Real-time online monitoring and early warning equipment, methods, systems and media for methane gas concentration

Abstract

This invention provides a real-time online monitoring and early warning device, method, system, and medium for methane gas concentration. The method includes: acquiring environmental data, including apparent methane concentration and other sensor data; converting the apparent methane concentration to obtain the true methane concentration; inputting the true methane concentration and other sensor data into a trained prediction model to obtain a prediction result; and performing threshold judgment based on the prediction result to output an early warning signal, including a first-level warning, a second-level warning, and a third-level warning. This invention determines the true methane concentration through conversion coefficients, and combined with the prediction model, can effectively detect abnormal changes in methane concentration in advance, improving the accuracy and timeliness of early warnings. Simultaneously, the real-time online monitoring and early warning device features high protection, high sensitivity, and low energy consumption, making it suitable for large-scale municipal drainage pipeline network deployment and ensuring municipal public safety.

Description

Technical Field

[0001] This invention belongs to the field of environmental monitoring and urban public safety technology, and in particular relates to a real-time online monitoring and early warning device, method, system and medium for methane gas concentration. Background Technology

[0002] With the continuous acceleration of urbanization, the amount of municipal sewage discharge continues to increase. During the long-term operation of municipal sewage pipelines, due to the decomposition of organic matter, hydraulic disturbance and anaerobic fermentation of microorganisms, flammable gas, mainly CH4, is easily accumulated in local spaces. When the methane concentration reaches the explosion limit (5% to 15%), it is very easy to cause an explosion accident, which seriously threatens public safety and the life safety of workers.

[0003] Currently, methane gas leakage or accumulation in municipal drainage pipes is a major factor leading to explosion accidents. Traditional methane detection methods mostly rely on single-point alarms, which are slow to respond and untimely in warning. Existing equipment often fails to combine real-time trend prediction for intelligent early warning, and the equipment deployment and maintenance are complex.

[0004] Existing technologies are limited to data monitoring of combustible gases, lacking targeted monitoring of methane gas and prediction of related formation trends, and the environmental parameters are relatively simple. Some studies focus on real-time monitoring of harmful gas concentrations in municipal sewage pipes, which lacks specificity and has a weak ability to perceive the formation patterns and concentration change trends of the monitored gases in advance, and lacks an effective time series prediction mechanism.

[0005] Meanwhile, in the complex and variable hydraulic conditions and pipeline environment of municipal sewage pipelines, the methane gas generation process is affected by a variety of environmental parameters (such as temperature, humidity, water level, wind speed, etc.). How to monitor and warn of methane concentration based on multiple environmental parameters has become a research direction and an urgent problem to be solved. Summary of the Invention

[0006] In view of the shortcomings of the prior art described above, the purpose of this invention is to provide a real-time online monitoring and early warning device, method, system and medium for methane gas concentration, to solve the problem of real-time monitoring and early warning of methane concentration in municipal drainage pipelines in the prior art.

[0007] In a first aspect, the present invention provides a real-time online monitoring and early warning device for methane gas concentration, comprising:

[0008] The equipment body contains a sensor array, a controller, and a power supply.

[0009] The sensor group includes a methane sensor, a temperature and humidity sensor, a liquid level sensor, and a wind speed sensor, used to collect environmental data inside the pipeline;

[0010] The controller includes a processing unit and a communication unit. The processing unit is used to receive environmental data, process it, and generate an early warning signal. The communication unit is used to communicate the early warning signal to the outside world.

[0011] The power supply includes a power supply, a power compartment, and a power connection cable. The power supply is used to supply power to the sensor group and the controller.

[0012] In one possible implementation of this application, the device body is placed on a support plate and limited by at least two fixing plates. The support plate and the fixing plates are both fixed to a fixing rod by fixing rings. The fixing rod is fixedly connected to the manhole ring of the pipeline inspection well. The materials of the support plate, the fixing plate and the fixing rod include metal.

[0013] In one possible implementation of this application, the device body includes a housing with an openable and closable lid, and a handle is installed on one side of the housing, wherein the housing is made of engineering plastic.

[0014] Secondly, the present invention provides a method for real-time online monitoring and early warning of methane gas concentration, applicable to any of the methane gas concentration real-time online monitoring and early warning devices described in any one of the claims, wherein the method includes:

[0015] Acquire environmental data, including apparent methane concentration and other sensor data;

[0016] The apparent value of methane concentration is converted to obtain the true value of methane concentration.

[0017] The prediction result is obtained by inputting the true value of methane concentration and the other sensor data into the trained prediction model.

[0018] Based on the prediction results, a threshold judgment is performed to output an early warning signal, which includes a first-level early warning, a second-level early warning, and a third-level early warning.

[0019] In one possible implementation of this application, the step of converting the apparent value of methane concentration to obtain the true value of methane concentration specifically includes:

[0020] Generate an experimental matrix, which includes numerical factors and categorical factors;

[0021] Response surface methodology is performed based on different factors in the experimental matrix to obtain the conversion coefficients.

[0022] The apparent value of methane concentration is multiplied by the conversion coefficient to complete the data conversion and obtain the true value of methane concentration.

[0023] In one possible implementation of this application, the training process of the prediction model includes:

[0024] Historical data is acquired and preprocessed to obtain training samples, which are then divided into a training set and a validation set according to a preset ratio.

[0025] The initial long short-term memory network model is trained based on the training set, and the training results are validated using the validation set. The training phase also includes a gradient pruning mechanism.

[0026] The root mean square error of the predicted concentration value sequence and the actual methane concentration in the validation set is evaluated to complete the model training and obtain the prediction model.

[0027] In one possible implementation of this application, the step of performing a threshold judgment based on the prediction result to output a warning signal, the warning signal including a first-level warning, a second-level warning, and a third-level warning, specifically including:

[0028] A preset concentration threshold is obtained, wherein the preset concentration threshold includes a first threshold range, a second threshold range, and a third threshold range;

[0029] The methane concentration value is compared based on the predicted result and the preset concentration threshold, wherein...

[0030] If the predicted methane concentration value is within the first threshold range, a first-level warning is output, including a yellow alert signal;

[0031] If the predicted methane concentration value is within the second threshold range, a level two warning is output, including an orange alert signal;

[0032] If the predicted methane concentration value is within the third threshold range, a level three warning is output, including a red alert signal.

[0033] Thirdly, the present invention provides a real-time online monitoring and early warning system for methane gas concentration, the system comprising:

[0034] An acquisition module is used to acquire environmental data, including apparent methane concentration and other sensor data.

[0035] The conversion module is used to convert the apparent value of methane concentration to obtain the true value of methane concentration.

[0036] The prediction module is used to input the true value of methane concentration and other sensor data into a trained prediction model to obtain the prediction result;

[0037] The output module is used to perform threshold judgment based on the prediction results to output early warning signals, including level one early warning, level two early warning and level three early warning.

[0038] Fourthly, the present invention provides an electronic device, the electronic device comprising: a processor and a memory;

[0039] The memory is used to store computer programs;

[0040] The processor is used to execute the computer program stored in the memory, so that the electronic device performs the above-described method for real-time online monitoring and early warning of methane gas concentration.

[0041] Fifthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by an electronic device, implements the above-described method for real-time online monitoring and early warning of methane gas concentration.

[0042] As described above, the methane gas concentration real-time online monitoring and early warning device, method, system, and medium of the present invention can continuously monitor and warn of methane gas in real time, integrate multi-parameter correction, trend prediction, and intelligent response. By introducing multiple environmental parameter variables to correct experimental data, the accuracy of the true methane value is effectively improved. Combined with the prediction model, it can effectively detect abnormal changes in methane concentration in advance, improving the accuracy and timeliness of the early warning. At the same time, the real-time online monitoring and early warning device has the characteristics of high protection, high sensitivity, and low energy consumption, making it suitable for large-scale municipal drainage pipeline network deployment and ensuring municipal public safety. Attached Figure Description

[0043] Figure 1 The diagram shown is a structural schematic of one embodiment of the methane gas concentration real-time online monitoring and early warning device of the present invention.

[0044] Figure 2 The diagram shown is an installation schematic of the methane gas concentration real-time online monitoring and early warning device of the present invention in one embodiment;

[0045] Figure 3 The diagram shows a step-by-step illustration of the real-time online monitoring and early warning method for methane gas concentration according to an embodiment of the present invention.

[0046] Figure 4 The diagram shown is a flowchart of an embodiment of the real-time online monitoring and early warning method for methane gas concentration of the present invention.

[0047] Figure 5 The diagram shows a flowchart illustrating the output of an early warning signal in one embodiment of the real-time online monitoring and early warning method for methane gas concentration according to the present invention.

[0048] Figure 6The diagram shown is a structural schematic of the real-time online monitoring and early warning system for methane gas concentration of the present invention in one embodiment.

[0049] Figure 7 The diagram shown is a structural schematic of an electronic device according to an embodiment of the present invention.

[0050] Component designation explanation

[0051] 12 Equipment Body

[0052] 121 handle

[0053] 122 Openable lid

[0054] 13 Methane Sensor

[0055] 131 Methane sensor chamber

[0056] 14 Temperature and humidity sensors

[0057] 141 Temperature and humidity sensor probe

[0058] 15 Liquid level sensor

[0059] 151 Ultrasonic transmitting probe

[0060] 16 Wind speed sensors

[0061] 161 Wind speed sensing component

[0062] 17 Power Supply Compartment

[0063] 18 controllers

[0064] 19 Power connection cable

[0065] 20 well rings

[0066] 21 Manhole Covers

[0067] 22 Fixed rod

[0068] 23. Fixing ring

[0069] 24 Fixing Plate

[0070] 25 Support Plate Detailed Implementation

[0071] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0072] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0073] Furthermore, the use of terms such as "first" and "second" in this application is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. If the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed in this application.

[0074] Specifically, invention patent (authorization announcement number: CN118038642B) discloses a mining digital gas parameter early warning system and method based on gas analysis. Invention patent (authorization announcement number: CN115019482B) discloses a dangerous gas patrol early warning method and device based on big data. Invention patent (authorization announcement number: CN111311882B) discloses a leak gas information feedback and alarm warning device that combines data transmission with the Internet of Things. However, the above patents are limited to data monitoring of combustible gases, lacking targeted monitoring of methane gas and prediction of related generation trends, and the environmental parameters are relatively simple. The research of Meng Qinglong et al. (2022) focuses on the real-time monitoring of harmful gas concentrations in municipal sewage pipes, which lacks specificity and has a weak ability to perceive the generation law and concentration change trend of the monitored gas in advance, and lacks an effective time series prediction mechanism.

[0075] This application proposes a real-time online monitoring and early warning device, method, system, and medium for methane gas concentration. It can be applied to municipal drainage pipelines. The device uses a real-time online monitoring and early warning system to output the apparent methane gas concentration data obtained from real-time monitoring by a sensor array, and then uses a conversion coefficient determined by response surface methodology to output the true methane concentration value. This value is combined with data from other sensors and input into a trained long short-term memory network prediction model to obtain a prediction result. Based on this prediction result, real-time early warnings can be issued. Specifically, threshold distribution judgment is performed to output corresponding early warning signals, helping municipal management personnel to respond and take immediate action.

[0076] Specifically, the technical solutions in the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0077] like Figure 1 As shown, in one embodiment of the invention, the methane gas concentration real-time online monitoring and early warning device of the present invention includes:

[0078] The equipment body contains a sensor array, a controller, and a power supply.

[0079] The sensor group includes a methane sensor, a temperature and humidity sensor, a liquid level sensor, and a wind speed sensor, used to collect environmental data inside the pipeline;

[0080] The controller includes a processing unit and a communication unit. The processing unit is used to receive environmental data, process it, and generate an early warning signal. The communication unit is used to communicate the early warning signal to the outside world.

[0081] The power supply includes a power supply, a power compartment, and a power connection cable. The power supply is used to supply power to the sensor group and the controller.

[0082] It should be noted that, in this embodiment, as Figure 1 The diagram shows a real-time online monitoring and early warning device, which can be applied to real-time online monitoring and early warning of methane gas concentration in municipal drainage pipelines. It can also be applied to other embodiments, such as the safety management of underground sewage pipe networks, inspection wells, and combined sewer systems. Specifically, the real-time online monitoring and early warning device includes a device body 12, a sensor group (not shown), a controller 18, and a power supply (not shown). The device body 12 includes a box with an openable cover 122, and a handle 121 is installed on one side of the box. The box is made of engineering plastics, such as ABS plastic and polycarbonate, with an IP68 protection rating, suitable for humid and highly corrosive environments. In actual deployment, it is installed at a vertical height of four-fifths of the way up from the bottom of the inspection well in the municipal drainage pipeline to meet the requirements of (CJ / T 307-2009 "Gas Detection Standard for Drainage Facilities").

[0083] Furthermore, in this embodiment, the sensor group includes a methane sensor 13, a temperature and humidity sensor 14, a liquid level sensor 15, and a wind speed sensor 16, as shown below. Figure 1 As shown, methane sensor 13 corresponds to methane sensor chamber 131, temperature and humidity sensor 14 corresponds to temperature and humidity sensor probe 141, liquid level sensor 15 corresponds to ultrasonic transmitting probe 151, and wind speed sensor 16 corresponds to wind speed sensing component 161. Specifically, the methane gas sensor supports NB-IoT communication and I²C interface, adopts the non-dispersive infrared (NDIR) principle, and periodically detects the apparent concentration of gaseous methane in the drainage pipe every three hours. Temperature and humidity sensor 14 is a digital temperature and humidity sensor 14, supports NB-IoT communication and I²C interface, has high precision and corrosion-resistant packaging, and is suitable for high humidity and harsh environments. Ultrasonic wind speed sensor 16 is used to monitor the ventilation of the pipeline, supports NB-IoT communication and I²C interface, has an IP68 protection rating, and can also work for a long time in the high humidity environment inside the pipeline.

[0084] Furthermore, in this embodiment, the controller 18 includes a microcontroller unit integrating a processing unit and a communication unit. The processing unit is used to receive environmental data and process it to generate an early warning signal. The communication unit is used to communicate the early warning signal to the outside world. Specifically, the processing unit uses an STM32L series MCU (Microcontroller Unit) to read sensor data and complete data analysis and processing steps, including converting the apparent value of methane concentration into the true value of methane concentration, activating a trained prediction model to predict the data, and determining the early warning response steps to generate an early warning signal based on the prediction results. The communication unit uses NB-IoT communication to transmit the processed early warning signal to the background management center.

[0085] Furthermore, in this embodiment, the power supply includes a power supply, a power compartment 17, and a power connection line 19. The power supply is used to supply power to the sensor group and the controller 18. The power supply uses multiple ER14505 lithium batteries connected in series, but other power supply batteries can also be used, and there is no limitation. Correspondingly, the power compartment 17 is used to hold multiple power supply lithium batteries, and the power connection line 19 is used to connect the power compartment 17 and the batteries in series, as well as to connect the power supply and different sensors and the controller 18 to supply power.

[0086] Furthermore, in one embodiment of the invention, the device body 12 is placed on a support plate 25 and is limited by at least two fixing plates 24. The support plate 25 and the fixing plates 24 are both fixed to a fixing rod 22 by fixing rings 23. The fixing rod 22 is fixedly connected to the manhole ring 20 of the pipeline inspection well. The materials of the support plate 25, the fixing plate 24 and the fixing rod 22 include metal.

[0087] It should be noted that, in this embodiment, as Figure 2 The diagram shown illustrates the installation and application of a real-time online monitoring and early warning device. This device is installed within municipal drainage pipes for monitoring and early warning purposes. Figure 2 As shown, a fixing rod 22 is fixedly connected to the manhole ring 20 of the pipeline inspection well, for example, by a threaded connection. Correspondingly, in practical applications, a manhole cover 21 is also included on the manhole ring 20. Furthermore, two fixing rings 23 are fixedly connected to the end of the fixing rod 22 away from the manhole cover 21. One fixing ring 23 is fixedly connected to a support plate 25, and the other fixing ring 23 is fixedly connected to several fixing plates 24. The device body 12 is placed on the support plate 25, and is limited by the mutual abutment of at least two fixing plates 24, so that the real-time online monitoring and early warning device can stably monitor and warn of the methane concentration in the municipal drainage pipeline. The support plate 25, the fixing plates 24 and the fixing rod 22 are made of metal with an anti-corrosion coating on the surface, which can be suitable for humid and highly corrosive environments while providing support.

[0088] like Figure 3 As shown, in one embodiment of the invention, the real-time online monitoring and early warning method for methane gas concentration of the present invention is applied to any of the real-time online monitoring and early warning devices for methane gas concentration described in any one of the claims, wherein the method includes the following steps:

[0089] Step S302: Obtain environmental data, including apparent methane concentration and other sensor data;

[0090] Step S304: Convert the apparent value of methane concentration to obtain the true value of methane concentration.

[0091] Step S306: Based on the true value of methane concentration and the other sensor data, input them into the trained prediction model to make a prediction and obtain the prediction result;

[0092] Step S308: Based on the prediction results, a threshold judgment is performed to output a warning signal, which includes a first-level warning, a second-level warning, and a third-level warning.

[0093] It should be noted that, in this embodiment, when applied to the municipal drainage pipeline application scenario, environmental data is first acquired through the sensor group of the real-time online monitoring and early warning device installed in the pipeline, including the apparent value of methane concentration and other sensor data, such as temperature data, humidity data, water level data, and wind speed data.

[0094] Furthermore, in this embodiment, the apparent methane concentration of methane gas obtained from real-time monitoring is converted into the true methane concentration using a conversion coefficient determined by response surface methodology. The specific response surface methodology process will be described in detail in the subsequent specification. Then, based on the true methane concentration and the other sensor data, the data is input into a trained prediction model to obtain the prediction result. Correspondingly, a deep learning neural network model is constructed based on the historical monitoring data of the sensor device and the experimental environment data and trained to obtain a trained prediction model. The specific training process will also be described in detail in the subsequent specification.

[0095] Furthermore, in this embodiment, the prediction results of methane gas concentration are combined with the set threshold to determine whether to issue an early warning. The early warning signals corresponding to different threshold ranges are different, specifically including level one, level two, and level three early warnings. By establishing an intelligent and automated multi-level early warning response mechanism, the accuracy and efficiency of monitoring and early warning of hazardous gases in municipal drainage pipelines can be improved, filling the gap in the field of early warning of hazardous gases in municipal drainage pipelines.

[0096] Furthermore, in one embodiment of the invention, the step of converting the apparent value of methane concentration to obtain the true value of methane concentration specifically includes:

[0097] Generate an experimental matrix, which includes numerical factors and categorical factors;

[0098] Response surface methodology is performed based on different factors in the experimental matrix to obtain the conversion coefficients.

[0099] The apparent value of methane concentration is multiplied by the conversion coefficient to complete the data conversion and obtain the true value of methane concentration.

[0100] It should be noted that, in this embodiment, the functional relationship between the actual methane concentration and the apparent methane concentration is as follows: ,in, This represents the actual methane concentration, in mg / L. This represents the apparent methane concentration in mg / L, and K is the conversion factor between the actual and apparent methane concentrations.

[0101] Furthermore, specifically, the experimental matrix was generated using the CustomDesign method in Design Expert 13 software. The factors and levels were set as numerical factors: temperature (15℃, 25℃, 35℃), humidity (70%, 80%, 90%), and wind speed (0.1 m / s, 0.3 m / s, 0.5 m / s); categorical factors included water level (high water level > 10 cm coded as 1, low water level ≤ 10 cm coded as 0). After generating the experimental matrix, experimental data were input. The response variable was set as the ratio of the actual methane value to the apparent methane value. Response surface methodology (RSM) analysis was then performed within the software. The built-in ANOVA tool was used to test the statistical significance of the model, combined with the coefficient of determination R-squared. 2 The indicators evaluate the goodness of fit of the model and further confirm its effectiveness. After confirming effectiveness, the function expression is output through "equation". The regular expression is as follows: , It is the response variable, which in this embodiment corresponds to the transformation coefficient K. It is the first One influencing factor; These are model coefficients; The experimental error is calculated by multiplying the apparent value of methane concentration with the conversion coefficients after performing response surface analysis based on different factors in the experimental matrix to obtain the true value of methane concentration.

[0102] Furthermore, in one embodiment of the invention, the training process of the prediction model includes:

[0103] Historical data is acquired and preprocessed to obtain training samples, which are then divided into a training set and a validation set according to a preset ratio.

[0104] The initial long short-term memory network model is trained based on the training set, and the training results are validated using the validation set. The training phase also includes a gradient pruning mechanism.

[0105] The root mean square error of the predicted concentration value sequence and the actual methane concentration in the validation set is evaluated to complete the model training and obtain the prediction model.

[0106] It should be noted that, in this embodiment, as Figure 4As shown, historical data is acquired and preprocessed to obtain training samples. In practical applications, real-time monitoring data also needs to be preprocessed. Furthermore, in this embodiment, data preprocessing specifically involves cleaning, identifying outliers, and smoothing the actual methane concentration value and other sensor data (such as temperature, humidity, wind speed, and water level environmental parameters) to ensure the quality of the input data and the stability of the model's prediction results. The process specifically includes:

[0107] (1) Data alignment: All types of monitoring data are aligned according to a unified time benchmark to ensure that the apparent value of methane concentration corresponds to the environmental parameters at each sampling time.

[0108] (2) Data cleaning: For abnormal fluctuations in the monitoring data, a multi-stage cleaning strategy is adopted, including ordinary outliers and extreme outliers. Ordinary outliers: For abnormal fluctuations in the monitoring data, a multi-stage cleaning strategy is adopted, that is, filtering out physically unacceptable outliers, such as negative methane concentration or water level exceeding the structural limit of drainage pipes; while extreme outliers: Extreme outliers in terms of statistical validity (extreme anomalies in sensor data) are identified and cleaned. The Z-score method and the interquartile range (IQR) method are introduced for processing. Specifically, (a) For variable data (temperature, humidity) that approximately follow a normal distribution, the Z-score standardization coefficient is used for determination, where Z-score is defined as: ,in, The sample value is μ; μ is the mean of the dataset. This represents the standard deviation of the dataset. This indicates the distance (in standard deviations) between the sample value and the mean. This indicates that the sample deviates from the mean by more than 3 standard deviations, and is considered an outlier. For non-normally distributed or skewed data (such as wind speed, water level, and methane concentration), the IQR upper and lower limit rules are used to identify abnormal fluctuations. and x is considered an outlier, where, Q1 is the lower quartile of the dataset, Q3 is the upper quartile of the dataset, and IQR is the interquartile range, representing the middle "50%" range of the dataset.

[0109] (3) Data smoothing: Specifically, outliers from the cleaned data are set as missing, and the data is smoothed using linear interpolation. The formula for linear interpolation is: In the formula, This is the current anomaly point; This is the previous normal value; The next normal value; i is the current abnormal index; This is the index of the previous normal point; This is the index for the next normal point.

[0110] Furthermore, in this embodiment, the time series sample construction uses historical monitoring data within "3" months. The data is imported into the library, and Python 3.9 software is used to construct the input samples of the LSTM (Long Short-Term Memory) model from the output data of step two according to the time series using a sliding window mechanism. The continuous environmental parameter data is constructed into samples according to the time window length, i.e., the time step T. The time window length T is selected as ten time steps, and each time step corresponds to sampling once every three hours. Each sample corresponds to the monitoring data of the past "30" hours. Thus, the environmental parameter data and methane concentration data of each T consecutive time step are used as input features, and the methane concentration value corresponding to time T+1 is used as the label output to construct the time series training samples.

[0111] Furthermore, in this embodiment, based on the constructed training samples, the first 80% of the dataset is divided into a training set, and the last 20% is divided into a validation set. An LSTM sequence prediction model is then established. The model structure includes at least one LSTM layer and an output fully connected layer. The LSTM layer contains four core structures: a forget gate, an input gate, an output gate, and a memory unit. The forget gate is used to determine and control whether the memory state of the previous time step is retained at the current time step. The formula is: ,in, Indicates the output of the forget gate. This is the hidden state from the previous moment. For the current input feature value, This is the weight matrix. For bias terms, This represents the Sigmoid activation function; the input gate determines the degree to which new input information at the current moment affects the current memory state, including the input gate control variable: and candidate memory states: In the formula, For input gate output, Candidate state This is the weight matrix. For bias terms, The hyperbolic tangent function is used to perform nonlinear fitting of the memory state; the memory unit updates its state by combining information controlled by the forget gate and the input gate, and the expression is: In the formula, The state is the memory state at time t; the output gate outputs the current state, expressed by the formula: , ,in, For output gate output, As the final hidden state, during training, the model automatically learns the temporal dependencies and parameter coupling effects in the input sequence, handles the temporal correlation of the sequence data, and uses fully connected layers to process the features output by the LSTM. The model maps the values ​​to predicted methane concentrations; it uses mean squared error (MSE) as the loss function. In the formula, This represents the actual methane concentration at time point i. is the methane concentration predicted by the LSTM model at the i-th time point; n is the number of evaluation samples; the Adam adaptive gradient optimizer is used for parameter learning; a validation set is introduced during training for model fitting and validation, and an early stopping strategy is used to avoid overfitting. If the validation error does not decrease for several consecutive rounds, training is stopped early.

[0112] Furthermore, in this embodiment, the training phase also includes a gradient clipping mechanism, that is, a gradient clipping mechanism is introduced in the LSTM model training phase. When the gradient value exceeds a set threshold (5.0) during the backpropagation of the model, the gradient size is automatically limited to prevent gradient explosion caused by differences in different input dimensions, thereby overcoming the problem of data errors caused by lack of normalization.

[0113] Furthermore, in this embodiment, the predicted concentration value sequence and the actual methane concentration in the validation set are evaluated using root mean square error (RMSE) to complete model training and obtain the prediction model. Specifically, the trained LSTM model is used to predict the corresponding input features in the validation set to obtain a predicted concentration value sequence, which corresponds to the actual methane concentration in the validation set. The RMSE is then evaluated. In the formula, This represents the actual methane concentration at time point i. is the predicted methane concentration value at the i-th time point corresponding to the LSTM model; n is the number of evaluation samples.

[0114] Furthermore, in one embodiment of the invention, the step of performing a threshold judgment based on the prediction result to output a warning signal, the warning signal including a first-level warning, a second-level warning, and a third-level warning, specifically includes:

[0115] A preset concentration threshold is obtained, wherein the preset concentration threshold includes a first threshold range, a second threshold range, and a third threshold range;

[0116] The methane concentration value is compared based on the predicted result and the preset concentration threshold, wherein...

[0117] If the predicted methane concentration value is within the first threshold range, a first-level warning is output, including a yellow alert signal;

[0118] If the predicted methane concentration value is within the second threshold range, a level two warning is output, including an orange alert signal;

[0119] If the predicted methane concentration value is within the third threshold range, a level three warning is output, including a red alert signal.

[0120] It should be noted that, in this embodiment, the latest T-step environmental parameters acquired in real time are input into the trained prediction model to predict the methane concentration at time T+1. Continuous rolling prediction is performed and the prediction sequence is recorded. If the methane concentration value in the prediction result exceeds the methane safe concentration threshold (10,000 ppm), a trend warning response mechanism is triggered, which then links the alarm device or control equipment to intervene. The prediction model adopts a rolling prediction mechanism, updating the input window and recalculating the future concentration every "3" hours based on the latest data, thus forming a continuous prediction capability.

[0121] Furthermore, in this embodiment, as Figure 5 The diagram illustrates the process of outputting early warning signals. The predicted methane concentration is compared to a preset concentration threshold. Risk grading is performed based on the distribution of the predicted concentration across different threshold ranges. If the predicted methane concentration falls within the first threshold range (10,000~15,000 ppm, excluding 15,000 ppm), a Level 1 warning is output, including a yellow alert signal, which is sent to the backend management center, suggesting manual review. If the predicted methane concentration falls within the second threshold range (15,000~18,000 ppm, excluding 18,000 ppm), a Level 2 warning is output, including an orange alarm signal, which is sent to the backend management center. If the predicted methane concentration falls within the third threshold range (18,000~20,000 ppm), a Level 3 warning is output, including a red alert signal, which is sent to the backend management center, prompting management personnel to respond promptly.

[0122] Please see Figure 6 In one embodiment, this embodiment provides a real-time online monitoring and early warning system 60 for methane gas concentration, the system comprising:

[0123] Acquisition module 61 is used to acquire environmental data, including apparent methane concentration and other sensor data;

[0124] Conversion module 62 is used to convert the apparent value of methane concentration to obtain the true value of methane concentration.

[0125] Prediction module 63 is used to input the true value of methane concentration and other sensor data into a trained prediction model to make a prediction result.

[0126] The output module 64 is used to perform threshold judgment based on the prediction result to output a warning signal, which includes a first-level warning, a second-level warning and a third-level warning.

[0127] Since the specific implementation of this embodiment corresponds to the aforementioned method embodiment, the same details will not be repeated here, and those skilled in the art should also understand this. Figure 6 The division of the modules in the embodiments is only a logical functional division. In actual implementation, they can be fully or partially integrated into one or more physical entities. These modules can be fully implemented in software through processing element calls, fully implemented in hardware, or some modules can be implemented in software through processing element calls and some modules can be implemented in hardware.

[0128] In the embodiments provided by this invention, it should be understood that the disclosed apparatus or method can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of modules / units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or units may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of the apparatus or module or unit may be electrical, mechanical, or other forms.

[0129] The modules / units described as separate components may or may not be physically separate. The components shown as modules / units may or may not be physical modules; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules / units can be selected to achieve the objectives of the embodiments of the present invention, depending on actual needs. For example, the functional modules / units in the various embodiments of the present invention may be integrated into one processing module, or each module / unit may exist physically separately, or two or more modules / units may be integrated into one module / unit.

[0130] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0131] This invention also provides an electronic device, such as... Figure 7 As shown, the electronic device includes a processor and a memory.

[0132] The memory is used to store computer programs.

[0133] The processor is configured to execute a computer program stored in the memory to cause the electronic device to perform any of the methods described above.

[0134] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, or methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of modules / units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or units may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of apparatuses or modules or units may be electrical, mechanical, or other forms.

[0135] The modules / units described as separate components may or may not be physically separate. The components shown as modules / units may or may not be physical modules; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules / units can be selected to achieve the objectives of the embodiments of this application, depending on actual needs. For example, the functional modules / units in the various embodiments of this application may be integrated into one processing module, or each module / unit may exist physically separately, or two or more modules / units may be integrated into one module / unit.

[0136] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0137] This application also provides a computer-readable storage medium. Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing a processor. The program can be stored in a computer-readable storage medium, which is a non-transitory medium, such as random access memory, read-only memory, flash memory, hard disk, solid-state drive, magnetic tape, floppy disk, optical disk, and any combination thereof. The storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., digital video disc (DVD)), or a semiconductor medium (e.g., solid-state drive (SSD)).

[0138] This application embodiment may also provide a computer program product comprising one or more computer instructions. When the computer instructions are loaded and executed on a computing device, all or part of the processes or functions described in this application embodiment are generated. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means.

[0139] When the computer program product is executed by a computer, the computer performs the method described in the foregoing method embodiments. The computer program product can be a software installation package; when the foregoing method is required, the computer program product can be downloaded and executed on the computer.

[0140] The descriptions of the processes or structures corresponding to the above figures each have their own emphasis. For parts of a process or structure that are not described in detail, please refer to the relevant descriptions of other processes or structures.

[0141] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in this application should still be covered by the claims of this application.

Claims

1. A method for real-time online monitoring and early warning of methane gas concentration, characterized in that, The method includes the following steps: Acquire environmental data, including apparent methane concentration and other sensor data; The apparent methane concentration is transformed to obtain the true methane concentration. Specifically, this includes: generating an experimental matrix containing numerical factors and categorical factors. The numerical factors include temperature, humidity, and wind speed, while the categorical factors include water level. Response surface methodology is then performed based on the different factors in the experimental matrix to obtain transformation coefficients. The conventional expression for this transformation is: Y = β0 + ∑β i X i +∑β ii X i 2 +∑β ij X i X j +ε, where Y is the response variable and K is the corresponding transformation coefficient. i It is the i-th influence factor; β0, β i β ii β ij ε is the model coefficient; ε is the experimental error; the data conversion is completed by multiplying the apparent value of methane concentration with the conversion coefficient to obtain the true value of methane concentration; The prediction results are obtained by inputting the actual methane concentration value and other sensor data into a trained prediction model. The training process of the prediction model includes: acquiring historical data and performing data preprocessing to obtain training samples; dividing the training samples into a training set and a validation set according to a preset ratio; training an initial long short-term memory network model based on the training set; validating each training result using the validation set; and performing a gradient pruning mechanism during the training phase. The root mean square error is then evaluated on the predicted concentration value sequence and the actual methane concentration in the validation set to complete the model training and obtain the prediction model. Based on the prediction results, a threshold judgment is performed to output an early warning signal, which includes a first-level early warning, a second-level early warning, and a third-level early warning.

2. The method for real-time online monitoring and early warning of methane gas concentration according to claim 1, characterized in that, The threshold judgment based on the prediction result is used to output a warning signal, which includes a first-level warning, a second-level warning, and a third-level warning, specifically including: A preset concentration threshold is obtained, wherein the preset concentration threshold includes a first threshold range, a second threshold range, and a third threshold range; The methane concentration value is compared based on the predicted result and the preset concentration threshold, wherein... If the predicted methane concentration value is within the first threshold range, a first-level warning is output, including a yellow alert signal; If the predicted methane concentration value is within the second threshold range, a level two warning is output, including an orange alert signal; If the predicted methane concentration value is within the third threshold range, a level three warning is output, including a red alert signal.

3. A real-time online monitoring and early warning system for methane gas concentration, characterized in that, include: An acquisition module is used to acquire environmental data, including apparent methane concentration and other sensor data. The conversion module is used to convert the apparent value of methane concentration to obtain the true value of methane concentration. Specifically, this includes: generating an experimental matrix, which includes numerical factors and categorical factors. The numerical factors include temperature, humidity, and wind speed, while the categorical factors include water level; and performing response surface analysis based on the different factors in the experimental matrix to obtain conversion coefficients. The conventional expression is: Y = β0 + ∑β i X i +∑β ii X i 2 +∑β ij X i X j +ε, where Y is the response variable and K is the corresponding transformation coefficient. i It is the i-th influence factor; β0, β i β ii β ij ε is the model coefficient; ε is the experimental error; the data conversion is completed by multiplying the apparent value of methane concentration with the conversion coefficient to obtain the true value of methane concentration; The prediction module is used to input the actual methane concentration and other sensor data into a trained prediction model to obtain a prediction result. The training process of the prediction model includes: acquiring historical data and performing data preprocessing to obtain training samples; dividing the training samples into a training set and a validation set according to a preset ratio; training an initial long short-term memory network model based on the training set; validating each training result using the validation set; the training phase also includes a gradient pruning mechanism; and performing root mean square error evaluation on the predicted concentration value sequence and the actual methane concentration in the validation set to complete model training and obtain the prediction model. The output module is used to perform threshold judgment based on the prediction results to output early warning signals, including level one early warning, level two early warning and level three early warning.

4. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the real-time online monitoring and early warning method for methane gas concentration as described in any one of claims 1 to 2.

5. An electronic device, characterized in that, The electronic device includes a processor and a memory; wherein the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to enable the electronic device to perform the real-time online monitoring and early warning method for methane gas concentration as described in any one of claims 1 to 2.

6. A real-time online monitoring and early warning device for methane gas concentration, characterized in that, The method for real-time online monitoring and early warning of methane gas concentration according to any one of claims 1-2, wherein the equipment includes: The equipment body contains a sensor array, a controller, and a power supply. The sensor group includes a methane sensor, a temperature and humidity sensor, a liquid level sensor, and a wind speed sensor, used to collect environmental data inside the pipeline; The controller includes a processing unit and a communication unit. The processing unit is used to receive environmental data, process it, and generate an early warning signal. The communication unit is used to communicate the early warning signal to the outside world. The power supply includes a power supply, a power compartment, and a power connection cable. The power supply is used to supply power to the sensor group and the controller.

7. The real-time online monitoring and early warning device for methane gas concentration according to claim 6, characterized in that, The device body is placed on a support plate and is limited by at least two fixing plates. The support plate and the fixing plates are both fixed to a fixing rod by fixing rings. The fixing rod is fixedly connected to the manhole ring of the pipeline inspection well. The support plate, the fixing plate and the fixing rod are made of metal.

8. The real-time online monitoring and early warning device for methane gas concentration according to claim 6, characterized in that, The device body includes a box with an openable and closable lid, and a handle is installed on one side of the box. The material of the box includes engineering plastics.

Images