Method and system for correcting harmful gas detection in a mine based on vehicle emission monitoring

Abstract

This invention relates to the field of coal mine safety monitoring and environmental sensing, and discloses a method and system for detecting and correcting harmful gases in underground mines based on vehicle emission monitoring. The method first acquires multi-source data on vehicle emission sources, spatial location, airflow field, and sensor observations, and performs time synchronization alignment. Next, it calculates the vehicle's flow field coupling factor to the sensor and calculates the exhaust gas transmission lag time based on the airflow field and location data. Based on emission source data, lag time, and diffusion parameters, it constructs an interference prediction model to quantitatively calculate the theoretical interference concentration generated by the vehicle. Furthermore, it calculates the dynamic confidence weight coefficient and combines the observed data with the theoretical value to calculate the corrected real environmental gas concentration. Finally, it compares the real concentration with a safety threshold and executes alarm triggering or suppression control based on the result. This invention can effectively isolate vehicle exhaust gas interference, restore the true background concentration in the roadway, reduce the false alarm rate, and improve the accuracy of coal mine monitoring and production efficiency.

Description

Technical Field

[0001] This invention relates to the field of coal mine safety monitoring and environmental sensing, specifically to a method and system for detecting and correcting harmful gases in underground mines based on vehicle emission monitoring. Background Technology

[0002] Coal mine safety monitoring relies on a network of gas sensors deployed at key locations in roadways to detect real-time changes in the concentration of harmful gases such as methane, carbon monoxide, and carbon dioxide. With the increasing mechanization of auxiliary transportation systems in coal mines, auxiliary vehicles using explosion-proof diesel engines or rubber-wheeled vehicles frequently travel back and forth within the roadways. The exhaust gases emitted by these vehicles contain high concentrations of carbon monoxide and hydrocarbon gases. When these vehicles pass through fixed monitoring points, the diffusion of these exhaust gases directly interferes with the normal readings of the gas sensors, causing the monitoring data to fail to accurately reflect the background gas concentrations in the roadway environment.

[0003] Existing mine gas monitoring technologies, such as monitoring systems based on CAN bus, industrial Ethernet, or Internet of Things (IoT) and LoRa wireless communication technologies, primarily focus on solving the problems of data acquisition, efficient transmission, and networking connectivity for large-scale sensor nodes underground, emphasizing the stability and coverage of communication links. However, these technologies often overlook the dynamic interference effects of mobile emission sources on fixed-point monitoring equipment within confined spaces. Regarding alarm logic, existing systems typically employ a single concentration threshold determination mechanism, lacking the ability to identify the signal source. Whenever a sensor reading exceeds a preset limit, the system determines it as an environmental anomaly and triggers an audible and visual alarm or a power-off mechanism. This mechanism cannot distinguish between instantaneous concentration spikes caused by vehicle exhaust and actual mine gas leaks, leading to frequent interruptions in underground production operations due to false alarms, thus impacting production efficiency.

[0004] Furthermore, underground tunnels are confined flow field environments with forced ventilation. There is a physical lag in the transmission of gas emitted by vehicles to fixed sensors, and this lag is dynamically affected by wind speed, wind direction, and the relative position of the vehicle and the sensor. Current technologies lack precise modeling of the multidimensional spatiotemporal relationships between vehicle trajectory, emission time, airflow field state, and sensor response, and do not perform rigorous time synchronization and flow field alignment processing on multi-source heterogeneous data. This prevents the monitoring system from establishing a causal mapping between historical emission behavior and current sensor response, making it difficult to accurately remove lag interference components superimposed on background concentration. Simultaneously, the underground environment presents uncertainties such as turbulent airflow, multipath effects, and sensor performance drift. Existing detection methods typically lack adaptive parameter adjustment and weighted correction mechanisms for complex flow field environments, making it difficult to dynamically optimize interference removal strategies based on environmental stability or data reliability. This results in distorted monitoring data under non-steady-state conditions, failing to meet the requirements of high-precision safety monitoring. Summary of the Invention

[0005] The first aspect of the present invention provides a method for detecting and correcting harmful gases in underground mines based on vehicle emission monitoring, which aims to solve the technical problem that existing fixed gas sensors in coal mines cannot distinguish between the interference of exhaust gas from auxiliary vehicles and the increase in the concentration of gases in the actual environment, thereby causing false alarms or power outages that interrupt production.

[0006] The method includes the following steps: First, the system acquires vehicle emission source data collected by the vehicle emission monitoring module, vehicle spatial location data collected by the downhole positioning system, airflow field data collected by the airflow field monitoring device, and fixed sensor observation data collected by the gas sensor network. Second, considering the inconsistent sampling frequencies of the above data sources, the acquired multi-source heterogeneous data undergoes time synchronization and alignment processing to construct a unified time reference.

[0007] Next, based on the synchronized data, the system calculates the flow field coupling factor between the underground auxiliary transport vehicle and the fixed gas sensor, and combines the airflow field data and vehicle spatial position data to accurately calculate the transmission lag time required for the exhaust gas cloud emitted by the vehicle to be transmitted to the fixed gas sensor.

[0008] Subsequently, based on vehicle emission source data, transmission lag time, airflow field data, and roadway diffusion parameters, an interference prediction model is constructed to quantitatively calculate the theoretical interference concentration generated by the underground auxiliary transport vehicle at the fixed gas sensor location at the current moment. After obtaining the theoretical prediction value, the system further calculates the dynamic confidence weight coefficient reflecting environmental stability, and obtains the corrected real environmental gas concentration through weighted difference calculation based on the fixed sensor observation data, theoretical interference concentration, and dynamic confidence weight coefficient.

[0009] Finally, the system compares the actual ambient gas concentration with the preset safe concentration threshold, and intelligently executes alarm triggering or alarm suppression control based on the comparison result.

[0010] Furthermore, to address the issue of asynchronous time processing of multi-source data, a circular data buffer mechanism is employed for the time synchronization and alignment process. Using the time series of fixed sensor observation data as the main axis, adjacent records are retrieved from the buffers of vehicle spatial location data and vehicle emission source data, and a linear interpolation algorithm is executed to calculate the instantaneous vehicle position, speed, and emission intensity that strictly correspond to the sensor sampling time. For slower-changing airflow field data, the nearest neighbor matching method is used for alignment.

[0011] Furthermore, to accurately determine whether a vehicle constitutes an effective interference source, the flow field coupling factor is calculated based on vector geometry. The system constructs a relative position vector of the vehicle pointing towards the sensor and calculates the cosine of the angle between this vector and the wind flow vector. When the cosine of the angle is greater than a preset angle tolerance threshold, the sensor is determined to be located in the effective downwind direction, and a non-zero flow field coupling factor is generated. This factor value reflects the consistency between the wind direction and the propagation path; otherwise, it is determined to be an invalid operating condition and set to zero, thereby eliminating invalid calculation paths.

[0012] Furthermore, to address the causal timing misalignment caused by gas transport, the calculation of the transport lag time integrates physical transport and sensor response characteristics. The system calculates the physical transport time by dividing the Euclidean distance between the vehicle and the sensor by the magnitude of the airflow vector, and then adds the sensor's inherent response time constant to obtain the total lag time. The system uses the current time minus the total lag time to pinpoint the target historical time and backtracks to match the corresponding emission intensity in historical vehicle emission data to ensure the correctness of the physical causality of the model input.

[0013] Furthermore, the interference prediction model is constructed based on the advection diffusion mechanism to quantify the diffusion behavior of vehicle exhaust in a confined space. In the model's operational logic, the theoretical interference concentration is positively correlated with the flow field coupling factor, the tunnel environment diffusion regulation constant, and the historical vehicle emission intensity matched in retrospect; it is inversely correlated with the product of the wind field vector magnitude (wind speed) and the horizontal and vertical diffusion parameters; and the theoretical interference concentration decreases exponentially with the increase of the vertical deviation distance of the sensor relative to the wind flow center axis.

[0014] Furthermore, to improve the robustness of the correction algorithm in complex environments, the system introduces dynamic confidence weight coefficients. The stability of the airflow field is assessed by statistically analyzing the standard deviation of wind speed and direction within a historical window, and the location reliability is evaluated by analyzing the jitter frequency of the positioning data. A normalized mapping function is used to convert these indicators into weight coefficients. When unsteady turbulence or multipath effects are detected, the system automatically reduces the weight coefficients, minimizing model correction and preventing computational errors introduced by environmental uncertainties.

[0015] Furthermore, the calculation of real-environment gas concentrations employs a weighted correction strategy. The system uses dynamic confidence weighting coefficients to weight the theoretical interference concentration, subtracting this interference component from the fixed sensor observation data. Simultaneously, the system performs physical constraint checks on the calculation results, clamping negative results to zero or historical baseline values, and uses an exponentially weighted moving average filter to eliminate numerical noise, ensuring the smoothness of the output data.

[0016] Furthermore, this invention employs dual threshold logic for alarm control, effectively distinguishing between vehicle interference and genuine exceedances. Only when both the fixed sensor's observed data and the corrected actual ambient gas concentration exceed the safe concentration threshold is the system determined to be a genuine gas exceedance event, triggering an alarm and power cutoff. If only the observed data exceeds the limit while the actual concentration does not, the system determines it to be a vehicle interference event, locks the alarm actuator, and records the event information, thereby preventing accidental power outages while ensuring safety.

[0017] Furthermore, this method also includes an online adaptive update mechanism for model parameters. The system automatically identifies a parameter calibration observation window where a single vehicle passes at a constant speed and the background is stable, and extracts the measured interference characteristic curves within this window. Using the root mean square error between the measured curve and the theoretical curve as the objective function, an iterative optimization algorithm is used to invert and fine-tune key parameters such as the diffusion adjustment constant, enabling the model to adapt to changes in the tunnel environment or the effects of sensor aging.

[0018] A second aspect of the present invention provides a downhole hazardous gas detection and correction system based on vehicle emission monitoring, the system comprising a vehicle emission monitoring module, a downhole positioning system, an airflow field monitoring device, a gas sensor network, a data acquisition and processing module, and a central control system.

[0019] The system comprises several subsystems: a vehicle emission monitoring module for real-time detection of vehicle exhaust composition and concentration; an underground positioning system for acquiring vehicle three-dimensional coordinates; an airflow monitoring device for monitoring wind speed and direction in the tunnel; and a gas sensor network for collecting ambient gas observation data. The data acquisition and processing module receives data from these subsystems, executes the correction method described in the first aspect, and outputs the actual ambient gas concentration. The central control system receives the actual ambient gas concentration and, based on the comparison between this concentration and a safety threshold, controls the activation and deactivation of the on-site audible and visual alarms and the power-off controller.

[0020] The technical solution provided in this invention integrates multi-dimensional heterogeneous data, including vehicle emission characteristics, vehicle trajectory, roadway airflow conditions, and sensor observations, to establish and solve a dynamic interference prediction model with spatiotemporal lag compensation. This solution not only removes the superimposed components of vehicle exhaust gas from sensor readings, restoring the true background gas concentration in the roadway, but also ensures the system's adaptability and correction accuracy in complex underground ventilation environments through dynamic weight adjustment and online parameter update mechanisms. This significantly reduces the false alarm rate of underground gas caused by auxiliary transport vehicles, improving the intelligence level and operational efficiency of coal mine safety monitoring systems.

[0021] This invention provides a method and system for detecting and correcting harmful gases in wells based on vehicle emission monitoring. It has the following beneficial effects:

[0022] 1. This invention establishes a rigorous spatiotemporal causal mapping between historical emission behavior and current sensor response by synchronizing vehicle emission source data, vehicle spatial location data, airflow field data, and fixed sensor observation data in a timely manner, and by combining wind flow vectors and relative position vectors to calculate precise transmission lag times. This mechanism solves the problem of data timing misalignment caused by airflow convection transmission in confined underground spaces, ensuring the physical correctness of the interference prediction model input, thereby significantly improving the accuracy of reconstructing the real environmental background gas concentration from superimposed signals.

[0023] 2. This invention utilizes flow field coupling factors and dynamic confidence weighting coefficients to achieve adaptive adjustment of model correction intensity. By analyzing airflow field stability and positioning data reliability, the system can automatically reduce correction weights in harsh environments such as unsteady turbulence or multipath interference, preventing over-correction or negative errors due to model prediction bias. Combined with an online adaptive update strategy for model parameters, the system can adapt to changes in roadway cross-sections and the effects of sensor aging, ensuring the long-term robustness of the monitoring system in complex ventilation environments.

[0024] 3. This invention employs a dual-threshold alarm logic based on parallel verification of real-environment gas concentration and original observation data, effectively distinguishing between vehicle exhaust interference events and actual gas over-limit events. When a high concentration reading is detected solely due to vehicle emissions, this logic automatically generates an alarm suppression signal, locking the audible and visual alarms and power-off controller, preventing unnecessary power outages and production interruptions caused by auxiliary vehicle operation. Simultaneously, it immediately triggers linkage control when the actual environmental concentration exceeds the limit, significantly improving mine production efficiency while ensuring the reliability of coal mine safety monitoring. Attached Figure Description

[0025] Figure 1 This is a diagram showing the overall architecture and data interaction topology of the downhole hazardous gas detection and correction system based on vehicle emission monitoring according to the present invention. Figure 2 This is a schematic diagram of the dual-threshold intelligent alarm and control logic of the present invention; Figure 3 This is a flowchart of the core algorithm for gas concentration correction based on multi-source data fusion in this invention. Detailed Implementation

[0026] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] Please see the appendix Figure 1 This invention provides an underground hazardous gas detection and correction system based on vehicle emission monitoring, including a vehicle emission monitoring module, an underground positioning system, a gas sensor network, an airflow field monitoring device, a data acquisition and processing module, and a central control system.

[0028] The data acquisition and processing module establishes communication connections with each of the aforementioned subsystems, receiving vehicle emission source data, vehicle spatial location data, fixed sensor observation data, and airflow field data, and performing time synchronization and alignment. This module stores a vehicle emission impact model and processes the data based on flow field coupling factors and interference concentration prediction algorithms: First, it calculates the flow field coupling factor based on the wind flow vector and the relative position vector between the vehicle and the sensor to determine whether the vehicle is located in the effective upwind direction of the sensor; then, it calculates the theoretical interference concentration by combining vehicle emission intensity, lag time, and roadway diffusion parameters; finally, it subtracts the theoretical interference concentration from the fixed sensor observation data using a dynamic confidence weighting coefficient to obtain the actual ambient gas concentration. The central control system receives the actual concentration and only alarms when the actual concentration exceeds the limit. If only the observed data exceeds the limit but the actual concentration does not, it is determined to be vehicle interference, and the alarm is suppressed.

[0029] The vehicle emissions monitoring module employs a pump-suction structure, collecting exhaust gas samples via a high-temperature resistant probe installed in the exhaust tailpipe. After pretreatment by a condenser and filter, the sample gas enters a multi-gas sensor array. The array includes a carbon monoxide sensor based on electrochemical principles and a methane sensor based on non-dispersive infrared principles. The module connects to the vehicle's OBD interface via a CAN interface to read parameters such as engine speed and intake manifold pressure. Combining this with gas concentration data, it estimates the emission mass flow rate, generates vehicle emission source data, and transmits it wirelessly.

[0030] The underground positioning system uses an intrinsically safe ultra-wideband (UWB) positioning tag mounted on the top of the driver's cab. The tag sends a sequence of ultrashort pulses containing an ID and timestamp to the positioning base station and receives synchronization signals, providing the physical layer raw data for calculating the vehicle's high-precision coordinates and trajectory. The tag is physically connected to the vehicle's emissions monitoring module for strict clock synchronization.

[0031] The gas sensor network consists of fixed multi-parameter gas detectors distributed in the return airway and transport main roadway. It integrates a catalytic combustion methane sensor (Wheatstone bridge principle), an electrochemical carbon monoxide sensor, and an infrared carbon dioxide sensor, and uploads background concentration data via industrial Ethernet or RS485 bus.

[0032] The airflow field monitoring device uses intrinsically safe ultrasonic anemometers and wind direction sensors suspended on the cross-section of the tunnel. The orthogonally arranged ultrasonic probes inside use the time-of-flight method to calculate the airflow velocity vector and direction, and decompose the velocity into longitudinal and transverse components to provide fluid dynamic boundary conditions for the model.

[0033] The positioning base stations of the underground positioning system are arranged at intervals (50-200 meters) along the roadway and connected to the data module via fiber optic or power line carrier. The base stations achieve nanosecond-level synchronization through the IEEE 1588PTP protocol, recording the arrival time of the tag signals for the backend to calculate the coordinates using the TDOA algorithm.

[0034] The data acquisition and processing module (industrial control computer or edge server) connects to the downhole ring network via a gigabit Ethernet interface and a fiber optic adapter, using a multi-threaded program to receive data in parallel via MQTT, Modbus TCP, and UDP protocols. This module has a built-in NTP / PTP server to provide time synchronization for each subsystem and uses interpolation to map multi-source data to a unified time axis. The processor calls the vehicle emission impact model to calculate interference concentrations in real time and outputs the corrected true concentrations.

[0035] The central control system runs configuration software that dynamically overlays vehicle positions, sensor icons, and airflow vectors onto a 3D topology map, and renders the exhaust gas diffusion plume range. The system provides a human-machine interface to display real-time curves and allows parameter adjustments (such as safety thresholds and confidence coefficients). The internal alarm logic control unit performs dual verification: it triggers audible and visual alarms and power-off commands only when the actual ambient gas concentration exceeds the safety threshold; if vehicle interference is detected, a suppression record is generated, achieving intelligent alarm management.

[0036] Please see the appendix Figure 3 The method for detecting and correcting harmful gases in wells based on vehicle emission monitoring provided by this invention includes the steps of simultaneous acquisition and preprocessing of multi-source heterogeneous data: The data acquisition and processing module is configured with multiple independent circular data buffers in memory. These buffers are used to temporarily store vehicle emission source data sent by the vehicle emission monitoring module, vehicle spatial location data sent by the downhole positioning system, airflow field data sent by the airflow field monitoring device, and fixed sensor observation data sent by the gas sensor network. Because the sampling frequencies of each subsystem differ—for example, the update frequency of vehicle spatial location data is typically higher than that of fixed sensor observation data—the data acquisition and processing module performs timestamp-based alignment.

[0037] The data acquisition and processing module uses the time series of fixed sensor observation data as the primary time reference axis. When the data acquisition and processing module receives a new fixed sensor observation data packet from the gas sensor network, it extracts the sampling timestamp from the fixed sensor observation data packet. The data acquisition and processing module is configured to use the sampling timestamp as the index key to perform searches in the vehicle emission source data buffer, vehicle spatial location data buffer, and airflow field data buffer.

[0038] For high-frequency updated data streams such as vehicle spatial location data and vehicle emission source data, the data acquisition and processing module is configured to execute a linear interpolation algorithm to obtain precise moment values. Specifically, if the sampling timestamp falls between two adjacent recording times in the vehicle spatial location data buffer, the data acquisition and processing module calculates the instantaneous vehicle position and instantaneous vehicle velocity that strictly correspond to the sampling timestamp by weighted averaging the three-dimensional coordinate values ​​and velocity vector values ​​corresponding to the two adjacent recording times based on the time difference between the sampling timestamp and the two adjacent recording times. For data streams with slower changes, such as airflow field data, the data acquisition and processing module is configured to use the nearest neighbor matching method to select the airflow field data record with the smallest time interval from the sampling timestamp as the current fluid boundary condition.

[0039] The data acquisition and processing module is configured to unify the spatial coordinate system of the acquired raw data. The module stores a map of the entire mine's absolute coordinate system. It maps the relative or local coordinates output by the underground positioning system to the mine's absolute coordinate system. Simultaneously, the module reads the physical installation coordinates of each fixed gas sensor pre-set in the memory. The module ensures that the vehicle's three-dimensional coordinates, the fixed gas sensor's three-dimensional coordinates, and the airflow vector of the airflow field monitoring device are all within the same three-dimensional Cartesian coordinate space to facilitate subsequent calculations of Euclidean distance and vector angles.

[0040] The data acquisition and processing module is further configured to perform outlier removal and filtering smoothing on the vehicle emission source data. Since the onboard sensor readings may generate instantaneous spike noise during vehicle bumpy driving, the data acquisition and processing module uses a moving average filter to smooth the vehicle emission source data. The data acquisition and processing module sets a sliding time window, calculates the arithmetic mean of multiple consecutive sampling points within the sliding time window, and uses this arithmetic mean as the effective emission intensity for model calculation. Through the above synchronous acquisition and preprocessing steps, the data acquisition and processing module constructs a data set containing time intervals... The vehicle emission characteristics, vehicle motion state, environmental airflow state, and environmental gas observation values ​​are simultaneously multidimensional feature vectors, and these simultaneous multidimensional feature vectors are input into the vehicle emission impact model.

[0041] The downhole hazardous gas detection correction method based on vehicle emission monitoring provided by this invention includes a geometric determination step for spatial airflow coupling relationship: The data acquisition and processing module is configured to construct the relative geometric relationship between the vehicle and the sensor based on synchronized and aligned vehicle spatial position data and fixed sensor observation data. The module first extracts the three-dimensional coordinates of the underground auxiliary transport vehicle in the absolute coordinate system of the entire mine, and the three-dimensional coordinates of the fixed gas sensor in the same system. The module then calculates the relative position vector from the vehicle's three-dimensional coordinates to the sensor's three-dimensional coordinates using vector subtraction. This relative position vector represents the straight-line physical path of exhaust gas propagating from the vehicle's emission source to the fixed gas sensor. Finally, the module calculates the magnitude of the relative position vector, which is the Euclidean distance between the vehicle and the fixed gas sensor.

[0042] The data acquisition and processing module is configured to import airflow field data for directionality determination to eliminate invalid interference sources. The module acquires the airflow field vector collected by the airflow field monitoring device at the current moment. The airflow field vector includes the magnitude of the airflow velocity and the direction of the airflow. The module is also configured to calculate the cosine of the angle between the airflow field vector and the relative position vector. Using the vector dot product formula, the module divides the dot product of the airflow field vector and the relative position vector by the product of the magnitudes of the airflow field vector and the relative position vector, obtaining a normalized value that reflects the consistency between wind direction and propagation path.

[0043] The data acquisition and processing module is configured to calculate the flow field coupling factor based on the calculated cosine of the included angle, using the flow field coupling factor formula to quantitatively determine whether the underground auxiliary transport vehicle is located on the effective upwind side of the fixed gas sensor. The formula for calculating the flow field coupling factor is as follows: ; in, Indicates time The flow field coupling factor is used to quantify the effective interference of the vehicle on the sensor. Indicates time The airflow field vector collected by the airflow field monitoring device; Indicates time The relative position vector of the vehicle's spatial position to the fixed gas sensor position; The magnitude of the wind flow vector (i.e., the wind speed); The magnitude of the relative position vector (i.e., the Euclidean distance between the vehicle and the sensor). This represents the vector dot product operation; This represents the preset angle tolerance threshold, used to define the effective upwind conical region; Let represent the Herveside step function.

[0044] The data acquisition and processing module is configured to perform logical judgments based on the calculation results of the flow field coupling factor. When the expression within the Herveyd step function is greater than zero, i.e., the cosine of the angle between the airflow vector and the relative position vector is greater than a preset angle tolerance threshold, the data acquisition and processing module determines that the fixed gas sensor is located in the effective downwind region of the underground auxiliary transport vehicle, and that the airflow direction can transport the vehicle's exhaust gas to the fixed gas sensor location. At this time, the data acquisition and processing module generates a non-zero flow field coupling factor. This flow field coupling factor retains the value of the angle weight, indicating that the higher the degree of wind alignment, the higher the exhaust gas transmission efficiency.

[0045] When the expression within the Herveside step function is less than or equal to zero—that is, when the cosine of the angle between the airflow vector and the relative position vector is less than a preset angle tolerance threshold, or when the airflow direction forms an obtuse angle with the relative position vector (i.e., the vehicle is located downwind or crosswind of the sensor)—the data acquisition and processing module determines this as an invalid interference condition. In this case, the flow field coupling factor is forcibly set to zero. The data acquisition and processing module is configured to directly terminate the interference concentration calculation process for the fixed gas sensor at the current moment when the flow field coupling factor is zero. This not only saves computational resources but also prevents vehicle exhaust data, which is physically impossible to transmit, from erroneously correcting the observations of the fixed gas sensor. Angle tolerance threshold. The setting value is usually close to 1.0, which is used to define a conical effective diffusion region with the airflow direction as the central axis. The subsequent interference quantification model is only activated when the fixed gas sensor is located within this conical effective diffusion region.

[0046] The downhole hazardous gas detection correction method based on vehicle emission monitoring provided by this invention includes the steps of calculating and matching dynamic lag time: The data acquisition and processing module is configured to identify and quantify the time delay characteristics required for exhaust gases emitted by underground auxiliary transport vehicles to travel to a fixed gas sensor. Because the underground coal mine environment is a confined space with forced ventilation, harmful gases emitted by vehicles do not reach the fixed gas sensor instantaneously, but are transported via convection with the airflow within the roadway. The data acquisition and processing module first calculates the physical transport lag time. The physical transport lag time is defined as the theoretical flight time required for the exhaust gas cloud to drift from the vehicle's current emission location to the fixed gas sensor location with the airflow.

[0047] The data acquisition and processing module calls upon the Euclidean distance between the vehicle and the stationary gas sensor, calculated in the geometric determination step of the spatial airflow coupling relationship. Simultaneously, the module reads the magnitude of the airflow field vector measured by the airflow field monitoring device. The module is configured to divide the Euclidean distance by the magnitude of the airflow field vector to obtain the basic physical transmission lag time. To prevent division by zero errors or calculated lag times approaching infinity due to static airflow or extremely low wind speeds, the module sets a minimum wind speed clamping threshold in the denominator of the division operation. When the magnitude of the airflow field vector is lower than the minimum wind speed clamping threshold, the module uses the minimum wind speed clamping threshold instead of the actual wind speed for calculation.

[0048] The data acquisition and processing module is further configured to compensate for the inherent response delay of the stationary gas sensors. Because there is a physical response time (commonly referred to as the T90 response time) between the contact with the gas and the output of a stable electrical signal by electrochemical carbon monoxide sensors or catalytic combustion methane sensors, the data acquisition and processing module has preset sensor response time constants for different types of stationary gas sensors in its memory. The data acquisition and processing module adds the basic physical transmission lag time to the sensor response time constant to obtain the total transmission lag time. The total transmission lag time represents the total time difference from the moment the vehicle emission occurs to the moment the stationary gas sensor produces the corresponding response.

[0049] The data acquisition and processing module is configured to perform a backtracking matching operation on historical data based on the calculated total transmission lag time. The module obtains the timestamp of the current system moment. It then subtracts the total transmission lag time from the current system timestamp to calculate the target historical moment. This target historical moment points to the specific vehicle's historical emission instant that caused the change in the current fixed gas sensor reading.

[0050] The data acquisition and processing module is configured to access a pre-built vehicle emission source data buffer. This buffer is a ring-shaped storage structure with a first-in, first-out (FIFO) characteristic, storing sequences of vehicle emission source data with historical timestamps. The data acquisition and processing module retrieves data from the vehicle emission source data buffer using the target historical time as the search key.

[0051] Given that vehicle emission source data is discretely sampled, the vehicle emission source data buffer may not contain timestamp records that are completely consistent with the target historical moment. The data acquisition and processing module is configured to perform a time-axis linear interpolation algorithm. The module searches the vehicle emission source data buffer for a data frame (forward frame) whose timestamp immediately precedes the target historical moment and a data frame (backward frame) whose timestamp immediately follows the target historical moment. Based on the time distance between the target historical moment and the timestamps of the forward and backward frames, the module calculates interpolation weighting coefficients. Using these weighting coefficients, the module weights the emission intensity recorded in the forward and backward frames to synthesize a vehicle emission intensity that precisely corresponds to the target historical moment.

[0052] The data acquisition and processing module is configured to use the matched and calculated historical vehicle emission intensity as the input variable for the interference quantification model. Through the above steps, the data acquisition and processing module establishes a causal mapping relationship between the sensor readings at the current moment and the vehicle emission behavior at past moments, ensuring the physical causal correctness of subsequent correction calculations in the time dimension and avoiding the timing misalignment error caused by directly using the vehicle emission data at the current moment to correct the current sensor readings.

[0053] The downhole hazardous gas detection correction method based on vehicle emission monitoring provided by this invention includes a quantitative prediction and modeling step for interference concentration: The data acquisition and processing module is configured to establish an interference prediction model based on the advection diffusion mechanism. Using this model, the module quantitatively calculates the theoretical interference concentration transmitted from the historical emission behavior of underground auxiliary vehicles to the fixed gas sensor location at the current moment. The module treats the underground roadway as a confined tubular space and employs an improved Gaussian plume model as its core algorithm.

[0054] The data acquisition and processing module first obtains a set of input variables required for model calculation. The input variables include: the flow field coupling factor calculated in the spatial airflow coupling relationship geometric determination step; the historical vehicle emission intensity that is causally related to the current moment obtained in the dynamic lag time calculation and matching step; the magnitude of the wind flow field vector at the current moment measured by the airflow field monitoring device; and the relative position vector of the vehicle's spatial position pointing to the fixed gas sensor position.

[0055] The data acquisition and processing module is configured to calculate gas diffusion parameters. Based on the straight-line distance between the vehicle and the stationary gas sensor, the module determines the horizontal and vertical diffusion parameters. The module's memory stores a diffusion parameter lookup table or empirical regression formula. The lookup table is constructed based on the turbulence intensity level and atmospheric stability classification of the underground roadway. The module retrieves the corresponding horizontal and vertical diffusion parameter values ​​from the lookup table based on the straight-line distance. These two parameters describe the degree of volume expansion and concentration dilution of the exhaust gas cloud as it travels with increasing distance.

[0056] The data acquisition and processing module is configured to calculate the vertical deviation distance of the fixed gas sensor relative to the central axis of the airflow. The module constructs a local coordinate system with the vehicle's instantaneous position as the origin and the airflow vector direction as the X-axis. The module projects the three-dimensional coordinates of the fixed gas sensor into this local coordinate system and calculates the projected distance of the fixed gas sensor on the plane perpendicular to the airflow direction, i.e., the magnitude of the vertical distance vector.

[0057] The data acquisition and processing module is configured to substitute the above parameters into the interference concentration prediction formula for calculation. The formula for calculating the interference concentration is as follows: ; in, Indicates time The theoretical interference concentration at the fixed gas sensor; Indicates time The flow field coupling factor (calculated from the aforementioned formula); This represents the diffusion regulation constant in a tunnel environment, which is related to the tunnel cross-section and boundary reflection. Indicates the time lag Vehicle emission intensity measured by the vehicle emission monitoring module; The transmission lag time is represented by the straight-line distance between the vehicle and the stationary gas sensor. With airflow speed The ratio is determined (which may include sensor response time compensation). Indicates time The vector magnitude of the airflow field; Indicates that it depends on distance Horizontal diffusion parameters; Indicates that it depends on distance Vertical diffusion parameters; Indicates time The straight-line distance between the vehicle and the stationary gas sensor (i.e. ); Represented by natural constant An exponential function with base 0; Indicates time The magnitude of the vertical distance vector between the fixed gas sensor position and the wind flow center axis (with the vehicle as the origin and the wind direction as the axis); The variance of the overall diffusion is usually expressed by... and The synthesis or specific empirical value is determined.

[0058] The theoretical interference concentration calculated by the data acquisition and processing module using this formula can dynamically reflect changes in vehicle emission source intensity, dilution effects caused by wind speed, diffusion attenuation caused by distance, and concentration gradient changes caused by sensor position deviation. The data acquisition and processing module is configured to pass the calculated theoretical interference concentration as a correction term to subsequent signal correction steps.

[0059] The method for detecting and correcting harmful gases in wells based on vehicle emission monitoring provided by this invention includes signal correction and true concentration restoration steps: The data acquisition and processing module is configured to receive real-time fixed sensor observation data uploaded from the gas sensor network, as well as the theoretical interference concentration calculated and output in the interference concentration quantification and prediction modeling step. The module is also configured to extract non-environmental background interference components caused by underground auxiliary vehicles from the fixed sensor observation data using a weighted difference algorithm. To address the impact of model prediction bias and environmental uncertainties on the correction results, the data acquisition and processing module introduces a dynamic confidence weighting coefficient mechanism.

[0060] The data acquisition and processing module first assesses the stability of the current airflow field data. It then extracts the airflow field vector sequence from a preset time window. The module calculates the standard deviation of the airflow velocity modulus and the standard deviation of the airflow direction angle. When the standard deviation of either the airflow velocity modulus or the standard deviation of the airflow direction angle exceeds a preset turbulence threshold, the module determines that the current airflow in the tunnel is in a non-steady-state turbulent flow mode. In this non-steady-state turbulent flow mode, the prediction accuracy of the Gaussian plume model decreases. Therefore, the data acquisition and processing module is configured to reduce the value of the dynamic confidence weight coefficient to prevent over-correction due to model errors.

[0061] The data acquisition and processing module simultaneously evaluates the reliability of the spatial location of the underground auxiliary transport vehicles. It analyzes the jitter frequency of the vehicle spatial location data uploaded by the underground positioning system. If the vehicle spatial location data experiences drastic changes that exceed the laws of physical motion within a short period, the data acquisition and processing module determines that the positioning data suffers from multipath interference. In this case, the module further reduces the value of the dynamic confidence weighting coefficient.

[0062] The data acquisition and processing module is configured to use a normalized mapping function to generate a dynamic confidence weight coefficient ranging from 0 to 1 in real time, integrating airflow stability and positioning reliability indicators. When the airflow is stable and the positioning is accurate, the dynamic confidence weight coefficient approaches 1, indicating that the data acquisition and processing module fully adopts the calculated results of the theoretical interference concentration. When environmental conditions are severe, the dynamic confidence weight coefficient approaches 0, indicating that the data acquisition and processing module only performs minor corrections or no corrections on the fixed sensor observation data.

[0063] The data acquisition and processing module is configured to perform the final calculation of the actual concentration. The formula for calculating the actual ambient gas concentration is as follows: ; in, Indicates the corrected time. The actual ambient gas concentration (i.e., the background concentration after removing vehicle interference); Indicates time Raw fixed sensor observation data collected by a gas sensor network; Indicates time The dynamic confidence weight coefficient has a range of values ​​of 100. It is used to adjust the correction intensity based on environmental stability and positioning reliability; Represents the time calculated based on the model. The theoretical interference concentration.

[0064] The data acquisition and processing module is configured to perform physical constraint boundary checks on the calculated real-environment gas concentration. If the calculation result is negative, which is physically impossible, it usually means that the theoretical interference concentration has been overestimated. The data acquisition and processing module is configured to force the real-environment gas concentration to be clamped to zero or to the historical baseline value of the fixed gas sensor before the vehicle enters.

[0065] The data acquisition and processing module is configured to perform post-processing smoothing on the corrected real-environment gas concentration sequence. The module uses an exponentially weighted moving average filter to filter the real-environment gas concentrations to eliminate high-frequency noise introduced by the correction calculation. The module then uses the smoothed real-environment gas concentration as the final system output value. This real-environment gas concentration is transmitted to the central control system via a communication network for subsequent alarm detection and ventilation linkage control, thereby ensuring the authenticity of the downhole environmental monitoring data and the accuracy of the alarm system.

[0066] Please see the appendix Figure 2The alarm logic control unit integrated within the central control system is configured to execute a hazard discrimination strategy based on a dual-verification mechanism. The alarm logic control unit has a preset safe concentration threshold in its memory. This safe concentration threshold is the upper limit of gas concentration that a fixed gas sensor is allowed to detect, as set according to coal mine safety regulations. The alarm logic control unit is configured to receive two sets of data streams in parallel from the data acquisition and processing module: one set is uncorrected fixed sensor observation data, and the other set is the actual ambient gas concentration corrected by the vehicle emission impact model.

[0067] The alarm logic control unit is configured to perform the first-level threshold discrimination operation. The alarm logic control unit compares the fixed sensor observation data with the safe concentration threshold in real time. When the fixed sensor observation data is below the safe concentration threshold, the alarm logic control unit determines that the current monitoring area is in a safe state, does not trigger any action, and continues to maintain monitoring. When the fixed sensor observation data is greater than or equal to the safe concentration threshold, the alarm logic control unit initiates the second-level verification procedure instead of immediately triggering the alarm actuator.

[0068] In the second-level verification procedure, the alarm logic control unit reads the actual ambient gas concentration at the same time and compares it again with the safe concentration threshold. The alarm logic control unit is configured to distinguish between vehicle interference events and actual gas over-limit events based on the comparison result.

[0069] When the data observed by the fixed sensor exceeds the safe concentration threshold, and the actual ambient gas concentration is less than the safe concentration threshold, the alarm logic control unit determines that the current high concentration reading is a false exceedance caused by exhaust emissions from underground auxiliary vehicles. In this logic state, the alarm logic control unit is configured to generate an alarm suppression control signal. This signal locks the audible and visual alarm and the power-off controller, keeping them silent and closed to prevent unnecessary power outages and production interruptions caused by passing vehicles. The alarm logic control unit is also configured to create a vehicle interference event record in the system's historical database. This record includes the timestamp of the event, the fixed gas sensor number that triggered the threshold, the unique identifier of the underground auxiliary vehicle that passed through the area at that time, a snapshot of the emission source data uploaded by the vehicle emission monitoring module, and the theoretical interference concentration calculated by the data acquisition and processing module. The vehicle interference event record is used for subsequent safety audits and model accuracy verification.

[0070] When the data observed by the fixed sensor exceeds the safe concentration threshold, and the actual ambient gas concentration also exceeds the safe concentration threshold, the alarm logic control unit determines that there is a real risk of hazardous gas leakage or accumulation in the current monitoring area, or determines that vehicle emissions superimposed on a high background concentration lead to a loss of overall environmental safety. In this logical state, the alarm logic control unit is configured to immediately generate an alarm trigger control signal. This signal is transmitted via an industrial fieldbus to the audible and visual alarm at the mine site, driving it to emit a high-decibel alarm sound and flashing red light. Simultaneously, the alarm trigger control signal is transmitted to the power-off controller of the underground power supply system, driving it to cut off the power to non-intrinsically safe electrical equipment in the relevant work area to prevent electrical sparks from igniting high-concentration methane.

[0071] The alarm logic control unit is also configured to handle a special operating condition: where the fixed sensor readings are below the safe concentration threshold, but the actual ambient gas concentration is abnormally high. Although this situation is physically rare (usually meaning a negative deviation in the correction algorithm), for system safety, the alarm logic control unit is configured to send a yellow warning to the central control system's human-machine interface when the actual ambient gas concentration approaches a preset percentage (e.g., 90%) of the safe concentration threshold. This prompts management personnel to check the zero-point drift of the gas sensor network or the parameter settings of the vehicle emission impact model. Through this dual-threshold intelligent alarm logic, the central control system minimizes the false alarm rate caused by vehicle operation while ensuring that no real danger is missed.

[0072] The data acquisition and processing module is configured to implement an online adaptive update strategy for model parameters to address model accuracy drift caused by changes in underground roadway ventilation conditions or sensor aging. The module maintains a dynamic parameter database in its memory. This database stores diffusion regulation constants and comprehensive diffusion variances corresponding to different monitoring areas or different fixed gas sensors underground. During system operation, the module is configured to automatically identify the optimal observation window for parameter calibration. The optimal observation window is defined as a specific time period that meets the following two conditions: First, the underground positioning system confirms that only one underground auxiliary transport vehicle passes through the monitoring range of the target fixed gas sensor, and the vehicle is traveling at a constant speed; second, within a preset time period before the vehicle's arrival, the fluctuation amplitude of the fixed sensor's observation data is less than a preset background noise threshold, indicating that the current ambient background gas concentration is stable.

[0073] When the data acquisition and processing module identifies the optimal observation window, it automatically initiates the parameter inversion algorithm. First, the module extracts the fixed sensor observation data sequence for that time period. It then calculates the average value of the fixed sensor observation data before the vehicle's arrival and sets this average value as the current baseline background concentration. Finally, the module subtracts the baseline background concentration from the fixed sensor observation data sequence within the optimal observation window to obtain the measured vehicle emission interference characteristic curve. This curve reflects the actual spatial diffusion distribution of vehicle exhaust gases under the current real-world ventilation and tunnel geometry conditions.

[0074] The data acquisition and processing module is configured to construct a parameter optimization objective function. This objective function is defined as the root mean square error between the measured vehicle emission interference characteristic curve and the theoretical interference concentration curve calculated based on the current model parameters. The module employs iterative optimization algorithms, such as the Levenberg-Marquardt algorithm or gradient descent, to fine-tune key parameters in the vehicle emission impact model. The module focuses on adjusting the diffusion regulation constant and the horizontal diffusion parameter. In each iteration, the module changes the value of the diffusion regulation constant, recalculates the theoretical interference concentration curve, and evaluates the numerical change of the parameter optimization objective function until it converges to its minimum. The corresponding parameter values ​​at this point represent the optimal model parameters for the current environment.

[0075] To prevent drastic fluctuations in model parameters due to single measurement errors, the data acquisition and processing module is configured to update the dynamic parameter database using an exponential moving average strategy. The module calculates the weighted average of the optimal model parameters and the currently stored historical parameters in the dynamic parameter database. An update learning rate factor is set within the module, determining the rate at which historical parameters are replaced by new parameters. The module then writes the calculated weighted average into the dynamic parameter database, serving as the latest parameter basis for calculating the vehicle emission impact model at the next time step.

[0076] The data acquisition and processing module is configured to perform boundary integrity checks on the updated model parameters. It stores a table of physical constraint ranges for the parameters. If the diffusion regulation constant obtained from the inversion calculation exceeds a physically reasonable range (e.g., becomes negative or exceeds the theoretical maximum value by several times), the data acquisition and processing module determines that the online adaptive update is invalid, discards the calculation results, and generates a parameter anomaly log, sending it to the central control system to prompt maintenance personnel to check for hardware faults in the airflow field monitoring device or gas sensor network in that area. Through this online adaptive update mechanism, the data acquisition and processing module enables the vehicle emission impact model to self-evolve as the roadway physical environment changes, ensuring continuous high-precision correction.

[0077] This invention provides a hardware structure for a downhole hazardous gas detection and correction system based on vehicle emission monitoring, which is specifically embodied in an industrial-grade electronic device or server.

[0078] Electronic devices include a central processing unit (CPU), system memory, mass storage device, system bus, and input / output interface (I / O) control bridge. The system bus is configured to transmit instruction and data streams between the CPU, system memory, mass storage device, and I / O interface control bridge. The system bus employs one or more of the following: an industry-standard architecture bus, a peripheral component interconnection standard bus, or a universal serial bus.

[0079] The central processing unit (CPU) is configured as the computational and control core of the electronic device. The CPU is physically implemented as a high-performance multi-core microprocessor, digital signal processor, or field-programmable gate array (FPGA). The CPU integrates an arithmetic logic unit (ALU) and a floating-point arithmetic unit. The CPU is configured to execute multi-threaded floating-point matrix operation instructions in parallel to meet the computational demands of solving gas diffusion equations and calculating flow field vectors in vehicle emission impact models. The CPU exchanges data with system memory via a cache.

[0080] The system memory is configured to provide the volatile data storage space required during the operation of the central processing unit. The system memory includes random access memory, specifically double-data-rate synchronous dynamic random access memory. The system memory stores the operating system kernel image, network communication protocol stack, and the running vehicle emissions monitoring and correction application.

[0081] Mass storage devices are configured as non-volatile computer-readable storage media. The physical implementation of the mass storage device is an enterprise-class solid-state drive or a redundant array of independent disks. The mass storage device connects to the system bus via a Serial Advanced Technology Accessory (TAIA) interface or a non-volatile memory host controller interface. The mass storage device is configured to persistently store the algorithm code library for vehicle emissions impact models, historical vehicle emissions source data, historical fixed sensor observation data, and a dynamic parameter database.

[0082] The input / output interface control bridge is configured to manage data exchange between electronic devices and external hardware peripherals. The input / output interface control bridge connects to various communication interface adapters. These adapters include Gigabit Ethernet network interface cards, Fibre Channel host bus adapters, and multi-channel serial communication cards. The Gigabit Ethernet network interface cards are configured to connect to the gas sensor network and airflow monitoring devices via a downhole industrial Ethernet ring network. The Fibre Channel host bus adapters are configured to connect to the core switch of the downhole positioning system. The multi-channel serial communication cards provide RS-485 fieldbus interfaces and controller area network interfaces for directly acquiring or debugging specific low-level sensor node data.

[0083] The electronic equipment further includes a human-machine interface (HMI) display terminal. The HMI display terminal is connected to the input / output interface control bridge via a video graphics array interface or a high-definition multimedia interface. The HMI display terminal is configured to display a 3D topology map interface of the central control system, real-time gas concentration curves, and alarm status information to the operator. The electronic equipment is connected to an industrial keyboard and industrial mouse for receiving threshold adjustment commands and parameter update commands input by management personnel.

[0084] The mainboard of the electronic equipment integrates a hardware watchdog timer. This hardware watchdog timer is configured to force a reset of the electronic equipment in the event of a deadlock or program crash in the central processing unit, ensuring the continuous and stable operation of the underground safety monitoring system. The electronic equipment employs a fanless, enclosed metal chassis structure with an outer shell that meets industrial dustproof and waterproof standards to withstand the harsh physical environment of a coal mine monitoring room.

[0085] Computer program instructions are stored on a computer-readable storage medium. When the computer program instructions are loaded and executed by the central processing unit, they implement the various steps of the downhole hazardous gas detection and correction method based on vehicle emission monitoring described in the foregoing embodiments of the present invention, including synchronous acquisition and preprocessing of multi-source heterogeneous data, geometric determination of spatial airflow coupling relationship, calculation and matching of dynamic lag time, quantitative prediction and modeling of interference concentration, and signal correction and restoration of true concentration.

[0086] This invention provides a computer-readable storage medium storing computer program instructions. When the computer program instructions are read and executed by the central processing unit of an electronic device, the electronic device causes the electronic device to implement various functions of the downhole hazardous gas detection and correction method based on vehicle emission monitoring.

[0087] The computer-readable storage medium specifically stores a multi-source heterogeneous data synchronization processing instruction set. When executed by the central processing unit (CPU), the instruction set first initializes multiple independent circular buffers in system memory, which are used to map the I / O address spaces of the vehicle emission monitoring module, the well positioning system, the airflow monitoring device, and the gas sensor network, respectively. The instruction set controls the CPU to parse the incoming raw data packets in real time and extract the timestamp field from the packet header. It then calls a high-precision timer in the system kernel to correct clock skew for data from different sources. To address the issue of inconsistent sampling frequencies, the instruction set uses linear interpolation functions from the mathematical computation library to calculate the interpolation points of the vehicle's spatial coordinates at each moment, based on the time series of the fixed gas sensors, thereby constructing a synchronization state matrix that is strictly aligned in the time dimension.

[0088] A computer-readable storage medium stores a spatial airflow coupling determination instruction set. This instruction set is configured to perform three-dimensional vector geometry operations. When the central processing unit (CPU) executes the instruction set, it reads the vehicle coordinate vector and sensor coordinate vector from the synchronization state matrix and performs vector subtraction to generate a relative position vector. The instruction set calls an inverse cosine function to calculate the angle between the airflow vector and the relative position vector. The instruction set includes logical decision-making branch code that compares the calculated cosine value of the angle with a preset angle tolerance threshold. When the determination result is true, the CPU marks the vehicle at the current moment as a valid interference source; when the determination result is false, the CPU sets the corresponding flow field coupling factor register to zero, thereby achieving software-level pruning and optimization of invalid calculation paths.

[0089] A computer-readable storage medium stores a dynamic hysteresis compensation instruction set. This instruction set includes a physical transmission time calculation subroutine and a historical data backtracking subroutine. The physical transmission time calculation subroutine causes the central processing unit (CPU) to perform floating-point division, dividing the Euclidean distance between the vehicle and the sensor by the current wind speed to calculate the air mass flight time. The historical data backtracking subroutine controls the CPU to access the historical data storage area and, based on the time index obtained by subtracting the total hysteresis time from the current time, retrieves past vehicle emission intensity records. The dynamic hysteresis compensation instruction set, through a weighted averaging algorithm, synthesizes vehicle emission source data that accurately corresponds to the moment of air mass formation, thus resolving the spatiotemporal misalignment problem caused by wind transport.

[0090] The computer-readable storage medium stores an instruction set for interference concentration prediction and correction. The kernel of this instruction set encapsulates the mathematical expression of a Gaussian plume diffusion model. When the central processing unit executes the instruction set, it indexes the corresponding horizontal and vertical diffusion coefficients from a diffusion parameter lookup table based on the currently calculated distance parameters. The instruction set calls an exponential function library to calculate the spatial decay distribution of the exhaust gas concentration and outputs the theoretical interference concentration value. The instruction set further includes a signal correction algorithm that calculates dynamic confidence weights and performs a weighted differential operation to subtract the theoretical interference concentration from the fixed sensor observation data. The instruction set also includes a smoothing filter code segment for denoising the corrected real concentration data.

[0091] A computer-readable storage medium stores an intelligent alarm and self-learning instruction set. This instruction set is configured to implement dual threshold logic. When the central processing unit executes the intelligent alarm and self-learning instruction set, it simultaneously loads a safe concentration threshold, fixed sensor observation data, and a corrected real-world gas concentration. The instruction set executes comparison instructions, writing an alarm trigger bit to the output interface register only when both the real-world gas concentration and the fixed sensor observation data simultaneously exceed the safe concentration threshold. Furthermore, the intelligent alarm and self-learning instruction set includes a parameter inversion optimization algorithm. In system idle or specific calibration modes, the instruction set utilizes gradient descent to automatically adjust the diffusion constant in the Gaussian model based on historical observation data, and writes the updated parameters to non-volatile memory, achieving adaptive model evolution at the software level.

Claims

1. A method for detecting and correcting harmful gases in wells based on vehicle emission monitoring, characterized in that, Includes the following steps: S1. Acquire vehicle emission source data collected by the vehicle emission monitoring module, vehicle spatial location data collected by the underground positioning system, airflow field data collected by the airflow field monitoring device, and fixed sensor observation data collected by the gas sensor network, respectively. S2. Perform time synchronization and alignment processing on the vehicle emission source data, vehicle spatial location data, airflow field data, and fixed sensor observation data. S3. Based on the synchronized data, calculate the flow field coupling factor of the underground auxiliary transport vehicle to the fixed gas sensor, and combine the airflow field data and vehicle spatial position data to calculate the transmission lag time of the vehicle exhaust gas to the fixed gas sensor. S4. Based on the vehicle emission source data, transmission lag time, airflow field data and diffusion parameters, construct an interference prediction model and calculate the theoretical interference concentration generated by the underground auxiliary transport vehicle at the fixed gas sensor at the current moment. S5. Calculate the dynamic confidence weighting coefficient based on the airflow field data and vehicle spatial position data, and calculate the corrected real ambient gas concentration based on the fixed sensor observation data, theoretical interference concentration and the dynamic confidence weighting coefficient. S6. Compare the actual ambient gas concentration with the preset safe concentration threshold, and execute alarm triggering or alarm suppression control based on the comparison result.

2. The method for detecting and correcting harmful gases in wells based on vehicle emission monitoring according to claim 1, characterized in that, The time synchronization and alignment processing of the vehicle emission source data, vehicle spatial location data, airflow field data, and fixed sensor observation data in step S2 specifically includes: A circular data buffer corresponding to different data sources is established in memory, and the time series of the fixed sensor observation data is used as the main time reference axis. Using the sampling timestamp of the fixed sensor observation data as the index key, retrieve data frames of adjacent recording times in the circular data buffer of the vehicle spatial location data and vehicle emission source data; A linear interpolation algorithm is performed on the data frames of the adjacent recording times to calculate the instantaneous vehicle position, instantaneous vehicle speed, and vehicle emission intensity that strictly correspond to the sampling timestamp; The nearest neighbor matching method is used to select the airflow field data record with the smallest time interval between the sampling timestamp.

3. The method for detecting and correcting harmful gases in wells based on vehicle emission monitoring according to claim 1, characterized in that, The calculation of the flow field coupling factor between the downhole haulage vehicle and the stationary gas sensor in step S3 specifically includes: Construct a relative position vector pointing from the vehicle spatial position data to the fixed gas sensor position, and obtain the wind flow field vector from the airflow field data; Calculate the cosine of the angle between the airflow field vector and the relative position vector; The cosine value of the included angle is compared with a preset angle tolerance threshold. If the cosine value of the included angle is greater than the angle tolerance threshold, it is determined that the fixed gas sensor is located in the effective downwind area, and a non-zero flow field coupling factor is generated. If the cosine value of the included angle is less than or equal to the angle tolerance threshold, it is determined to be an invalid interference condition, and the flow field coupling factor is set to zero.

4. The method for detecting and correcting harmful gases in wells based on vehicle emission monitoring according to claim 1, characterized in that, The calculation of the transmission lag time from vehicle exhaust gas to the fixed gas sensor in step S3 specifically includes: Calculate the Euclidean distance between the vehicle's spatial position data and the fixed gas sensor position, and obtain the wind flow field vector magnitude in the airflow field data; Divide the Euclidean distance by the magnitude of the wind flow field vector to obtain the physical transmission lag time; Obtain the sensor response time constant of the fixed gas sensor, and add the physical transmission lag time to the sensor response time constant to obtain the total transmission lag time; The target historical time is obtained by subtracting the total transmission lag time from the current system time, and the historical vehicle emission intensity corresponding to the target historical time is back-matched in the vehicle emission source data.

5. The method for detecting and correcting harmful gases in wells based on vehicle emission monitoring according to claim 4, characterized in that, The calculation of the theoretical interference concentration generated by the downhole auxiliary transport vehicle at the fixed gas sensor at the current moment, as described in step S4, specifically includes: Based on the straight-line distance between the vehicle's spatial location data and the location of the fixed gas sensor, the horizontal and vertical diffusion parameters are determined. Calculate the vertical deviation distance of the fixed gas sensor position relative to the axis with the direction of the airflow field vector as the axis; The flow field coupling factor, the diffusion regulation constant in the tunnel environment, the historical vehicle emission intensity, the wind flow field vector magnitude, the horizontal diffusion parameters, the vertical diffusion parameters, and the vertical deviation distance are substituted into the interference prediction model. The interference prediction model is configured to perform the following calculation logic: the theoretical interference concentration is directly proportional to the flow field coupling factor, the diffusion regulation constant and the historical vehicle emission intensity, inversely proportional to the product of the wind flow field vector magnitude and the diffusion parameter, and decreases exponentially with the increase of the vertical deviation distance.

6. The method for detecting and correcting harmful gases in wells based on vehicle emission monitoring according to claim 1, characterized in that, The calculation of the dynamic confidence weight coefficient in step S5 specifically includes: Extract airflow field data within a historical time window, calculate the standard deviation of airflow velocity modulus and the standard deviation of airflow direction angle to assess airflow field stability; The jitter frequency of the vehicle spatial location data is analyzed to assess the reliability of the vehicle spatial location. By using a normalized mapping function to integrate airflow field stability and vehicle spatial position confidence, a dynamic confidence weight coefficient with values ​​ranging from 0 to 1 is generated. When the airflow field is in an unsteady turbulent mode or when the vehicle spatial location data is subject to multipath interference, the value of the dynamic confidence weight coefficient is reduced.

7. The method for detecting and correcting harmful gases in wells based on vehicle emission monitoring according to claim 1, characterized in that, The calculation of the corrected real ambient gas concentration in step S5 specifically includes: The theoretical interference concentration is weighted using the dynamic confidence weighting coefficient to obtain the weighted interference component; The true ambient gas concentration is obtained by subtracting the weighted interference component from the data observed by the fixed sensor. Physical constraint boundary checks are performed on the actual ambient gas concentration. If the calculation result is negative, the actual ambient gas concentration is clamped to zero or clamped to the historical baseline value. An exponentially weighted moving average filter is used to smooth the actual ambient gas concentration.

8. The method for detecting and correcting harmful gases in wells based on vehicle emission monitoring according to claim 1, characterized in that, The step S6, which describes performing alarm triggering or alarm suppression control based on the comparison results, specifically includes: When the data observed by the fixed sensor is greater than or equal to the safe concentration threshold, and the actual ambient gas concentration is less than the safe concentration threshold, it is determined to be a vehicle interference event. An alarm suppression control signal is generated to lock the audible and visual alarm and the power-off controller, and the vehicle interference event is recorded. When the data observed by the fixed sensor is greater than or equal to the safe concentration threshold, and the actual ambient gas concentration is greater than the safe concentration threshold, it is determined to be a real gas over-limit event, and an alarm trigger control signal is generated to drive the audible and visual alarm and the power-off controller.

9. The method for detecting and correcting harmful gases in wells based on vehicle emission monitoring according to claim 5, characterized in that, It also includes an online adaptive update step for model parameters: Parameter calibration observation window is established when only one underground auxiliary transport vehicle passes by at a constant speed and the background gas concentration is stable. Extract the measured vehicle emission interference characteristic curves within the parameter calibration observation window; A parameter optimization objective function is constructed, wherein the parameter optimization objective function is the root mean square error between the measured vehicle emission interference characteristic curve and the theoretical interference concentration curve calculated based on the current model parameters; An iterative optimization algorithm is used to adjust the diffusion regulation constant and the horizontal diffusion parameter until the parameter optimization objective function converges to the minimum value, and the updated parameters are written into the dynamic parameter database.

10. A downhole hazardous gas detection and correction system based on vehicle emission monitoring, characterized in that, The method for detecting and correcting harmful gases in wells based on vehicle emission monitoring, as described in any one of claims 1-9, is used in a vehicle emission monitoring module configured to detect the composition and concentration of exhaust gas emitted by underground auxiliary vehicles in real time and generate vehicle emission source data. The underground positioning system is configured to acquire the three-dimensional spatial coordinates of the underground auxiliary transport vehicle and generate vehicle spatial position data. The airflow field monitoring device is configured to monitor the airflow velocity and direction in underground roadways in real time and generate airflow field data. A gas sensor network, consisting of fixed gas sensors arranged in the monitoring area, is configured to generate fixed sensor observation data; The data acquisition and processing module is configured to receive the above data and execute the downhole harmful gas detection correction method based on vehicle emission monitoring as described in any one of claims 1 to 9, and output the real ambient gas concentration; The central control system is configured to receive the actual ambient gas concentration and control the start and stop of the on-site alarm equipment based on the comparison result between the actual ambient gas concentration and the safe concentration threshold.

Images