A non-contact gas leak detection method and system

Abstract

The present application relates to the technical field of gas detection, in particular to a non-contact gas leakage detection method and system, the method provided by the present application mainly includes configuring a refrigeration type infrared detector, environment map establishment and path planning, autonomously performing fixed point inspection navigation and gas imaging scanning, image preprocessing and gas identification, leakage source positioning and leakage rate calculation. The present application detects the leakage of typical colorless and transparent gas medium, and realizes automatic unmanned inspection from the aspect of automatic algorithm. Secondly, remote intelligent monitoring can be realized, the monitoring range is large, and the application range is wide. The purpose of improving the gas leakage detection positioning efficiency and calculating the leakage rate is to further improve the safety of the device.

Description

Technical Field

[0001] This invention relates to the field of gas detection technology, and more specifically, to a non-contact gas leak detection method and system. Background Technology

[0002] Pipeline transmission systems using new media place higher demands on leak detection. A leak can lead not only to economic losses but also, in severe cases, safety accidents. Currently, most pipeline leak measurements are done visually, with estimations made for minor leaks. However, for transmission systems using colorless and transparent gases as the medium, leaks are invisible to the naked eye, and due to the high temperature, high pressure, and gaseous properties of the medium, human observation poses safety risks.

[0003] Invention patent CN 107121238 A discloses a method for detecting high-altitude gas leaks in high-temperature gas pipelines. This method uses an aircraft carrying a thermal imager to photograph the high-temperature gas pipeline to determine the leak location and area. It also simulates the ambient temperature of the pipeline to determine the heat of the leak. Invention patent CN 111562055 A discloses a helium leak detection device and method based on ultrasonic time-of-flight. This includes a detection box, a signal processing and control module, and two sets of ultrasonic sensors. Helium is used as a tracer gas, and the concentration of leaked helium is indirectly measured by measuring the flight time of the ultrasonic waves within the detection box. While these methods can achieve contact-type gas leak detection under certain conditions, the use of pressure flow meters and acoustic sensors presents challenges such as difficulty in accurately locating the leak, high detection cost per unit area, and limited detection flexibility. Summary of the Invention

[0004] The purpose of this invention is to provide a non-contact gas leak detection method and system to solve the problems existing in the prior art.

[0005] This invention is achieved through the following technical solution: A non-contact gas leak detection method, comprising: Configure and calibrate a gas-cooled infrared detector for the corresponding band; Acquire environmental data, build maps based on the environmental data, and use the maps for route planning; Based on path planning, signals for autonomous point inspection and navigation are sent, and gas imaging scanning is performed when the fixed inspection position is reached. The scanned images are preprocessed, and the obtained gas data is used for gas identification. Based on the results of image preprocessing and gas identification, the leak source is located and the leak rate is calculated.

[0006] Preferably, the configuration calibration of the corresponding band gas-cooled infrared detector includes: Infrared detectors matching the incident light intensity are selected using a gas absorption model of infrared radiation.

[0007] In the formula, The intensity of transmitted light. For the incident light intensity, The gas absorption coefficient is... For gas concentration, This is the optical path length.

[0008] Preferably, the route planning via map includes: A point cloud map of the inspection area is generated by combining laser vision with SLAM, and high-risk points of the target are marked. Reinforcement learning algorithms are used to plan the optimal path, avoiding obstacles and covering key areas.

[0009] Preferably, the gas imaging scan includes: Acquire GPS data, visual LiDAR datasets, and IMU data to locate the robot's position; Send control signals to the robot to make it reach the fixed inspection position; The area at a fixed inspection location is imaged and scanned by carrying a gimbal.

[0010] Preferably, the preprocessing of the scanned image and the gas identification of the obtained gas data include: Infrared and visible light images are aligned using the Mutual Information Maximization (MI) algorithm. Gas region extraction based on dual-spectral difference method; By improving the YOLOv8 model, gas leak areas can be captured.

[0011] Preferably, the step of aligning infrared and visible light images using the Mutual Information Maximization (MI) algorithm includes:

[0012] In the formula, For spatial transformation parameters, Let be the affine transformation matrix. Infrared image, Visible light image, , For image coordinates, In a given transformation Under the given conditions, the joint probability distribution of the intensity of the two images, The edge probability distribution of the infrared image. This represents the edge probability distribution of a visible light image.

[0013] Preferably, the gas extraction region based on dual-spectral difference method includes:

[0014] In the formula, This is a binary graph of the gas. For an infrared image frame at time t, For a visible light image frame at time t, Infrared differential threshold, This is the visible light differential threshold.

[0015] Preferably, the process of locating the leak source and calculating the leak rate based on the results of image preprocessing and gas identification includes: The gas pixel particle swarm position is updated based on the plume tracking algorithm, converges to the leakage source, and the effective cross-sectional area of ​​the gas plume in the infrared image is calculated. After the robot detects a leak, it measures the gas flow in several directions to identify the leak source and calculate the gas wind speed at the location. By combining gas wind speed and effective cross-sectional area, the leakage rate is estimated by fitting a typical gas leakage cloud.

[0016] Preferably, the leakage rate calculation includes:

[0017] In the formula, Leakage rate, The total conversion factor is... For gas wind speed, This represents the effective cross-sectional area of ​​the gas plume in the infrared image. The apparent temperature difference between the gas plume and the background environment. This is the temperature-to-concentration conversion coefficient. The heat exchange coefficient, The thickness of the smoke plume.

[0018] Secondly, the present invention also provides a non-contact gas leak detection system for performing the above-described non-contact gas leak detection method, comprising: The data processing module is configured to configure and calibrate the gas-cooled infrared detector in the corresponding band; acquire environmental data and build a map based on the environmental data, and perform path planning through the map; send autonomous point inspection and navigation signals based on the path planning, and perform gas imaging scanning when it reaches the fixed inspection position; The identification module is configured to preprocess the scanned images, identify the gas data, and locate the leak source and calculate the leak rate based on the results of image preprocessing and gas identification.

[0019] The technical solution of the present invention has at least the following advantages and beneficial effects: The method provided by this invention mainly includes configuring a cooled infrared detector, establishing an environmental map and planning a path, autonomously performing fixed-point inspection navigation and gas imaging scanning, image preprocessing and gas identification, and locating the leak source and calculating the leak rate. This invention detects leaks in typical colorless and transparent gaseous media and, through automated algorithms, achieves automated, unmanned inspection. Furthermore, it enables remote intelligent monitoring with a large monitoring range and wide applicability. By improving the efficiency of gas leak detection and location and calculating the leak rate, it further enhances the safety of the device. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the process of the present invention; Figure 2 This is a hardware composition diagram of the present invention; Figure 3 This is a flowchart illustrating the execution of the detection algorithm of the present invention; Figure 4 This is a schematic diagram illustrating the leakage location and calculation of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of 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. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0023] The module division in this application is a logical division. In actual application, there may be other division methods. For example, multiple modules may be combined into or integrated into another system, or some features may be ignored or not executed.

[0024] Furthermore, the connection, coupling, or communication in this application can be a direct connection, coupling, or communication between related objects, or an indirect connection, coupling, or communication through other devices. Moreover, the connection, coupling, or communication between objects can be electrical or other similar forms, and this application does not impose any limitations on these. Independently described modules or sub-modules may or may not be physically separated; they may be implemented in software or hardware, and some modules or sub-modules may be implemented in software, with the processor calling the software to implement the function of these modules or sub-modules, while other modules or sub-modules may be implemented in hardware, such as through hardware circuits. Furthermore, some or all of the modules can be selected to achieve the purpose of this application's solution according to actual needs.

[0025] Please refer to Figures 1-2 The present invention provides a non-contact gas leak detection method, comprising: S101: Configure and calibrate a gas-cooled infrared detector for the corresponding band; To configure and calibrate a gas-cooled infrared detector for the corresponding band, first, by matching the typical gas leakage spectral properties, a mid-wave infrared detector is selected and equipped with a portable cooling module to capture minute temperature difference data.

[0026] Preferably, using a mid-wave infrared detector means selecting an appropriate wavelength range and measurement distance, and choosing an infrared detector with matching incident light intensity based on the gas absorption formula for infrared radiation.

[0027] In the formula, The intensity of transmitted light. For the incident light intensity, The gas absorption coefficient is... For gas concentration, This is the optical path length.

[0028] Preferably, the camera is equipped with a cooling Stirling engine to reduce the temperature of the detector chip, thereby reducing the detector noise to below that of the imaging scene and increasing the temperature difference ΔT between the leaked gas and the background, resulting in a difference in radiation intensity.

[0029] S102: Acquire environmental data, build a map based on the environmental data, and perform route planning using the map; A point cloud map of the inspection area is generated by combining laser vision with SLAM, and high-risk points (pipe joints, valves) are marked. Reinforcement learning algorithms are used to plan the optimal path, avoiding obstacles and covering key areas.

[0030] Specifically, regarding path planning, problems in the physical world are mapped to standard elements of RL.

[0031] 1. State (S): Robot's own state: position (x, y), orientation (θ), velocity (v), angular velocity (ω). Environmental perception state: obstacle information: perception data from LiDAR, depth camera, or map. For example, a distance vector [d1, d2, ..., d...]. n [] represents the distance to the nearest obstacle in n surrounding directions. Coverage Status: This is the core of achieving "coverage of critical areas". An internal map representation is needed to record which areas have been visited / covered. For example, a 2D grid corresponding to the global map, where each cell has a value (0 = uncovered, 1 = covered, or a decaying coverage value). Target Information: The relative position or ID of the next "critical area" to be covered, or a global "coverage" progress scalar.

[0032] 2. Action (a) Depending on the robot platform, for a differential wheeled robot, it might be [linear velocity, angular velocity]. For an omnidirectional wheeled robot, it might be [v_x, v_y, ω]. The action space needs to be either continuous or discretized. Reward function (Reward, r): Constructed to balance multiple sub-objectives. Coverage reward: A large positive reward is given whenever the robot enters a new, uncovered critical area cell. This is the primary signal driving coverage behavior. Obstacle avoidance penalty: A negative reward is given when the robot approaches an obstacle, inversely proportional to the distance. If a collision occurs, a very large negative reward is given and the training round terminates. Efficiency / survival reward: A small negative reward (e.g., -0.1) is given for each step taken, encouraging the robot to complete the coverage using the shortest path and avoid meaningless wandering. Completion reward: A large positive reward is given when all critical areas are covered, or when the coverage exceeds a certain threshold. Exploration reward (optional, used to improve efficiency): To encourage exploration of unknown areas, an intrinsic curiosity mechanism can be introduced, giving extra small rewards for states with low access frequency.

[0033] 3. Termination conditions: Success: All critical areas are covered. Failure: A collision occurs. Timeout: The number of exploration steps in a single attempt exceeds the maximum limit (to prevent the agent from getting stuck).

[0034] S103: Sends autonomous point inspection and navigation signals based on path planning, and performs gas imaging scanning when it reaches the fixed inspection position; S104: Preprocess the scanned image and perform gas identification on the acquired gas data; S105: Locate the leak source and calculate the leak rate based on the results of image preprocessing and gas identification.

[0035] The method provided by this invention mainly includes configuring a cooled infrared detector, establishing an environmental map and planning a path, autonomously performing fixed-point inspection navigation and gas imaging scanning, image preprocessing and gas identification, and locating the leak source and calculating the leak rate. This invention detects leaks in typical colorless and transparent gaseous media and, through automated algorithms, achieves automated, unmanned inspection. Furthermore, it enables remote intelligent monitoring with a large monitoring range and wide applicability. By improving the efficiency of gas leak detection and location and calculating the leak rate, it further enhances the safety of the device. In one exemplary embodiment of the present invention, performing a gas imaging scan includes: S201: Acquire GPS data, visual LiDAR dataset, and IMU data to locate the robot's position; Time synchronization and sensor calibration, front-end data processing, IMU: pre-integration, compensation for Earth's rotation and Coriolis force, LiDAR: feature extraction, motion distortion removal, vision: feature extraction and tracking, descriptor calculation, GPS: coordinate transformation (WGS84 → local ENU), quality inspection, state estimation and fusion, back-end optimization and mapping, pose graph optimization: GPS and loop closure provide global constraints, sub-map management: hierarchical mapping reduces computation, global consistency: when GPS signal is restored, corrects accumulated errors.

[0036] S202: Send a control signal to the robot to make it reach the fixed inspection position; S203: Performs imaging scans of areas at fixed inspection locations by carrying a gimbal.

[0037] Look for low-temperature dark areas (gas temperature < background) or high-temperature bright areas (gas temperature > background).

[0038] In one exemplary embodiment of the present invention, preprocessing the scanned image and performing gas identification on the obtained gas data includes: Infrared and visible light images are aligned using the Mutual Information Maximization (MI) algorithm. Gas region extraction based on dual-spectral difference method; By improving the YOLOv8 model, gas leak areas can be captured.

[0039] Specifically, the mutual information maximization (MI) algorithm for aligning infrared and visible light images includes:

[0040] In the formula, For spatial transformation parameters, Let be the affine transformation matrix. Infrared image, Visible light image, , For image coordinates, In a given transformation Under the given conditions, the joint probability distribution of the intensity of the two images, The edge probability distribution of the infrared image. This represents the edge probability distribution of a visible light image.

[0041] Gas extraction regions based on dual-spectral difference method include:

[0042] In the formula, This is a binary graph of the gas. For an infrared image frame at time t, For a visible light image frame at time t, Infrared differential threshold, This is the visible light differential threshold.

[0043] An exemplary embodiment of the present invention, which involves locating the leak source and calculating the leak rate based on the results of image preprocessing and gas identification, includes: The gas pixel particle swarm position is updated based on the plume tracking algorithm, converges to the leakage source, and the effective cross-sectional area of ​​the gas plume in the infrared image is calculated. After the robot detects a leak, it measures the gas flow in several directions to identify the leak source and calculate the gas wind speed at the location. By combining gas wind speed and effective cross-sectional area, the leakage rate is estimated by fitting a typical gas leakage cloud.

[0044] Leakage rate calculation includes:

[0045] In the formula, Leakage rate, The total conversion factor is... For gas wind speed, This represents the effective cross-sectional area of ​​the gas plume in the infrared image. The apparent temperature difference between the gas plume and the background environment. This is the temperature-to-concentration conversion coefficient. The heat exchange coefficient, The thickness of the smoke plume.

[0046] Secondly, the present invention also provides a non-contact gas leak detection system for performing the above-described non-contact gas leak detection method, comprising: The data processing module is configured to configure and calibrate the gas-cooled infrared detector in the corresponding band; acquire environmental data and build a map based on the environmental data, and perform path planning through the map; send autonomous point inspection and navigation signals based on the path planning, and perform gas imaging scanning when it reaches the fixed inspection position; The identification module is configured to preprocess the scanned images, identify the gas data, and locate the leak source and calculate the leak rate based on the results of image preprocessing and gas identification.

[0047] This invention utilizes a cooled infrared camera to effectively leverage the sensitivity of typical gases in specific wavelength bands, minimizing the impact of thermal fields and enabling 24-hour monitoring of high and low temperature gases.

[0048] By installing it on a quadruped mobile device, the ability to pass through and the detection angle are improved. Through non-contact detection, a wide range of monitoring can be achieved for multiple leak points on different floors.

[0049] After image processing alignment and differentiation, the improved intelligent detection algorithm is used to capture the leakage area, which effectively enhances image features and improves the accuracy of visual detection.

[0050] By combining a contact-type anemometer, it is possible to fit and estimate typical gas leak clouds.

[0051] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0052] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. This computer software product, stored in a storage medium, includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0053] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A non-contact gas leak detection method, characterized in that, include: Configure and calibrate a gas-cooled infrared detector for the corresponding band; Acquire environmental data, build maps based on the environmental data, and use the maps for route planning; Based on path planning, signals for autonomous point inspection and navigation are sent, and gas imaging scanning is performed when the fixed inspection position is reached. The scanned images are preprocessed, and the obtained gas data is used for gas identification. Based on the results of image preprocessing and gas identification, the leak source is located and the leak rate is calculated.

2. The non-contact gas leak detection method according to claim 1, characterized in that, The configuration calibration corresponding band gas-cooled infrared detector includes: Infrared detectors matching the incident light intensity are selected using a gas absorption model of infrared radiation. In the formula, The intensity of transmitted light. For the incident light intensity, The gas absorption coefficient is... For gas concentration, This is the optical path length.

3. The non-contact gas leak detection method according to claim 2, characterized in that, The route planning via map includes: A point cloud map of the inspection area is generated by combining laser vision with SLAM, and high-risk points of the target are marked. Reinforcement learning algorithms are used to plan the optimal path, avoiding obstacles and covering key areas.

4. The non-contact gas leak detection method according to claim 3, characterized in that, The gas imaging scan includes: Acquire GPS data, visual LiDAR datasets, and IMU data to locate the robot's position; Send control signals to the robot to make it reach the fixed inspection position; The area at a fixed inspection location is imaged and scanned by carrying a gimbal.

5. The non-contact gas leak detection method according to claim 3, characterized in that, The preprocessing of the scanned image and the gas identification of the obtained gas data include: Infrared and visible light images are aligned using the Mutual Information Maximization (MI) algorithm. Gas region extraction based on dual-spectral difference method; By improving the YOLOv8 model, gas leak areas can be captured.

6. The non-contact gas leak detection method according to claim 5, characterized in that, The method of aligning infrared and visible light images using the mutual information maximization algorithm (MI) includes: In the formula, For spatial transformation parameters, Let be the affine transformation matrix. Infrared image, Visible light image, , For image coordinates, In a given transformation Under the given conditions, the joint probability distribution of the intensity of the two images, The edge probability distribution of the infrared image. This represents the edge probability distribution of a visible light image.

7. The non-contact gas leak detection method according to claim 6, characterized in that, The gas extraction region based on dual-spectral difference method includes: In the formula, This is a binary graph of the gas. For an infrared image frame at time t, For a visible light image frame at time t, Infrared differential threshold, This is the visible light differential threshold.

8. The non-contact gas leak detection method according to claim 7, characterized in that, The process of locating the leak source and calculating the leak rate based on the results of image preprocessing and gas identification includes: The gas pixel particle swarm position is updated based on the plume tracking algorithm, converges to the leakage source, and the effective cross-sectional area of ​​the gas plume in the infrared image is calculated. After the robot detects a leak, it measures the gas flow in several directions to identify the leak source and calculate the gas wind speed at the location. By combining gas wind speed and effective cross-sectional area, the leakage rate is estimated by fitting a typical gas leakage cloud.

9. A non-contact gas leak detection method according to claim 8, characterized in that, The leakage rate calculation includes: In the formula, Leakage rate, The total conversion factor is... For gas wind speed, This represents the effective cross-sectional area of ​​the gas plume in the infrared image. The apparent temperature difference between the gas plume and the background environment. This is the temperature-to-concentration conversion coefficient. The heat exchange coefficient, The thickness of the smoke plume.

10. A non-contact gas leak detection system, characterized in that, A non-contact gas leak detection method according to any one of claims 1-9, characterized in that it comprises: The data processing module is configured to configure and calibrate the gas-cooled infrared detector in the corresponding band; acquire environmental data and build a map based on the environmental data, and perform path planning through the map; send autonomous point inspection and navigation signals based on the path planning, and perform gas imaging scanning when it reaches the fixed inspection position; The identification module is configured to preprocess the scanned images, identify the gas data, and locate the leak source and calculate the leak rate based on the results of image preprocessing and gas identification.

Images