Wireless remote transmission industrial and commercial laser methane alarm and alarm system

Abstract

The invention discloses a wireless remote transmission industrial and commercial laser methane alarm and an alarm system, relates to the technical field of gas leakage alarms, and solves the problem that a sensor in a combustible gas alarm is greatly influenced by cross interference and vibration. The laser methane alarm comprises a light source emitter, an optical module, an illumination detector, a control module, a power supply module, an interface module, an explosion-proof shell and external equipment, the laser methane alarm system comprises a processing unit, an early warning unit, a display unit and a plurality of laser methane alarms, performing gas leakage prediction and early warning on the lossless data blocks formed by preprocessing through a preset leakage prediction model, correcting alarm thresholds of the monitoring areas with different volumes, generating an alarm signal according to the real-time methane concentration data and the corrected alarm thresholds, and performing alarm and ventilation emission through external equipment; the detection efficiency of the laser methane alarm is improved, and the error rate of gas leakage prediction is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of gas leakage alarms, and in particular to a wireless remote transmission industrial and commercial laser methane alarm and an alarm system. Background Art

[0002] With the development of industrialization and urbanization, the number of crowded places such as chemical plants, factories and medical institutions is increasing, and the demand for safety precautions is also increasing. Methane is a flammable gas and there is a risk of leakage. In addition, traditional methane alarms are not easy to move and have problems such as inflexible cable wiring. They are difficult to meet the safety requirements of modern places. Therefore, a more advanced alarm system is needed, and the application of wireless remote transmission technology provides a solution to this demand.

[0003] Patent No. CN2022102427260 discloses an intelligent gas safety linkage control system based on the Internet of Things, including a gas meter and an alarm; the alarm is equipped with a methane sensor and a wireless communication module, and the methane sensor is wirelessly connected to the gas meter through the wireless communication module; the gas meter is equipped with an automatic valve, a valve control module, an Internet of Things module and an LCD display module, and a gas connecting pipe is provided inside the gas meter, and an intake pipe and an exhaust pipe are provided on the left and right sides of the gas connecting pipe, respectively. The automatic valve is installed on the gas connecting pipe and is controlled by the valve control module. The alarm realizes wireless connection with the gas meter and the mobile phone such as Bluetooth / Lora, collects and uploads the data information of the gas user meter to the gas company's smart platform for supervision. When a fault or alarm occurs, the platform can obtain the fault information in the first time, and the meter valve can be remotely shut down through the valve control module to ensure the safety of the product.

[0004] Patent No. CN2017104699540 discloses a natural gas leak detector, which includes: a shell, a methane light detector, an air intake device, a solar cell module, a battery, a wireless transmission device, a controller and an alarm; a solar cell module is arranged on the top of the shell, and a methane light detector, a battery, a wireless transmission device, a controller and an alarm are arranged inside the shell; the air intake device includes a trumpet-shaped air intake port, a wind wheel and a motor; the solar cell module is arranged in an umbrella shape, including an umbrella surface, an umbrella handle and an umbrella rib connecting the umbrella surface and the umbrella handle so that the umbrella surface can be opened and folded, and a solar cell panel is arranged on the umbrella surface; the controller is respectively connected to the methane light detector, the solar cell module, the wireless transmission device, the motor and the alarm. The natural gas leak detector of the above invention can convert solar energy into electrical energy to power it, does not require an external power supply, has high detection sensitivity, simple structure, small size, is easy to carry, and is suitable for long-term outdoor detection.

[0005] Although the combustible gas alarm in the above patent can detect and warn of natural gas in industrial and commercial places, the sensor in the combustible gas alarm is a semiconductor and catalytic element, so the sensor is greatly affected by cross interference and vibration, resulting in an excessively high error rate in predicting gas leaks. Summary of the invention

[0006] The purpose of the present invention is to provide a wireless remote transmission industrial and commercial laser methane alarm and alarm system, which can predict gas leakage based on the collected methane concentration data using a preset leakage prediction model, and then correct the alarm thresholds of monitoring areas of different volumes. Subsequently, based on the real-time methane concentration data and the corrected alarm thresholds, early warning and ventilation emission processing are performed on the monitoring areas where each laser methane alarm is located; so as to improve the detection efficiency of the laser methane alarm and reduce the error rate of gas leakage prediction.

[0007] The present invention utilizes the following technical solutions:

[0008] A wireless remote transmission industrial and commercial laser methane alarm includes a light source transmitter, an optical module, a light detector, a control module, a power module, an interface module, an explosion-proof housing and external equipment; wherein,

[0009] A light source transmitter, used to generate laser light to serve as a detection light source;

[0010] An optical module, used for focusing the detection light source and irradiating it to the light detector;

[0011] A light detector, used to detect light intensity data of a detection light source in a monitoring area;

[0012] The control module is used to calculate the methane concentration data using the spectral absorption technology based on the light intensity data of the detection light source; at the same time, according to the alarm thresholds of the monitoring areas of different volumes corrected in real time, an alarm signal is generated based on the real-time methane concentration data and the corrected alarm thresholds, and an alarm and ventilation emission are performed through external equipment;

[0013] The power module is used to power the laser methane alarm and manage the battery charging and discharging status;

[0014] Interface module, used to transmit methane concentration data, alarm threshold and light intensity data to external devices for display and setting;

[0015] An explosion-proof housing is used to protect the internal components of the laser methane alarm; the internal components include a light source emitter, an optical module, a light detector, a control module, a power module and an interface module;

[0016] External equipment, used for alarm and ventilation discharge according to the alarm signal sent by the control module;

[0017] The power module is connected to the light source emitter, optical module, light detector, control module and interface module respectively through the power management circuit; the light source emitter and light detector are connected through the signal processing circuit, and the control module is connected to the light source emitter, optical module, light detector and interface module respectively through the signal transmission line; each internal component is installed in the explosion-proof casing by screws or buckles.

[0018] Preferably, the external device includes a buzzer, a display screen, a gas valve and an exhaust control fan.

[0019] A wireless remote transmission industrial and commercial laser methane alarm system, comprising a processing unit, an early warning unit, a display unit and a plurality of laser methane alarms; wherein,

[0020] Laser methane alarm, used to obtain the light intensity data of the detection light source in the monitoring area and calculate the corresponding methane concentration data;

[0021] The processing unit is used to pre-process the methane concentration data and light intensity data, obtain lossless data blocks after division, classification, cleaning, inspection, aggregation and compression, and send them to the early warning unit for gas leakage prediction;

[0022] The early warning unit is used to predict the gas leak alarm threshold value based on the pre-processed methane concentration data and light intensity data according to the volume of the monitoring area, and to correct the alarm threshold value according to the prediction result of the gas leak alarm threshold value; and then to warn the monitoring area where each laser methane alarm is located according to the real-time methane concentration data and the corrected alarm threshold value; the gas leak prediction result includes the leak location, leak amount and leak time; the volume of the monitoring area is selected and determined by the assembler according to the installation space of each laser methane alarm;

[0023] The display unit is used to display the working status and methane concentration data of each laser methane alarm and external equipment.

[0024] Preferably, the laser methane alarm obtains the light intensity data of the detection light source in the monitoring area through a light detector; and obtains the methane concentration data calculated by the control module.

[0025] Preferably, the process of preprocessing the methane concentration data and the light intensity data by the processing unit is:

[0026] S1: Divide the methane concentration data and the light intensity data into a number of data blocks according to a preset time interval; the preset time interval may be minutes;

[0027] S2: using a clustering algorithm to classify the internal data of the data block according to the data type; the data type includes methane concentration and light intensity;

[0028] S3: Use data cleaning algorithms to clean and correct abnormal data in the classified internal data; data cleaning algorithms include data deduplication algorithms, data filling algorithms, data conversion algorithms, and data simplification algorithms;

[0029] S4: For the internal data that has completed data cleaning, the methane concentration data and light intensity data of the same collection time are respectively checked;

[0030] S5: According to the data collection time and the volume of the monitoring area, the inspected data blocks are correlated and normalized to construct an aggregated data block;

[0031] S6: Losslessly compressing the aggregated data block using a data compression algorithm to form a corresponding lossless data block;

[0032] S7: The lossless data block is transmitted to the early warning unit through the wireless communication protocol using the parallel heuristic transmission algorithm for gas leakage prediction.

[0033] Preferably, the early warning unit uses a preset leakage prediction model to predict the gas leakage alarm threshold of the lossless data block formed by preprocessing, and corrects the alarm threshold of the monitoring areas of different volumes; the leakage prediction model includes a feature extraction layer, an iterative training layer and a recognition prediction layer; first, the feature extraction layer extracts features from the lossless data block to form a methane concentration-light intensity feature correlation value; then, the iterative training layer performs several iterative training on the methane concentration-light intensity feature correlation value to form a methane-light feature weight matrix; finally, the recognition prediction layer predicts the methane-light feature weight matrix according to the volume of the monitoring area, and corrects the alarm threshold of the monitoring areas of different volumes.

[0034] Preferably, the feature extraction layer uses two 3×3 convolutional layers and one 5×5 convolutional layer to segment the lossless data block to form a methane concentration-light intensity association pair and uses two convolutional blocks consisting of a 3×1 convolutional layer and a 1×3 convolutional layer and two 3×3 average pooling layers to extract features of the methane concentration-light intensity association pair, thereby obtaining a methane concentration-light intensity feature association value:

[0035]

[0036] Among them, Model 0 represents the feature extraction layer, represents two 1×3 convolutional layers with a stride of 1 in series, represents two 3×1 convolutional layers with a stride of 1 in series, represents two 3×3 average pooling layers in series, represents a 3×3 convolutional layer with a stride of 2, represents a 5×5 convolutional layer with a stride of 1, represents a 3×3 convolutional layer with a stride of 1, and M represents a lossless data block.

[0037] Preferably, the iterative training layer uses three iterative branches to learn and train the methane concentration-light intensity feature correlation values ​​several times, and obtains a methane shallow matrix, a light shallow matrix, and a methane-light shallow matrix respectively; and the three shallow matrices are feature-fused to obtain a methane-light feature weight matrix; the iterative training layer includes 5 3×3 convolutional layers, 3 InceptionV1 blocks, 2 InceptionV3 blocks, 2 ReLU activation functions, 1 softmax activation function, 6 3×3 maximum pooling layers, 4 5×5 adaptive pooling layers, and 3 residual blocks:

[0038]

[0039] Among them, Y 1 represents the first iteration branch, Y 2 represents the second iteration branch, Y 3 Represents the third iteration branch, Model 1 represents the iterative training layer, Relu() represents the activation function, AutoPool 5×5 () represents a 5×5 adaptive pooling layer, MaxPool 3×3 2 () represents two 3×3 maximum pooling layers in series, represents two 3×3 convolutional layers with a stride of 1 in series, ResBlock() represents a residual block, Represents the Hadamard operation.

[0040] Preferably, the recognition prediction layer uses 1 shuffle block, 2 1×1 convolutional layers, 2 fully connected layers, 2 5×5 average pooling layers and a feedback optimization function to verify the gas leak prediction result of the methane-illumination feature weight matrix, and correct the alarm thresholds of monitoring areas of different volumes:

[0041]

[0042] Among them, Model 2 represents the recognition prediction layer, L represents the feedback optimization function, MeanPool 5×5 2 () represents two 5×5 average pooling layers in series, Shuff represents the shuffled block, V represents the volume of the monitoring area, G represents the quantization function, σ represents the number of iterative training layers, and E σ represents the optimization efficiency of iterative training for each layer, E0 represents the preset maximum value of iterative training optimization efficiency, τ represents the total number of iterative training layers, μ1 represents the calculation accuracy coefficient, and B σRepresents the iterative training loss function of each layer, and μ2 represents the preset optimization weight coefficient.

[0043]

[0044] Among them, H ′ V It represents the corrected alarm threshold of the monitoring area with different volumes, H V Indicates the current alarm threshold of the monitoring area of ​​different volumes, L V Represents the feedback optimization function corresponding to monitoring areas of different volumes.

[0045] The present invention predicts and warns of gas leakage on lossless data blocks formed by preprocessing through a preset leakage prediction model, corrects the alarm thresholds of monitoring areas of different volumes, generates an alarm signal based on real-time methane concentration data and the corrected alarm thresholds, and performs alarm and ventilation emissions through external equipment; thereby improving the detection efficiency of the laser methane alarm and reducing the error rate of gas leakage prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or related technologies, the drawings required for use in the embodiments or related technical descriptions are briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0047] Figure 1 This is the principle block diagram of the wireless remote transmission industrial and commercial laser methane alarm system;

[0048] Figure 2 The preprocessing flow chart of the processing unit. DETAILED DESCRIPTION

[0049] The present invention is described in detail below with reference to the accompanying drawings and embodiments:

[0050] like Figure 1-Figure 2 As shown, the wireless remote transmission industrial and commercial laser methane alarm described in the present invention includes a light source transmitter, an optical module, a light detector, a control module, a power module, an interface module, an explosion-proof housing and external equipment; wherein,

[0051] A light source transmitter, used to generate laser light to serve as a detection light source;

[0052] An optical module, used for focusing the detection light source and irradiating it to the light detector;

[0053] A light detector, used to detect light intensity data of a detection light source in a monitoring area;

[0054] The control module is used to calculate the methane concentration data using the spectral absorption technology based on the light intensity data of the detection light source; at the same time, according to the alarm thresholds of the monitoring areas of different volumes corrected in real time, an alarm signal is generated based on the real-time methane concentration data and the corrected alarm thresholds, and an alarm and ventilation emission are performed through external equipment;

[0055] The power module is used to power the laser methane alarm and manage the battery charging and discharging status;

[0056] Interface module, used to transmit methane concentration data, alarm threshold and light intensity data to external devices for display and setting;

[0057] An explosion-proof housing is used to protect the internal components of the laser methane alarm; the internal components include a light source emitter, an optical module, a light detector, a control module, a power module and an interface module;

[0058] External equipment, used for alarm and ventilation discharge according to the alarm signal sent by the control module;

[0059] The power module is connected to the light source emitter, optical module, light detector, control module and interface module respectively through the power management circuit; the light source emitter (such as a laser diode) is installed in the optical module through a fixing device; the connection between the optical module and the light detector (such as a photodiode or a phototransistor) can be through an optical path, that is, the laser beam passes through or reflects through a certain optical system to reach the light detector; the light source emitter and the light detector are connected through a signal processing circuit, and the control module is connected to the light source emitter, optical module, light detector and interface module respectively through a signal processing circuit; each internal component is installed in an explosion-proof housing by screws or buckles; the control module includes a microcontroller, a logic circuit and a signal processing circuit for processing the light intensity signal output by the detector; the interface module can be connected to the explosion-proof housing by a cable or wirelessly;

[0060] In the present invention, the light detector includes a detection chamber and a photoelectric detector;

[0061] In the present invention, the light source emitter, optical module, light detector, power module, interface module, and explosion-proof housing are all conventional technologies in the art and will not be described in detail here.

[0062] When the wireless remote transmission industrial and commercial laser methane alarm described in the present invention is in use, a light source transmitter such as a laser diode emits a laser beam, which is focused and transmitted to the detection chamber of the light detector through an optical module, and then the beam passes through the air containing methane; methane molecules absorb the energy in the beam, causing the intensity of light passing through the detection chamber to change. The photoelectric detector of the light detector measures the light intensity data of the transmitted light beam and converts the light intensity data into an electrical signal. The control module processes the electrical signal to calculate the methane concentration and triggers an alarm when necessary;

[0063] The principle of using spectral absorption technology to calculate methane concentration data is as follows: after methane molecules absorb the energy of the laser beam, they change the direction of light propagation; the light beam passing through the detection chamber is guided to a spectrometer using optical fibers or light guides, and the spectrometer analyzes the wavelength changes of the light that has absorbed energy; the spectrometer converts the absorption data into electrical signals, and the intensity of the electrical signals is proportional to the concentration of methane gas; the control module processes the electrical signals to calculate the methane concentration.

[0064] The external device is used to adjust the working state according to the alarm signal sent by the control module: when the real-time methane concentration is greater than the corrected alarm threshold, the external device receives the alarm signal sent by the control module and performs alarm and ventilation discharge;

[0065] In the present invention, the external equipment includes a buzzer, a display screen, a gas valve and an exhaust control fan;

[0066] A wireless remote transmission industrial and commercial laser methane alarm system, comprising a processing unit, an early warning unit, a display unit and a plurality of laser methane alarms; wherein,

[0067] Laser methane alarm, used to obtain the light intensity data of the detection light source in the monitoring area and calculate the corresponding methane concentration data;

[0068] In the present invention, the laser methane alarm obtains the light intensity data of the detection light source in the monitoring area through the light detector; and calculates the corresponding methane concentration data through the control module;

[0069] In the present invention, the principle of using spectral absorption technology to measure methane concentration data is as follows: after methane molecules absorb the energy of a laser beam, the propagation direction of the light will be changed; the light beam passing through the detection chamber is guided to a spectrometer by using an optical fiber or a light guide, and the spectrometer analyzes the wavelength change of the light that has absorbed the energy; the spectrometer converts the absorbed energy data into an electrical signal, and the intensity of the electrical signal is proportional to the concentration of the methane gas; the control module processes the electrical signal to calculate the methane concentration.

[0070] The processing unit is used to pre-process the methane concentration data and light intensity data, obtain lossless data blocks after division, classification, cleaning, inspection, aggregation and compression, and send them to the early warning unit for gas leakage prediction;

[0071] In the present invention, the process of preprocessing the methane concentration data and the light intensity data by the processing unit is as follows:

[0072] S1: Divide the methane concentration data and the light intensity data into a number of data blocks according to a preset time interval; the preset time interval may be minutes;

[0073] S2: using a clustering algorithm to classify the internal data of the data block according to the data type; the data type includes methane concentration and light intensity;

[0074] S3: Use data cleaning algorithms to clean and correct abnormal data in the classified internal data; data cleaning algorithms include data deduplication algorithms, data filling algorithms, data conversion algorithms, and data simplification algorithms;

[0075] S4: For the internal data that has completed data cleaning, the methane concentration data and light intensity data of the same collection time are respectively checked;

[0076] S5: According to the data collection time and the volume of the monitoring area, the inspected data blocks are correlated and normalized to construct an aggregated data block;

[0077] S6: Losslessly compressing the aggregated data block using a data compression algorithm to form a corresponding lossless data block;

[0078] S7: using a parallel heuristic transmission algorithm, the lossless data block is transmitted to the early warning unit through a wireless communication protocol to predict gas leakage;

[0079] The early warning unit is used to predict the gas leak alarm threshold value based on the pre-processed methane concentration data and light intensity data according to the volume of the monitoring area, and to correct the alarm threshold value according to the prediction result of the gas leak alarm threshold value; and then to warn the monitoring area where each laser methane alarm is located according to the real-time methane concentration data and the corrected alarm threshold value; the gas leak prediction result includes the leak location, leak amount and leak time; the volume of the monitoring area is selected and determined by the assembler according to the installation space of each laser methane alarm;

[0080] In the present invention, the early warning unit uses a preset leakage prediction model to predict the gas leakage alarm threshold of the lossless data block formed by preprocessing, and corrects the alarm threshold of the laser methane alarm under the monitoring area of ​​different volumes; the leakage prediction model includes a feature extraction layer, an iterative training layer and an identification prediction layer;

[0081] First, the feature extraction layer extracts features from the lossless data block to form a methane concentration-light intensity feature correlation value;

[0082] In this embodiment, the feature extraction layer uses two 3×3 convolutional layers and one 5×5 convolutional layer to segment the lossless data block to form a methane concentration-light intensity association pair, and the methane concentration-light intensity association pair is subjected to feature extraction using two convolutional blocks consisting of a 3×1 convolutional layer and a 1×3 convolutional layer and two 3×3 average pooling layers, thereby obtaining a methane concentration-light intensity feature association value:

[0083]

[0084] Among them, Model 0 represents the feature extraction layer, represents two 1×3 convolutional layers with a stride of 1 in series, represents two 3×1 convolutional layers with a stride of 1 in series, represents two 3×3 average pooling layers in series, represents a 3×3 convolutional layer with a stride of 2, represents a 5×5 convolutional layer with a stride of 1, represents a 3×3 convolutional layer with a stride of 1, and M represents a lossless data block;

[0085] Then, the iterative training layer performs several iterative training on the methane concentration-light intensity feature correlation values ​​to form a methane-light feature weight matrix;

[0086] In this embodiment, the iterative training layer uses three iterative branches to learn and train the methane concentration-light intensity feature correlation values ​​several times, and obtains the methane shallow matrix, the light shallow matrix and the methane-light shallow matrix respectively; and the three shallow matrices are feature fused to obtain the methane-light feature weight matrix; the iterative training layer includes 5 3×3 convolutional layers, 3 InceptionV1 blocks, 2 InceptionV3 blocks, 2 ReLU activation functions, 1 softmax activation function, 6 3×3 maximum pooling layers, 4 5×5 adaptive pooling layers and 3 residual blocks:

[0087]

[0088] Among them, Y 1 represents the first iteration branch, Y 2 represents the second iteration branch, Y 3 Represents the third iteration branch, Model 1 represents the iterative training layer, Relu() represents the activation function, AutoPool 5×5 () represents a 5×5 adaptive pooling layer, MaxPool3×3 2 () represents two 3×3 maximum pooling layers in series, represents two 3×3 convolutional layers with a stride of 1 in series, ResBlock() represents a residual block, represents the Hadamard operation;

[0089] Finally, the recognition prediction layer predicts the methane-light feature weight matrix according to the volume of the monitoring area and modifies the alarm thresholds of monitoring areas with different volumes;

[0090] In this embodiment, the recognition prediction layer uses 1 shuffle block, 2 1×1 convolutional layers, 2 fully connected layers, 2 5×5 average pooling layers and a feedback optimization function to verify the gas leak prediction results of the methane-illumination feature weight matrix, and correct the alarm thresholds of monitoring areas of different volumes:

[0091]

[0092] Among them, Model 2 represents the recognition prediction layer, L represents the feedback optimization function, MeanPool 5×5 2 () represents two 5×5 average pooling layers in series, Shuff represents the shuffled block, V represents the volume of the monitoring area, G represents the quantization function, σ represents the number of iterative training layers, and E σ represents the optimization efficiency of iterative training for each layer, E0 represents the preset maximum value of iterative training optimization efficiency, τ represents the total number of iterative training layers, μ1 represents the calculation accuracy coefficient, and B σ represents the iterative training loss function of each layer, and μ2 represents the preset optimization weight coefficient;

[0093]

[0094] Among them, H ′ V It represents the corrected alarm threshold of the monitoring area with different volumes, H V Indicates the current alarm threshold of the monitoring area of ​​different volumes, L V Indicates the feedback optimization function corresponding to the monitoring areas of different volumes;

[0095] The display unit is used to display the working status and methane concentration data of each laser methane alarm and external equipment.

[0096] A wireless remote transmission industrial and commercial laser methane alarm method comprises the following steps in sequence:

[0097] A: The power module supplies power to the laser methane alarm, and the laser methane alarm starts working;

[0098] B: The laser methane alarm uses a light detector to collect the light intensity data of the detection light source in the monitoring area, and then uses spectral absorption technology in the control module to calculate the light intensity data into methane concentration data;

[0099] C: Transmitting the methane concentration data and light intensity data to the processing unit of the laser methane alarm system for pre-processing;

[0100] D: The pre-processed methane concentration data and light intensity data are used to predict gas leakage through the early warning unit, and the alarm thresholds of the monitoring areas of different volumes are corrected according to the gas leakage prediction results, and the corrected alarm thresholds of the monitoring areas of different volumes are transmitted to the laser methane alarm;

[0101] E: The control module of the laser methane alarm generates an alarm signal based on the real-time corrected alarm thresholds of the monitoring areas of different volumes and the real-time methane concentration data and the corrected alarm thresholds, and performs alarm and ventilation emissions through external equipment;

[0102] F: The laser methane alarm system displays the working status and methane concentration data of each laser methane alarm and external equipment on the display unit.

[0103] Example:

[0104] The power module supplies power to the laser methane alarm, and the laser methane alarm starts to work; the laser methane alarm uses a light detector to collect light intensity data of the detection light source in the monitoring area, and then uses spectral absorption technology in the control module to calculate the light intensity data into methane concentration data; at the same time, the methane concentration data and light intensity data are transmitted to the laser methane alarm system;

[0105] Then, the laser methane alarm system receives the methane concentration data and the light intensity data, and uses the processing unit to divide the methane concentration data and the light intensity data into several data blocks according to the preset time interval; the preset time interval is 2 minutes; the internal data of the data block is classified according to the data type; the data cleaning algorithm is used to clean and correct the abnormal data of the classified internal data; the data cleaning algorithm includes a data deduplication algorithm, a data filling algorithm, a data conversion algorithm and a data simplification algorithm; the internal data that has completed data cleaning is first inspected and second inspected according to the methane concentration data and light intensity data of the same collection time; according to the data collection time and the volume of the monitoring area, the inspected data blocks are correlated and normalized to construct an aggregated data block; the aggregated data block is losslessly compressed using a data compression algorithm to form a corresponding lossless data block; the lossless data block is transmitted to the early warning unit through wireless communication using a parallel heuristic transmission algorithm for gas leak prediction;

[0106] Subsequently, the early warning unit in the laser methane alarm system uses a preset leakage prediction model to predict the gas leakage alarm threshold of the lossless data block formed by preprocessing, and corrects the alarm threshold of the monitoring area with different volumes; the leakage prediction model includes a feature extraction layer, an iterative training layer and a recognition prediction layer; first, the feature extraction layer extracts features from the lossless data block to form a methane concentration-light intensity feature correlation value; then, the iterative training layer performs several iterative training on the methane concentration-light intensity feature correlation value to form a methane-light feature weight matrix; finally, the recognition prediction layer predicts the methane-light feature weight matrix according to the volume of the monitoring area, and corrects the alarm threshold of the monitoring area with different volumes;

[0107] The feature extraction layer uses two 3×3 convolutional layers, one 5×5 convolutional layer, two convolutional blocks consisting of 3×1 convolutional layers and 1×3 convolutional layers, and two 3×3 average pooling layers to extract features from the lossless data blocks and obtain the methane concentration-light intensity feature correlation value:

[0108] The iterative training layer uses 5 3×3 convolutional layers, 3 InceptionV1 blocks, 2 InceptionV3 blocks, 2 ReLU activation functions, 1 softmax activation function, 6 3×3 maximum pooling layers, 4 5×5 adaptive pooling layers and 3 residual blocks to train the methane concentration-light intensity feature correlation values ​​several times to obtain the methane-light feature weight matrix;

[0109] The recognition prediction layer uses 1 shuffle block, 2 1×1 convolutional layers, 2 fully connected layers, 2 5×5 average pooling layers and feedback optimization function to verify the gas leak prediction results of the methane-light feature weight matrix, and correct the alarm thresholds of monitoring areas of different volumes;

[0110] Finally, the control module in the laser methane alarm controls the working status of the laser methane alarm and external equipment according to the corrected alarm thresholds of the monitoring areas with different volumes; the display unit of the laser methane alarm system displays the working status and methane concentration data of each laser methane alarm and external equipment.

Claims

1. A wireless remote transmission industrial and commercial laser methane alarm, characterized by: It includes a light source emitter, an optical module, a light detector, a control module, a power module, an interface module, an explosion-proof housing and external equipment; wherein, A light source transmitter, used to generate laser light to serve as a detection light source; An optical module, used for focusing the detection light source and irradiating it to the light detector; A light detector, used to detect light intensity data of a detection light source in a monitoring area; The control module is used to calculate the methane concentration data using the spectral absorption technology based on the light intensity data of the detection light source; at the same time, the working state of the external equipment is adjusted according to the alarm thresholds of the monitoring areas of different volumes corrected in real time; The power module is used to power the laser methane alarm and manage the battery charging and discharging status; Interface module, used to transmit methane concentration data, alarm threshold and light intensity data to external devices for display and setting; An explosion-proof housing is used to protect the internal components of the laser methane alarm; the internal components include a light source emitter, an optical module, a light detector, a control module, a power module and an interface module; An external device for adjusting the working state according to the alarm threshold received by the control module; The power module is connected to the light source emitter, optical module, light detector, control module and interface module respectively through the power management circuit; the light source emitter and light detector are connected through the signal processing circuit, and the control module is connected to the explosion-proof housing with the light source emitter, optical module, light detector and interface module respectively through the signal transmission line.

2. The wireless remote transmission industrial and commercial laser methane alarm according to claim 1 is characterized by: The external devices include a buzzer, a display screen, a gas valve and an exhaust control fan.

3. A wireless remote transmission industrial and commercial laser methane alarm system, applied to the wireless remote transmission industrial and commercial laser methane alarm according to claims 1-2, characterized in that: It includes a processing unit, an early warning unit, a display unit and several laser methane alarms; among which, Laser methane alarm, used to obtain the light intensity data of the detection light source in the monitoring area and calculate the corresponding methane concentration data; The processing unit is used to pre-process the methane concentration data and light intensity data, obtain lossless data blocks after division, classification, cleaning, inspection, aggregation and compression, and send them to the early warning unit for gas leakage prediction; The early warning unit is used to predict the gas leakage alarm threshold value based on the pre-processed methane concentration data and light intensity data according to the volume of the monitoring area, and to correct the alarm threshold value according to the gas leakage prediction result; and then to warn the monitoring area where each laser methane alarm is located according to the real-time methane concentration data and the corrected alarm threshold value; the gas leakage prediction result includes the leakage location, leakage amount and leakage time; the volume of the monitoring area is selected and determined by the assembler according to the installation space of each laser methane alarm; The display unit is used to display the working status and methane concentration data of each laser methane alarm and external equipment.

4. The wireless remote transmission industrial and commercial laser methane alarm system according to claim 3 is characterized by: The laser methane alarm obtains the light intensity data of the detection light source in the monitoring area through the light detector; and the methane concentration data is calculated by the control module.

5. The wireless remote transmission industrial and commercial laser methane alarm system according to claim 3 is characterized by: The process of preprocessing the methane concentration data and light intensity data by the processing unit is as follows: S1: Divide the methane concentration data and the light intensity data into a number of data blocks according to a preset time interval; the preset time interval may be minutes; S2: using a clustering algorithm to classify the internal data of the data block according to the data type; the data type includes methane concentration and light intensity; S3: Use data cleaning algorithms to clean and correct abnormal data in the classified internal data; data cleaning algorithms include data deduplication algorithms, data filling algorithms, data conversion algorithms, and data simplification algorithms; S4: For the internal data that has completed data cleaning, the methane concentration data and light intensity data of the same collection time are respectively checked; S5: According to the data collection time and the volume of the monitoring area, the inspected data blocks are correlated and normalized to construct an aggregated data block; S6: Losslessly compressing the aggregated data block using a data compression algorithm to form a corresponding lossless data block; S7: The lossless data block is transmitted to the early warning unit through the wireless communication protocol using the parallel heuristic transmission algorithm for gas leakage prediction.

6. The wireless remote transmission industrial and commercial laser methane alarm system according to claim 3 is characterized by: The early warning unit uses a preset leakage prediction model to predict gas leakage on the lossless data block formed by preprocessing, and corrects the alarm thresholds of monitoring areas of different volumes; the leakage prediction model includes a feature extraction layer, an iterative training layer and an identification prediction layer; first, the feature extraction layer extracts features from the lossless data block to form a methane concentration-light intensity feature correlation value; Then, the iterative training layer performs several iterative training on the methane concentration-light intensity feature correlation values ​​to form a methane-light feature weight matrix; Finally, the recognition prediction layer predicts the methane-light feature weight matrix according to the volume of the monitoring area and modifies the alarm thresholds of monitoring areas with different volumes.

7. The wireless remote transmission industrial and commercial laser methane alarm system according to claim 6 is characterized by: The feature extraction layer uses two 3×3 convolutional layers and one 5×5 convolutional layer to segment the lossless data block to form a methane concentration-light intensity association pair, and uses two convolutional blocks consisting of a 3×1 convolutional layer and a 1×3 convolutional layer and two 3×3 average pooling layers to extract features of the methane concentration-light intensity association pair, thereby obtaining a methane concentration-light intensity feature association value.

8. The wireless remote transmission industrial and commercial laser methane alarm system according to claim 6 is characterized by: The first iteration branch of the iterative training layer uses two serially connected 3×3 convolutional layers, one InceptionV1 block, one staggered block, two 3×3 maximum pooling layers, one 5×5 adaptive pooling layer and one ReLU activation function to perform Hadamard operation on the methane concentration-light intensity feature correlation value and the lossless data block, and then performs several learning trainings to obtain a methane shallow matrix; The second iterative branch uses 1 3×3 convolutional layer, 1 InceptionV3 block, 1 staggered block, 2 cascaded 3×3 maximum pooling layers, 2 cascaded 5×5 adaptive pooling layers and 1 ReLU activation function to perform several learning and training on the methane concentration-light intensity feature correlation value, thereby obtaining the shallow illumination matrix; the third iterative branch uses 1 3×3 convolutional layer, 2 cascaded InceptionV1 blocks, 2 cascaded 3×3 maximum pooling layers, 1 5×5 adaptive pooling layer and 1 softmax activation function to perform several learning and training on the lossless data block, thereby obtaining the methane-light shallow matrix; The three shallow matrices are then feature-fused using a 3×3 convolutional layer with a stride of 2 and an InceptionV3 block to obtain the methane-light feature weight matrix.

9. The wireless remote transmission industrial and commercial laser methane alarm system according to claim 6 is characterized by: The identified prediction layer uses 1 shuffle block, 2 fully connected layers, 2 serial 5×5 average pooling layers, 2 1×1 convolutional layers with a step size of 2, and a feedback optimization function to verify the gas leak prediction results of the methane-illumination feature weight matrix, and uses the feedback optimization function to correct the alarm thresholds of monitoring areas of different volumes.

Images