for the nearest gateway through the LoRaWAN network; After the nearest LoRa gateway receives the signal, it demodulates and repackages the data, and after packaging, it forwards it to the cloud server for data transmission over the Internet; The cloud server receives the data from the gateway and uses the same encryption algorithm and key as the device to decrypt the data containing CO 2 Information on concentration, temperature and humidity; The server integrates the decrypted data with data from other monitoring points into a unified data lake and monitors the CO 2 Concentration data, temperature and humidity data are classified and stored.

5. The method for testing carbon dioxide concentration as claimed in claim 4, Features: The specific steps of fusing and integrating the data in the data lake into a data set on the cloud platform are as follows: CO in the data lake 2 Clean, verify, standardize and normalize the concentration data and temperature and humidity data; Based on the common time frame requirements of all monitoring point data in the data lake, a unified benchmark time interval is established as the time benchmark for all data in the data lake; Adopt the nearest neighbor matching strategy to accurately find the data record closest to each benchmark time, and then use the linear interpolation method to fill in the missing monitoring point data near the benchmark time to align the time of data in different geographical locations in the data lake; The address information of the monitoring points is converted into longitude and latitude coordinates through geocoding, and then these coordinates are unified into a common projection coordinate system through projection transformation; Use spatial indexing technology to accelerate the retrieval and processing of geospatial data and perform spatial alignment of data from different geographic locations in the data lake; Through comprehensive time alignment and space alignment, it facilitates the deep integration of cross-regional monitoring data in the data lake; Use ETL to integrate the fused data into data set D.

6. The method for testing carbon dioxide concentration as claimed in claim 5, Features: The CO is constructed based on the integrated data set. 2 Concentration correction model, adjusting CO 2 Concentration measurement value, the specific steps are as follows: Based on spectral data and temperature and humidity information, a CO 2 The concentration correction model adjusts the CO 2 The concentration prediction result is expressed as: in, is the corrected CO 2 concentration, C is the CO obtained by preliminary analysis 2 concentration, T is the ambient temperature, H is the ambient humidity, T r is the reference temperature, H r is the reference humidity, β is the temperature sensitivity coefficient, γ is the humidity sensitivity coefficient, η is the seasonal fluctuation amplitude, ω is the angular frequency, t is the time, and φ is the periodic adjustment coefficient.

7. The method for testing carbon dioxide concentration as claimed in claim 6, Features: The combined adjusted CO 2 The concentration measurement and historical data analysis are used to calculate the dynamic threshold for intelligent early warning. The specific steps are as follows: Using LSTM to identify CO 2 The trend and periodic pattern of concentration change over time, and the determination of CO 2 Baseline concentration level; Determine CO from historical data 2 The baseline level of concentration is used to calculate the dynamic threshold based on the fluctuation of baseline and historical data. The expression is: Among them, τ t is the alert level set at time point t based on the volatility of historical data, m t is the moving average at time point t, and w is the dynamic threshold τ t Sensitivity to recent data changes, k is a positive constant factor, x i is the CO at time point i 2 The measured value of concentration; Real-time monitoring of corrected carbon dioxide concentration and with the dynamic threshold τ t Make comparisons to control the warning mechanism and the severity level of the warning; when When CO 2 The concentration does not exceed the normal range, so there is no need to trigger the warning mechanism; when When CO 2 The concentration exceeds the normal range, triggering the warning mechanism. Exceed τ t Determine the severity of the alarm and make corresponding strategies based on the severity level.

8. The method for testing carbon dioxide concentration as claimed in claim 7, Features: The warning records are stored in the cloud platform database, and the CO 2 The long-term trend of concentration is as follows: The warning records, CO 2 The concentration and environmental parameters are stored in the cloud platform database, and the historical data are analyzed using time series to identify periodic patterns, seasonal changes and potential nonlinear relationships, and predict CO 2 Long-term trends in concentrations, while integrating external data sources to enhance the comprehensiveness and accuracy of forecasts; Based on the results of trend forecasting, a long-term trend forecast report is generated.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program. Features: When the processor executes the computer program, the steps of the method for testing the carbon dioxide concentration according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, Features: When the computer program is executed by a processor, the steps of the method for testing carbon dioxide concentration according to any one of claims 1 to 7 are implemented.

Images