Leakage source positioning method and device based on multi-source gas data fusion

Abstract

The invention provides a leakage source positioning method and device for multi-source gas data fusion, and relates to the technical field of sensor systems.The method comprises the steps that multi-source data collected by a sensor system arranged in a target area is obtained, the sensor system comprises a plurality of gas sensors and a plurality of environment sensors, and the gas sensors are arranged in the target area; the gas sensor is used for collecting gas concentration data, and the environment sensor is used for collecting physical environment data; the gas concentration data, the physical environment data and the leakage process of the leaked gas are fused for inversion calculation; and based on the leakage process, the leakage source position is determined and mapped back to the physical structure of the gas pipe network. According to the invention, the gas leakage position can be positioned, so that the disposal efficiency of gas leakage detection is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sensor systems, in particular to a multi-source gas data fusion leak source positioning method and device. BACKGROUND

[0002] Household gas leakage detection mainly monitors the concentration changes of methane, propane and other combustible gases in indoor air in real time by deploying high-sensitivity gas sensors, identifies leakage events in combination with threshold judgment mechanism, triggers sound-light alarms and links to close electromagnetic valves to cut off gas supply once the detection value exceeds the preset safety concentration threshold, and can report the leakage alarm information to the user terminal or property platform through the wireless communication module to realize rapid response and remote intervention and protect the safety of life and property of users.

[0003] Firstly, at the gas input end, if the gas meter air inlet interface seal fails, the connection is loose, or the pipeline deforms due to foundation settlement, small or intermittent leaks will occur. Secondly, in the gas delivery path, metal pipes may leak due to long-term stress concentration, thermal expansion and contraction, corrosion perforation, or third-party construction disturbance; and soft hose connections have the risks of pipe aging, external force pulling, connection loosening, or interface seal ring failure, especially non-stainless steel corrugated pipes or PVC hoses are more vulnerable. Thirdly, in terms of connection components, such as tees, elbows, valves, quick couplings, etc., if the sealing materials such as rubber gaskets fail due to aging, material incompatibility or improper installation, they will become high-risk points for leakage. At the use terminal, if the gas stove, water heater, wall-mounted furnace, etc. have problems such as gas path system rupture, abnormal opening of the ignition device, failure of the flameout protection, etc., they may also continuously leak gas without obvious sound.

[0004] The existing household gas leakage detection technology generally uses fixed-point gas sensors, which can only monitor the concentration of combustible gases in the air at fixed points and determine whether there is gas leakage through threshold triggering, lacks the ability to perceive the spatial location of the leakage source, and thus although it can issue an alarm in the event of a leak, it cannot provide the specific location of the gas leak, and users need to check each place themselves in the presence of safety hazards, which is not only time-consuming and labor-intensive, but also easy to cause misjudgment and secondary risks, seriously restricting the practicality and disposal efficiency of the detection system. Therefore, a method is needed to locate the position of the gas leak to improve the disposal efficiency of the gas leak detection. SUMMARY

[0005] The present application provides a multi-source gas data fusion leak source positioning method and device, which can locate the position of the gas leak to improve the disposal efficiency of the gas leak detection.

[0006] In a first aspect of the present application, a multi-source gas data fusion leak source positioning method is provided, the method comprising: acquire multi-source data collected by a sensor system deployed in a target area, wherein the sensor system comprises a plurality of gas sensors for collecting gas concentration data and a plurality of environmental sensors for collecting physical environment data; fuse the gas concentration data and the physical environment data, and perform inversion calculation on a leakage path of the leaked gas; determine a leakage source position based on the leakage process and map the leakage source position back to a physical structure of the gas pipe network.

[0007] On the basis of the above technical solutions, preferably, the acquired multi-source data collected by the sensor system deployed in the target area specifically comprises: deploy a plurality of gas sensors according to a gas structure topology map, wherein nodes of the gas structure topology map correspond to gas source positions, gas transmission paths, gas connection position and gas use terminal positions; deploy a plurality of environmental sensors, wherein the environmental sensors comprise temperature sensors, humidity sensors, air pressure sensors and air flow velocity sensors, and are respectively used to collect temperature data, humidity data, air pressure data and gas flow velocity data; establish a structured index of time stamps, spatial positions, sensor types and sampling data according to the data collected by the plurality of gas sensors and the plurality of environmental sensors, and form an original data set with complete space-time attributes.

[0008] On the basis of the above technical solutions, preferably, the fusion of the gas concentration data and the physical environment data is used to perform inversion calculation on the leakage path of the leaked gas, and specifically comprises: calculate gas concentration change data according to the gas concentration data in the original data set; construct a gas diffusion physical model with the gas concentration change data as a core observation, wherein the gas diffusion physical model takes a three-dimensional space grid as a basic unit, divides the target area into a regular voxel grid structure, takes temperature data, humidity data, air pressure data and gas flow velocity data as state variables of each grid node in the grid, and describes gas diffusion behavior by establishing a partial differential equation; In the process of constructing the gas diffusion physical model, the local gas molecular mean free path and the thermal motion coefficient are determined by the temperature data, the humidity data and the air pressure data, so as to calculate the diffusion coefficient of each grid node, the local air flow field is solved according to the air flow velocity data, the flow velocity vector in each space unit is obtained, the diffusion coefficient and the flow velocity vector are input into the partial differential equation, and the first evolution sequence of the theoretical concentration field is obtained by numerical solution.

[0009] On the basis of the above technical solutions, preferably, the leakage source position is determined based on the leakage process and mapped back to the physical structure of the gas pipe network, and specifically comprises: differences between the gas concentration change data and the first evolution sequence are calculated to construct a residual vector sequence; A gradient descent method based on a back propagation mechanism is used to perform multiple rounds of gradient updates on parameters of a source term function of the partial differential equation, wherein the source term function is used to describe time-varying estimated values of the leakage source position and intensity, and the residual vector sequence is used as a gradient source of an optimization objective function of the source term function; After each round of gradient update is completed, the partial differential equation is numerically solved again to update a second evolution sequence of the theoretical concentration field, and the target residual vector is calculated again until the residual mean is lower than a threshold value, and the target residual vector is output; The target source term function corresponding to the target residual vector is output, and a non-zero distribution interval of the target source term function is determined as the leakage source position; The leakage source position is mapped to a gas structure topology graph, so as to determine the physical position of the leakage source.

[0010] On the basis of the above technical solutions, preferably, the multi-source data collected by the sensor system arranged in the target area is obtained, and specifically further includes: A plurality of gas sensors are deployed according to the gas structure topology graph, and nodes of the gas structure topology graph correspond to gas source positions, gas delivery paths, gas connection part positions, and gas use terminal positions; A first time at which gas initially reaches a first sensor and second times at which gas initially reaches each second sensor are determined based on a set concentration mutation threshold value, wherein the first sensor is any one of the plurality of gas sensors, and the second sensor is any one of the plurality of gas sensors except the first sensor; A plurality of environmental sensors are respectively arranged at each gas sensor position, and the environmental sensors include temperature sensors, humidity sensors, air pressure sensors, and air flow speed sensors, and are respectively used to collect temperature data, humidity data, air pressure data, and gas flow speed data.

[0011] On the basis of the above technical solutions, preferably, the gas concentration data and the physical environment data are fused, and the leakage process of the leaked gas is calculated by inversion, and specifically includes: A gas propagation speed function is defined based on the temperature data, the humidity data, the air pressure data, and the gas flow speed data; An arrival time constraint equation is constructed based on the gas propagation speed function, and the arrival time constraint equation is used to represent a path propagation time of gas from the leakage source position to each second sensor, which is equal to a time difference between the second time and the first time; A system of simultaneous constraint equations is established for the plurality of gas sensors through the arrival time constraint equation, and a residual function of each gas sensor is constructed; The residual functions of all the gas sensors are constituted into a vector form of a nonlinear least squares optimization objective function, wherein the nonlinear least squares optimization objective function takes the first time and the estimated leakage source position as parameters to be solved, and minimizes the sum of squares of the difference between the node path propagation time and the actual arrival time.

[0012] On the basis of the above technical solutions, preferably, based on the leakage process, the leakage source position is determined and mapped back to the physical structure of the gas pipe network, specifically comprising: The nonlinear least squares optimization objective function is iteratively solved, in each iteration process, the residual value of the residual function is recalculated according to the current estimated value of the parameter to be solved, and the derivative matrix of the residual function to the parameter to be solved is calculated, and the step size is adjusted by constructing the Jacobian matrix and the damping term; In the process of multiple iterations, if the change of the parameter to be solved is less than the set threshold, the estimated leakage source position corresponding to the last parameter to be solved is output, and the leakage source position is obtained; The leakage source position is mapped to the gas structure topology graph, so as to determine the physical position of the leakage source.

[0013] In the second aspect of the present application, a multi-source gas data fusion leakage source positioning device is provided, which is used to execute any one of the above multi-source gas data fusion leakage source positioning methods, and the device comprises an acquisition module, a processing module and an output module, wherein: The acquisition module is used to acquire multi-source data collected by a sensor system arranged in a target area, wherein the sensor system comprises a plurality of gas sensors and a plurality of environmental sensors, the gas sensors are used to collect gas concentration data, and the environmental sensors are used to collect physical environment data; The processing module is used to fuse the gas concentration data and the physical environment data, and perform inverse calculation on the leakage process of the leakage gas; The output module is used to determine the leakage source position based on the leakage process and map it back to the physical structure of the gas pipe network.

[0014] On the basis of the above technical solutions, preferably, the acquisition module is used to deploy a plurality of gas sensors according to a gas structure topology graph, and the nodes of the gas structure topology graph correspond to gas source positions, gas transmission paths, gas connection part positions and gas use terminal positions; The acquisition module is used to deploy a plurality of environmental sensors, and the environmental sensors comprise temperature sensors, humidity sensors, air pressure sensors and air flow speed sensors, which are respectively used to collect temperature data, humidity data, air pressure data and gas flow speed data; The acquisition module is configured to establish a structured index of timestamps, spatial positions, sensor types and sampling data according to data collected by the plurality of gas sensors and the plurality of environmental sensors, and form a raw data set with complete space-time attributes.

[0015] On the basis of the above technical solutions, preferably, the processing module is configured to calculate gas concentration change data according to gas concentration data in the raw data set. The processing module is configured to construct a gas diffusion physical model with the gas concentration change data as a core observation, wherein the gas diffusion physical model takes a three-dimensional space grid as a basic unit, divides the target region into a regular voxel grid structure, takes temperature data, humidity data, gas pressure data and gas flow rate data as state variables of each grid node in the grid, and describes the gas diffusion behavior by establishing a partial differential equation. The processing module is configured to determine a local gas molecular mean free path and a thermal motion coefficient by the temperature data, the humidity data and the gas pressure data during construction of the gas diffusion physical model, thereby calculating diffusion coefficients of each grid node, solving a local airflow field according to airflow velocity data to obtain a flow velocity vector in each space unit, inputting the diffusion coefficients and the flow velocity vector into the partial differential equation, and obtaining a first evolution sequence of a theoretical concentration field by numerical solving.

[0016] On the basis of the above technical solutions, preferably, the processing module is configured to calculate a difference between the gas concentration change data and the first evolution sequence, and construct a residual vector sequence. The processing module is configured to perform multiple rounds of gradient updates on parameters of a source term function of the partial differential equation by using a gradient descent method based on a back propagation mechanism, wherein the source term function is used to describe time-varying estimated values of the leakage source position and intensity, and the residual vector sequence is used as a gradient source of an optimization objective function of the source term function. The processing module is configured to re-solve the partial differential equation after each round of gradient update is completed, update a second evolution sequence of the theoretical concentration field, and calculate a target residual vector again until a residual average is lower than a threshold value, and output the target residual vector. The output module is configured to output a target source term function corresponding to the target residual vector, and determine a non-zero distribution interval of the target source term function as the leakage source position. The output module is configured to map the leakage source position to a gas structure topology graph, thereby determining a physical position of the leakage source.

[0017] On the basis of the above technical solutions, preferably, the acquisition module is configured to deploy a plurality of gas sensors according to the gas structure topology graph, and nodes of the gas structure topology graph correspond to gas source positions, gas delivery paths, gas connection position and gas use terminal positions. The acquisition module is configured to determine a first time at which the gas initially reaches the first sensor and a second time at which the gas initially reaches each second sensor based on the set concentration mutation threshold, wherein the first sensor is any one of the plurality of gas sensors, and the second sensor is any one of the plurality of gas sensors except the first sensor. The acquisition module is configured to deploy a plurality of environmental sensors at each gas sensor position respectively, wherein the environmental sensors include a temperature sensor, a humidity sensor, a barometric pressure sensor, and an air flow speed sensor, and are respectively configured to collect temperature data, humidity data, barometric pressure data, and gas flow speed data.

[0018] In the above technical solution, the processing module is configured to define a gas propagation speed function based on the temperature data, the humidity data, the barometric pressure data, and the gas flow speed data. The processing module is configured to construct a time-of-arrival constraint equation based on the gas propagation speed function, wherein the time-of-arrival constraint equation is configured to represent a path propagation time of the gas from the leak source position to each second sensor, and the path propagation time is equal to a time difference between the second time and the first time. The processing module is configured to establish a system of simultaneous constraint equations for the plurality of gas sensors by the time-of-arrival constraint equation, and construct a residual function of each gas sensor. The processing module is configured to construct a vector-form nonlinear least squares optimization objective function by the residual functions of all gas sensors, wherein the nonlinear least squares optimization objective function takes the estimated leak source position and the first time as to-be-solved parameters, and minimizes a sum of squares of differences between each node path propagation time and an actual arrival time.

[0019] In the above technical solution, the processing module is configured to iteratively solve the nonlinear least squares optimization objective function, and in each iteration, the residual value of the residual function is recalculated according to a current to-be-solved parameter estimate value, and a derivative matrix of the residual function with respect to the to-be-solved parameter is calculated, and a step size is adjusted by constructing a Jacobian matrix and a damping term. The processing module is configured to, in the plurality of iterations, if it is determined that a change in the to-be-solved parameter is less than a set threshold, output an estimated leak source position corresponding to a last to-be-solved parameter, and obtain the leak source position. The output module is configured to map the leak source position to a gas structure topology, and thus determine a physical position of the leak source.

[0020] In a third aspect of the application, an electronic device is provided, comprising a processor, a memory, a user interface and a network interface, the memory being configured to store instructions, the user interface and the network interface each being configured to communicate with other devices, and the processor being configured to execute the instructions stored in the memory to cause the electronic device to perform the method according to any one of the preceding aspects.

[0021] In a fourth aspect of the application, a computer-readable storage medium is provided, which stores instructions that, when executed, perform the method according to any one of the preceding aspects.

[0022] In summary, the one or more technical solutions provided in the embodiments of the application have at least the following technical effects or advantages: 1. The application provides a multi-source gas data fusion leakage source positioning method and device, relating to the technical field of sensor systems, comprising: obtaining multi-source data collected by a sensor system arranged in a target area, wherein the sensor system comprises a plurality of gas sensors and a plurality of environmental sensors, the gas sensors are used to collect gas concentration data, and the environmental sensors are used to collect physical environment data; fusing the gas concentration data and the physical environment data, performing inversion calculation on the leakage process of the leaked gas; determining the position of the leakage source based on the leakage process, and mapping back to the physical structure of the gas pipe network. The application can position the position of the gas leakage, thereby improving the disposal efficiency of the gas leakage detection.

[0023] 2. The application constructs a multi-source fusion modeling system with gas concentration change as the observation core and physical environment state as the propagation constraint, combines partial differential diffusion equation to invert the leakage path, and based on the gradient optimization mechanism, accurately solves the source function, finally realizes the quantitative identification of the spatial position of the leakage source and the mapping of the pipe network structure, not only significantly improves the spatial perception ability and positioning accuracy of the gas leakage detection, but also greatly enhances the robust analysis ability of the system to complex diffusion behavior under environmental disturbance, thereby realizing the efficient closed loop from leakage perception to structure positioning, improving the timeliness, safety and intelligent level of leakage disposal.

[0024] 3. The application introduces a time difference observation mechanism based on sudden concentration response, constructs a gas propagation speed function combined with environmental variables such as temperature, humidity, air pressure and gas flow rate, forms a fine space-time propagation constraint model, and based on this, establishes a nonlinear least squares optimization framework to iteratively solve the spatial position and occurrence time of the leakage source, and finally maps the results to the gas structure topology graph to realize the quantitative positioning of the source point from the gas dynamic diffusion process. Significantly improves the accuracy and dynamic adaptability of gas leakage positioning, has the technical advantages of fast response, stable algorithm and strong physical interpretability, and enhances the intelligent level and practical safety performance of the leakage detection system. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a flow chart of a method for locating a leak source by fusing multi-source gas data disclosed in an embodiment of the present invention; Figure 2 This is a module schematic diagram of a leakage source locating device for multi-source gas data fusion disclosed in an embodiment of the present invention; Figure 3 It is a structural diagram of an electronic device disclosed in an embodiment of the present invention.

[0026] Explanation of the reference numerals: 201, acquisition module; 202, processing module; 203, output module; 301, processor; 302, communication bus; 303, user interface; 304, network interface; 305, memory. DETAILED DESCRIPTION

[0027] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments.

[0028] In describing the embodiments of the present invention, words such as "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "for example" or "for instance" in the embodiments of the present invention should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "for example" or "for instance" is intended to present the relevant concepts in a concrete manner.

[0029] In the description of the embodiments of the present invention, the term "multiple" means two or more. For example, multiple systems refers to two or more systems, and multiple screen terminals refers to two or more screen terminals. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0030] The existing household gas leakage detection technology mainly monitors the concentration changes of methane, propane and other combustible gases in the air in real time through fixedly arranging high-sensitivity gas sensors, and triggers alarm and gas cutting linkage in combination with a threshold judgment mechanism, which can realize timely response and remote early warning of leakage events. However, due to the lack of recognition ability of the spatial position of the leakage source, accurate positioning information cannot be provided after the alarm, which leads the user to manually investigate in the environment with safety risks, easily causes misjudgment and secondary harm, and seriously affects the actual disposal efficiency. Especially in the case of multiple complex leakage situations such as sealing failure of the gas input end, aging of the conveying path pipeline, aging of the sealing of the connecting part, and abnormality of the terminal equipment, the technical limitations of the traditional fixed-point monitoring scheme in the leakage tracing ability are more prominent.

[0031] The embodiment discloses a leakage source positioning method based on multi-source gas data fusion, referring to Figure 1 , comprising the following steps S110-S130: S110, acquiring multi-source data collected by a sensor system arranged in a target area.

[0032] S120, fusing gas concentration data and physical environment data, and performing inverse calculation on the leakage process of the leakage gas.

[0033] S130, determining the position of the leakage source based on the leakage process, and mapping back to the physical structure of the gas pipeline network.

[0034] The leakage source positioning method based on multi-source gas data fusion disclosed by the embodiment of the application is applied to a server. The server includes but is not limited to electronic devices such as mobile phones, tablet computers, wearable devices, PC (Personal Computer), and the like, and can also be a background server running a leakage source positioning method based on multi-source gas data fusion. The server can be realized by an independent server or a server cluster composed of multiple servers.

[0035] In the first embodiment, multi-source data collected by a sensor system arranged in a target area is acquired, specifically including: deploying a plurality of gas sensors according to a gas structure topology map, the nodes of the gas structure topology map corresponding to gas source positions, gas conveying paths, gas connection part positions and gas use terminal positions; deploying a plurality of environment sensors, the environment sensors including temperature sensors, humidity sensors, air pressure sensors and air flow speed sensors, respectively used for collecting temperature data, humidity data, air pressure data and gas flow speed data; establishing a structured index of time stamps, spatial positions, sensor types and sampling data according to the data collected by the plurality of gas sensors and the plurality of environment sensors, to form a raw data set with complete space-time attributes.

[0036] Specifically, in the deployment phase, first, the spatial layout of gas sensors is completed according to the gas structure topology map of the interior of the residence. The gas structure topology map is a directed graph structure representing the connection relationship between each component unit in the gas supply system, including the gas source position, the gas transmission path, the gas connection position, and the gas use terminal position. Each node represents a key functional position, and each edge represents a pipeline connection relationship. According to the topology map, gas sensors are selected and deployed at each node position to ensure that the sensors can cover all key areas where gas flows or may leak, forming a logically complete observation network in space. For example, in a typical residential scenario, four gas sensor nodes can be deployed at the house gas main valve, near the kitchen gas stove air inlet, at the kitchen wall surface pipe bending interface, and at the water heater interface to cover the source point, straight path, connection interface, and use terminal, respectively.

[0037] Subsequently, environmental sensors are deployed to synchronously collect temperature data, humidity data, air pressure data, and gas flow rate data. Temperature sensors measure air temperature based on thermocouples or thermistors, humidity sensors reflect air relative humidity based on capacitive or resistive sensing units, air pressure sensors measure atmospheric pressure changes using piezoresistive MEMS structures, and airflow speed sensors obtain local air flow speed based on hot-wire anemometers or ultrasonic anemometers. Various environmental sensors and gas sensors are spatially co-pointed or adjacent to each other, ensuring that their observation areas have direct intersections, enabling spatial matching and response coupling in subsequent data fusion modeling.

[0038] In the data collection phase, all sensors are uniformly configured with millisecond-level precision local timestamp recording mechanisms, and are spatially encoded based on physical installation location numbers and indoor plan coordinate systems. Each sensor's collected data includes sensor type identification, timestamp, spatial coordinates, and current sample value, forming a structured four-tuple. The above four-tuple is standardized and organized to build a data structure indexed primarily by time and secondarily by spatial location and sensor type, ensuring complete spatiotemporal consistency of the data. For example, the gas sensor node at the water heater interface collects a concentration value of 180 ppm at 2025-07-25 10:00:00, and the record in the original data set is constructed as: {Type: Gas Sensor, Location: Node H1, Time: 2025-07-25 10:00:00, Concentration Value: 180 ppm}. This forms multi-channel observation data at the same location with the temperature sensor record at the same time, providing a unified data input basis for subsequent leakage modeling, path inversion, and source positioning. The entire process builds a unified description framework for multi-source heterogeneous data, paving the way for gas leakage spatial modeling and intelligent identification.

[0039] In one possible implementation, the gas concentration data and the physical environment data are fused, and the leakage path of the leakage gas is inversely calculated, specifically including: calculating gas concentration change data according to the gas concentration data in the original data set; taking the gas concentration change data as the core observation, a gas diffusion physical model is constructed, wherein the gas diffusion physical model takes a three-dimensional space grid as a basic unit, divides the target region into a regular voxel grid structure, each grid node in the grid takes temperature data, humidity data, air pressure data and gas flow rate data as state variables, and a partial differential equation is established to describe the gas diffusion behavior; in the process of constructing the gas diffusion physical model, the local gas molecular mean free path and the thermal motion coefficient are determined through the temperature data, humidity data and air pressure data, so as to calculate the diffusion coefficient of each grid node, the local airflow field is solved according to the airflow velocity data, the flow velocity vector in each space unit is obtained, the diffusion coefficient and the flow velocity vector are input into the partial differential equation, and the first evolution sequence of the theoretical concentration field is obtained by numerical solution.

[0040] Specifically, first, after forming the original data set, gas concentration change data needs to be extracted from the gas concentration data collected by the gas sensor. The gas concentration change data refers to the change rate of the gas concentration value at a certain space node over time, reflecting the diffusion dynamic process of the leakage event. Taking each gas sensor node as a reference, the gas concentration increment in the adjacent sampling time interval is calculated and normalized to form a differential concentration sequence. Let the gas sensor node The concentration value at time is , and the gas concentration change amount is:

[0041] The above differential sequence is used to describe the mutation degree of local gas concentration, which is an important observation quantity for judging the gas propagation front and diffusion path.

[0042] Next, a gas diffusion physical model is constructed with the gas concentration change data as the core driving factor. The entire target region is divided into a regular voxel grid structure, and each cubic grid is taken as a three-dimensional space discrete unit. Each voxel grid node is defined to have complete physical state variables, including temperature data , humidity data , air pressure data and airflow velocity data . A partial differential control equation is introduced in the space grid field to describe the propagation behavior of the gas in space and time. A convection-diffusion type control equation is used to describe the dynamic evolution relationship of the gas concentration :

[0043] wherein the diffusion coefficient Characterize the molecular thermal diffusion capacity, flow velocity vector Characterize the macroscopic gas flow driven convection diffusion trend. Indicates the source term function, which is used to describe the time-varying estimated value of the leakage source position and intensity. The control equation is a kind of second-order partial differential equation, the first term is the diffusion term, and the second term is the convection term, which describes the diffusion behavior of gas under the condition of non-uniform thermal field and gas flow disturbance.

[0044] In order to obtain the diffusion coefficient, it is necessary to estimate the average free path and diffusion constant of gas molecules in combination with local temperature, humidity and air pressure parameters. Diffusion coefficient It can be given by the following empirical formula:

[0045] Among them, is the Boltzmann constant, is the dynamic viscosity coefficient of the gas, is the effective radius of the gas molecule, is the humidity adjustment coefficient, is the air pressure influence function. The higher the temperature, the lower the air pressure, the smaller the humidity, the larger the diffusion coefficient, and the stronger the gas diffusion capacity.

[0046] Flow velocity vector The interpolation field is constructed by sampling data from the air flow velocity sensor and the boundary conditions of the building space, and the three-dimensional air flow velocity component is estimated at each grid node by using linear interpolation or finite volume method . If the wind speed value of each sensor node is , the corresponding position is , then the velocity value at the center position of each voxel is constructed by using multi-point weighted interpolation:

[0047] Input the diffusion coefficient and flow velocity vector field at all space points into the control equation, and use the implicit Crank-Nicolson difference format for numerical solution, the time step is set to , the space step is , and the solution target is the theoretical concentration value of all grid nodes at the next time, forming the concentration field evolution sequence .

[0048] In the embodiments of the present application, the source term function is not only used to describe whether the leakage source exists, but also used to quantify the leakage rate and diffusion influence range. Its mathematical meaning is that in the theoretical concentration evolution process, the model actively injects gas at the specified space position, so as to simulate the continuous release process of gas in the real leakage scene. Its physical meaning can be understood as the "source function" of the leakage point.

[0049] The source term function generally satisfies the following conditions: when located in the spatial region and time interval of the leak source at the stage of leak occurrence, takes non-zero value; otherwise, it is zero or near zero. For stable leak events, the source term function can be modeled as the product of spatial Gaussian distribution and time constant function:

[0050] wherein denotes the leak intensity, denotes the estimated location of the leak source, denotes the diffusion scale of the leak region, is a time indicator function, controlling the leak duration interval.

[0051] In the leak path inversion and leak source localization, the source term function is also a key parameter to be optimized. By comparing the residual error between the model simulated concentration field and the sensor measured concentration field, and performing reverse optimization on the spatial position parameter and intensity parameter in the source term function, the form of is adjusted so that the theoretical concentration field approximates the measured concentration field, thereby realizing the joint inference of the spatial location and time characteristics of the leak source. In other words, the source term function is not only an input quantity of the diffusion equation, but also a core inversion object in the leak source tracing process.

[0052] The theoretical concentration evolution sequence, as a physical fitting result of the real gas diffusion process, provides a comparable analysis benchmark for subsequent leak source location inversion, path reconstruction and time tracing. Through the above modeling process, the gas diffusion trend can be accurately described under the conditions of thermal disturbance and air flow disturbance with actual physical background, and the numerical controllable simulation of the leak process can be realized.

[0053] In one possible implementation, based on the leak process, the leak source location is determined and mapped back to the physical structure of the gas pipe network, specifically including: calculating the difference between the gas concentration change data and the first evolution sequence to construct a residual vector sequence; using a gradient descent method based on a back propagation mechanism, performing multiple rounds of gradient updates on the parameters of the source term function of the partial differential equation, wherein the source term function is used to describe the time-varying estimated value of the leak source location and intensity, and the residual vector sequence is used as the gradient source of the optimization objective function of the source term function; after each round of gradient update is completed, the partial differential equation is re-solved numerically to update the second evolution sequence of the theoretical concentration field, and the target residual vector is calculated again until the residual mean is lower than the threshold value, and the target residual vector is output; the target source term function corresponding to the target residual vector is output, and the non-zero distribution interval of the target source term function is determined as the leak source location; the leak source location is mapped to the gas structure topology graph, thereby determining the physical location of the leak source.

[0054] Specifically, first, the first evolution sequence has been obtained in the theoretical simulation stage, that is, the simulated concentration field obtained by inputting the current source function parameter value into the gas diffusion partial differential equation. At the same time, the gas concentration change data sequence is constructed based on the original concentration data collected by the gas sensor. , which is the actual observation value at discrete spatial locations and time points. The two sequences are compared point by point at the same grid nodes and time nodes to construct the residual vector sequence:

[0055] in It represents the error between the simulated concentration and the measured concentration at each sampling point. The residual vector is used as the source of gradient calculation in the optimization process to measure whether the current source function parameters are reasonable.

[0056] Next, for the source function in the gas diffusion partial differential equation ,in Construct a loss function for a set of parameters including leakage location, leakage intensity and diffusion scale Represents the total residual sum of squares:

[0057] The optimization of the source function parameters is performed using the gradient descent method with backpropagation mechanism. The partial derivatives of the loss function with respect to the parameters are calculated by automatic differentiation:

[0058] Then set the learning rate Perform iterative updates:

[0059] After each parameter update, the new source function Input into the diffusion partial differential equation and solve it numerically again to obtain the updated theoretical concentration field , and recalculate the residual vector. This iterative process continues until the mean residual of the objective function meets the set convergence condition, that is:

[0060] in is the total number of sampling points, is the residual convergence threshold.

[0061] After convergence is completed, the corresponding source function is output , the continuum formed by the non-zero area in the spatial position of the source function is defined as the leakage source spatial distribution area. In other words, if there is a spatial point satisfy and the non-zero value persists over multiple time steps, the point is included in the set of leak source estimates.

[0062] Finally, the estimated spatial coordinates of the leak source are mapped to the gas structure topology graph, and the node identifier closest to its location is found through a spatial matching mechanism, completing the mapping of the leak source's structure number in the actual gas pipe network physical structure, for example, matching the leak source positioning result to the node "connection interface J7", which is the identification result of the physical location of the leak source.

[0063] In the second embodiment, the multi-source data collected by the sensor system deployed in the target area is obtained, and specifically further includes: deploying a plurality of gas sensors according to the gas structure topology graph, the nodes of the gas structure topology graph corresponding to the gas source position, the gas transmission path, the gas connection position and the gas use terminal position; determining the first time when the gas first reaches the first sensor and the second time when the gas first reaches each second sensor based on a set concentration mutation threshold, wherein the first sensor is any one of the plurality of gas sensors, and the second sensor is any one of the plurality of gas sensors except the first sensor; respectively deploying a plurality of environmental sensors at each gas sensor position, the environmental sensors including temperature sensors, humidity sensors, air pressure sensors and air flow speed sensors, respectively for collecting temperature data, humidity data, air pressure data and gas flow rate data.

[0064] Specifically, in actual implementation, first, a gas structure topology graph needs to be constructed according to the physical structure information of the indoor gas supply system. The topology graph is a directed graph structure, and its nodes correspond to key positions in the physical pipe network, including the gas source position (such as the house total valve or gas meter), the main trunk section in the gas transmission path, the gas connection position (such as the tee, elbow, interface connection), and the gas use terminal position (such as the stove, water heater, wall-mounted boiler, etc.). Each node has a clear spatial coordinate and forms a topology edge through the pipeline connection. According to the topology graph, a plurality of gas sensors are arranged at the above-mentioned key nodes, so that the gas sensors cover the starting point, transmission process, key connection point and end consumption point of gas flow in space, ensuring that once a leak occurs at any link, the corresponding node can respond immediately. For example, in the kitchen, four gas sensors are installed below the gas meter, in the middle section of the exposed pipe, at the interface valve, and at the gas inlet of the stove.

[0065] After the gas leakage event occurs, the gas sensor will continue to monitor the gas concentration value in the air at a preset sampling frequency. In order to capture the spatial propagation process of the front of the leaked gas, a preset concentration mutation threshold is set in the system, for example, set to 100 ppm. When any gas sensor detects that its concentration value rises from the stable background concentration and first exceeds the mutation threshold, the time is recorded as the initial arrival time of the gas at the gas sensor. The time is the time reference point of the actual leakage path, and the first sensor triggered is defined as the first sensor, and the trigger time is the first time T1. The gas sensor that is not triggered first is the second sensor, and the initial arrival time of the gas is recorded as the second time T2. i . The set of time data is used as the key input quantity for subsequent diffusion path time difference inversion. For example, if the sensor under the gas meter first reaches the mutation threshold, the initial mutation time is recorded as 10:00:12, and the other positions are 10:00:15, 10:00:20, and 10:00:26, respectively. The sequence forms a time sequence chain of gas propagation.

[0066] To ensure that the modeling of the physical environment of gas diffusion at each gas sensor node has spatiotemporal consistency, a set of environmental sensors is deployed at the same point as each gas sensor node. The temperature sensor is used to collect local thermal environment data in the gas diffusion process in real time, the humidity sensor is used to reflect the influence of water vapor content on the viscosity of gas molecule movement, the air pressure sensor is used to monitor the modulation effect of local atmospheric pressure on gas density and diffusion path, and the air flow velocity sensor is used to quantify the influence of wind speed and convection on gas propagation at this position. This set of sensors cooperatively constitutes a physical adjustment parameter field of gas diffusion behavior. For example, a set of MEMS temperature sensors (sampling range -20~85℃), capacitive humidity sensors (accuracy ±2%RH), piezoresistive air pressure sensors (0.1kPa accuracy), and hot-wire air flow velocity sensors (0.05m / s resolution) are installed at the same time as the gas sensor under the cooking appliance. The node will form five-channel spatiotemporal alignment of original data records at each moment, realize the fusion monitoring of complete gas response and dynamic physical environment, and provide complete input for leakage positioning and path inversion.

[0067] In one possible implementation, gas concentration data and physical environment data are integrated, and the leakage process of the leaked gas is inversely calculated, specifically including: defining a gas propagation velocity function based on temperature data, humidity data, air pressure data, and gas flow rate data; constructing an arrival time constraint equation based on the gas propagation velocity function, where the arrival time constraint equation is used to represent the path propagation time of the gas from the leakage source location to each second sensor, which is equal to the time difference between the second time and the first time; establishing a set of simultaneous constraint equations for multiple gas sensors through the arrival time constraint equation, and constructing the residual function of each gas sensor; constructing a nonlinear least squares optimization objective function in vector form through the residual functions of all gas sensors, wherein the nonlinear least squares optimization objective function takes the estimated leakage source location and the first time as parameters to be determined, and minimizes the sum of the squares of the difference between the path propagation time of each node and the actual arrival time.

[0068] Specifically, in the embodiment of the present application, it is first necessary to combine the temperature data, humidity data, air pressure data and gas flow rate data collected by the environmental sensor to establish a gas propagation velocity function at the location of the gas sensor. , used to estimate the average path velocity of gas propagating from the leak source to each gas sensor. This propagation velocity function is the result of empirical modeling of the composite physical field, established based on the diffusion theory and convection coupling mechanism. For example, it can be constructed as follows:

[0069] in It represents the macroscopic airflow convection velocity measured by the airflow velocity sensor. is the fitting coefficient, is the Boltzmann constant, is the dynamic viscosity coefficient of the gas, 、 are the regulating factors of humidity and air pressure, respectively, thereby quantifying the coupling effect of the local environment on molecular thermal diffusion and macroscopic gas migration velocity.

[0070] Next, the arrival time constraint equation of the gas propagation path is constructed. Assume that the estimated location of the leakage source is , No. The gas sensor locations are , the gas propagation distance is the Euclidean distance:

[0071] Then the path propagation time is:

[0072] The propagation time is constrained with the actual time difference observed, and the time difference is defined as ,in For the a second time of the gas sensor, is a first time. The constraint equation is obtained as:

[0073] The equation is established on all second sensors, forming a multi-nonlinear simultaneous constraint system.

[0074] Further, each constraint equation is constructed in the form of a residual function, representing the deviation of the difference between the theoretical propagation time and the observed time, and the first residual function is defined as:

[0075] On this basis, all residual functions are summarized to form a nonlinear least squares optimization objective function:

[0076] The objective function takes the three-dimensional coordinates of the leak source position and the first time as the optimization variables, and uses a high-precision nonlinear optimization algorithm (such as the Levenberg-Marquardt method) to iteratively solve until the squared deviation between all path propagation times and observed time differences is minimized.

[0077] In one possible implementation, based on the leakage process, the leak source position is determined and mapped back to the physical structure of the gas pipe network, specifically including: iteratively solving the nonlinear least squares optimization objective function, in each iteration, recalculating the residual value of the residual function according to the current parameter estimate, and calculating the derivative matrix of the residual function with respect to the parameter to be solved, and adjusting the update step by constructing the Jacobian matrix and the damping term; in the process of multiple iterations, if it is determined that the change of the parameter to be solved is less than a set threshold, the estimated leak source position corresponding to the last parameter to be solved is output, and the leak source position is obtained; mapping the leak source position to the gas structure topology to determine the physical location of the leak source.

[0078] Specifically, in the parameter optimization process of the leak source inversion, the nonlinear least squares optimization method is used to iteratively solve the objective function, and the Levenberg-Marquardt method is specifically selected to balance the convergence speed and numerical stability. In each iteration, the current parameter estimate, i.e., the three-dimensional position parameters of the leak source and the leak start time parameters are used to recalculate all residual function values, which are in the form of:

[0079] Meanwhile, the partial derivative matrix of the residual function with respect to all unknown parameters is constructed to obtain the Jacobian matrix , each row of which represents the partial derivative of the residual function of the corresponding sensor with respect to , for linear approximation of the gradient distribution of the current solution space in the local region. The parameter update direction of the current iteration step satisfies:

[0080] , where is a column vector composed of the current values of all residual functions, is a damping coefficient that controls the adjustment of the iteration step size, is an identity matrix used to prevent the divergence of the solution caused by the ill-conditioned Jacobian matrix when the parameter space is close to a singular point. If the residual sum of squares decreases after the current iteration, then is reduced to relax the constraint; if the residual increases, then is increased to increase the regularization penalty to converge to a local minimum.

[0081] As the iteration process advances, the system continuously updates the parameter estimate and continuously calculates the residual vector and the Jacobian matrix of the next round. When the change in the parameters between two consecutive iterations is less than the set convergence threshold (e.g., set to meters and seconds), it is considered that the iteration has converged, the optimization process is terminated, and the last round of parameters is output as the final estimation result, obtaining the location of the leakage source:

[0082] After completing the spatial location estimation of the leakage source, the physical coordinates need to be mapped to the gas structure topology graph. The gas structure topology graph has already labeled and numbered all functional nodes, including gas source nodes, transmission path nodes, connection node, and use terminal nodes. The mapping process is based on the principle of spatial proximity and uses the minimum distance matching method to match the estimated location with the node set in the topology graph to calculate the Euclidean distance:

[0083] The node number that makes the minimum is selected as the physical structure location corresponding to the leakage source, completing the mapping from the mathematical solution to the gas pipe network structure identification. For example, if the distance to the node "connection interface J12" is the closest, then the leakage source is marked as the structure location corresponding to the J12 node, which is used for subsequent repair intervention operations or automatic response mechanism.

[0084] The embodiment also discloses a leakage source positioning device based on multi-source gas data fusion, which refers to Figure 2 comprises an acquisition module 201, a processing module 202 and an output module 203, and the device is used for executing any one of the above multi-source gas data fusion leakage source positioning methods, wherein: The acquisition module 201 is used for acquiring multi-source data collected by a sensor system arranged in a target area, wherein the sensor system comprises a plurality of gas sensors and a plurality of environmental sensors, the gas sensors are used for collecting gas concentration data, and the environmental sensors are used for collecting physical environment data.

[0085] The processing module 202 is used for fusing the gas concentration data and the physical environment data, and performing inversion calculation on a leakage process of a leakage gas.

[0086] The output module 203 is used for determining a leakage source position based on the leakage process and mapping back a physical structure of a gas pipe network.

[0087] In a possible implementation, the acquisition module 201 is used for deploying a plurality of gas sensors according to a gas structure topology map, and nodes of the gas structure topology map correspond to gas source positions, gas transmission paths, gas connection position and gas use terminal positions.

[0088] The acquisition module 201 is used for deploying a plurality of environmental sensors, and the environmental sensors comprise temperature sensors, humidity sensors, air pressure sensors and air flow speed sensors, and are respectively used for collecting temperature data, humidity data, air pressure data and gas flow speed data.

[0089] The acquisition module 201 is used for establishing a structured index of a timestamp, a spatial position, a sensor type and sampling data according to data collected by the plurality of gas sensors and the plurality of environmental sensors, and forming an original data set with complete space-time attributes.

[0090] In a possible implementation, the processing module 202 is used for calculating gas concentration change data according to the gas concentration data in the original data set.

[0091] The processing module 202 is used for taking the gas concentration change data as a core observation quantity, constructing a gas diffusion physical model, wherein the gas diffusion physical model takes a three-dimensional space grid as a basic unit, divides the target area into a regular voxel grid structure, takes temperature data, humidity data, air pressure data and gas flow speed data as state variables of each grid node in the grid, and describes a gas diffusion behavior by establishing a partial differential equation.

[0092] The processing module 202 is configured to determine a local gas molecular mean free path and a thermal motion coefficient by using the temperature data, the humidity data and the air pressure data in the process of constructing the gas diffusion physical model, to calculate diffusion coefficients of each grid node, to solve a local airflow field according to airflow velocity data, to obtain a flow velocity vector in each space unit, and to input the diffusion coefficients and the flow velocity vector into a partial differential equation to obtain a first evolution sequence of a theoretical concentration field by numerical solving.

[0093] In a possible implementation, the processing module 202 is configured to calculate a difference between the gas concentration change data and the first evolution sequence, and to construct a residual vector sequence.

[0094] The processing module 202 is configured to perform a plurality of rounds of gradient updates on parameters of a source function of the partial differential equation by using a gradient descent method based on a back propagation mechanism, wherein the source function is used to describe time-varying estimated values of the leakage source position and intensity, and the residual vector sequence is used as a gradient source of an optimization objective function of the source function.

[0095] The processing module 202 is configured to re-solve the partial differential equation after each round of gradient update is completed, to update a second evolution sequence of the theoretical concentration field, and to calculate a target residual vector again until a residual average is lower than a threshold value, and to output the target residual vector.

[0096] The output module 203 is configured to output a target source function corresponding to the target residual vector, and to determine a non-zero distribution interval of the target source function as the leakage source position.

[0097] The output module 203 is configured to map the leakage source position to a gas structure topology map, to determine a physical position of the leakage source.

[0098] In a possible implementation, the root acquisition module 201 is configured to deploy a plurality of gas sensors according to the gas structure topology map, and nodes of the gas structure topology map correspond to gas source positions, gas delivery paths, gas connection position and gas use terminal positions.

[0099] The acquisition module 201 is configured to determine a first time at which gas initially reaches a first sensor and a second time at which gas initially reaches each second sensor based on a set concentration mutation threshold value, wherein the first sensor is any one of the plurality of gas sensors, and the second sensor is any one of the plurality of gas sensors except the first sensor.

[0100] The acquisition module 201 is configured to deploy a plurality of environment sensors at each gas sensor position respectively, and the environment sensors include temperature sensors, humidity sensors, air pressure sensors and airflow velocity sensors, which are respectively used to collect temperature data, humidity data, air pressure data and gas flow rate data.

[0101] In a possible implementation, the processing module 202 is configured to define a gas propagation speed function based on the temperature data, the humidity data, the air pressure data, and the gas flow rate data.

[0102] The processing module 202 is configured to construct a time-of-arrival constraint equation based on the gas propagation speed function, where the time-of-arrival constraint equation is used to represent a path propagation time of the gas from the leak source position to each second sensor, which is equal to a time difference between the second time and the first time.

[0103] The processing module 202 is configured to establish a simultaneous constraint equation set for the plurality of gas sensors by the time-of-arrival constraint equation, and construct a residual function of each gas sensor.

[0104] The processing module 202 is configured to construct a vector-form nonlinear least squares optimization objective function by the residual functions of all the gas sensors, where the nonlinear least squares optimization objective function takes the estimated leak source position and the first time as unknown parameters, and minimizes a sum of squares of differences between the node path propagation times and actual arrival times.

[0105] In a possible implementation, the processing module 202 is configured to iteratively solve the nonlinear least squares optimization objective function, and in each iteration, recompute residual values of the residual functions according to current estimated values of the unknown parameters, and compute a derivative matrix of the residual functions with respect to the unknown parameters, and adjust an update step size by constructing a Jacobian matrix and a damping term.

[0106] The processing module 202 is configured to output, in the plurality of iterations, the estimated leak source position corresponding to the last iteration of the unknown parameters, to obtain the leak source position, if it is determined that a change of the unknown parameters is less than a preset threshold.

[0107] The output module 203 is configured to map the leak source position to a gas structure topology, to determine a physical position of the leak source.

[0108] It should be noted that the apparatus provided in the above embodiments is only used to illustrate the division of the above functional modules, and in actual applications, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the apparatus is divided into different functional modules to complete all or part of the above described functions. In addition, the apparatus and method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process is described in the method embodiments, which will not be repeated here.

[0109] The embodiment further discloses an electronic device, which refers to Figure 3 The electronic device can include at least one processor 301, at least one communication bus 302, a user interface 303, a network interface 304, and at least one memory 305.

[0110] The communication bus 302 is configured to realize the connection communication between the components.

[0111] The user interface 303 can include a display screen and a camera. Optionally, the user interface 303 can further include a standard wired interface and a wireless interface.

[0112] The network interface 304 can optionally include a standard wired interface and a wireless interface (e.g., a WI-FI interface).

[0113] The processor 301 can include one or more processing cores. The processor 301 is connected to various parts of the server through various interfaces and lines, and performs various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 305, and calling data stored in the memory 305. Optionally, the processor 301 can be implemented in at least one of a digital signal processing (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA). The processor 301 can be integrated with a combination of one or more of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU is mainly used to process an operating system, a user interface, and an application program. The GPU is used to render and draw the content to be displayed on the display screen. The modem is used to process wireless communication. It can be understood that the above-mentioned modem can also not be integrated into the processor 301, but can be implemented by a separate chip.

[0114] The memory 305 can include a random access memory (RAM) and can also include a read-only memory (ROM). Optionally, the memory includes a non-transitory computer-readable storage medium. The memory 305 can be used to store instructions, programs, codes, code sets, or instruction sets. The memory 305 can include a program storage area and a data storage area, where the program storage area can store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing the various method embodiments described above, and the like. The data storage area can store data related to the various method embodiments described above, and the like. The memory 305 can also optionally be at least one storage device located away from the aforementioned processor 301. The memory 305, as a computer storage medium, can include an operating system, a network communication module, a user interface 303 module, and an application program of a multi-source gas data fusion leak source positioning method.

[0115] In Figure 3 In the electronic device shown, the user interface 303 is mainly used to provide an interface for user input and obtain data input by the user. The processor 301 can be used to call the application program of a multi-source gas data fusion leak source positioning method stored in the memory 305, and when executed by one or more processors 301, the electronic device performs the method of one or more of the above embodiments.

[0116] It should be noted that, for the foregoing method embodiments, in order to simply describe, they are all expressed as a combination of a series of actions, but those skilled in the art should know that the present application is not limited by the order of the actions described, because according to the present application, certain steps can be performed in other orders or at the same time. Secondly, those skilled in the art should know that the embodiments described in the specification all belong to preferred embodiments, and the actions and modules involved are not necessarily essential to the present application.

[0117] In the above embodiments, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0118] In several embodiments provided by the present application, it should be understood that the disclosed apparatus can be implemented in other manners. For example, the division of the apparatus embodiments is merely illustrative, and the division of units can be changed according to actual needs. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections can be indirect couplings or communication connections through some interfaces, devices or units, and can be in electrical, mechanical or other forms.

[0119] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one place, or they may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0120] In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can be physically present separately, or two or more units can be integrated into one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0121] If the integrated unit is realized in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the part of the prior art that contributes to the technical solutions or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium 305 and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the method of each embodiment of the present application. The aforementioned storage medium 305 includes: a U disk, a mobile hard disk, a magnetic disk or an optical disk, and various media that can store program codes.

[0122] The present application also discloses a computer readable storage medium, which stores instructions. When executed by one or more processors 301, the electronic device executes one or more methods as described in the above embodiments.

[0123] The above merely show example embodiments of the present disclosure, and cannot limit the scope of the present disclosure. That is, equivalent changes and modifications made in accordance with the teachings of the present disclosure are still within the scope of the present disclosure. Other embodiments of the present disclosure will be readily apparent to those skilled in the art upon considering the specification and practice of the present disclosure. The present disclosure is intended to cover any variations, uses, or adaptive changes of the present disclosure following the general principles thereof and including those art-known or customary practices not recited in the present disclosure. The specification and examples are to be regarded as merely illustrative, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A leak source location method based on multi-source gas data fusion, characterized in that: The method comprises: Acquiring multi-source data collected by a sensor system deployed in a target area, wherein the sensor system includes a plurality of gas sensors and a plurality of environmental sensors, the gas sensors being used to collect gas concentration data, and the environmental sensors being used to collect physical environment data; The gas concentration data and the physical environment data are integrated to perform an inverse calculation of the leakage process of the leaked gas; Based on the leakage process, the location of the leakage source is determined and mapped back to the physical structure of the gas pipeline network.

2. The method for locating a leak source by fusion of multi-source gas data according to claim 1, characterized in that: The acquiring of multi-source data collected by the sensor system deployed in the target area specifically includes: Deploy a plurality of gas sensors according to a gas structure topology diagram, wherein nodes of the gas structure topology diagram correspond to gas source locations, gas transmission paths, gas connection locations, and gas use terminal locations; Deploy a plurality of the environmental sensors, wherein the environmental sensors include a temperature sensor, a humidity sensor, an air pressure sensor, and an air flow velocity sensor, for collecting temperature data, humidity data, air pressure data, and air flow velocity data, respectively; Based on the data collected by the multiple gas sensors and the multiple environmental sensors, a structured index is established for the timestamp, spatial position, sensor type and sampling data to form an original data set with complete spatiotemporal attributes.

3. The method for locating a leak source by fusion of multi-source gas data according to claim 2, characterized in that: The fusing of the gas concentration data and the physical environment data to perform inverse calculation of the leakage path of the leaked gas specifically includes: Calculating gas concentration change data based on the gas concentration data in the original data set; A gas diffusion physical model is constructed using the gas concentration change data as a core observation. The gas diffusion physical model uses a three-dimensional space grid as a basic unit, divides the target area into a regular voxel grid structure, and uses the temperature data, the humidity data, the air pressure data, and the gas flow rate data as state variables for each grid node within the grid. A partial differential equation is established to describe the gas diffusion behavior. In the process of constructing the gas diffusion physical model, the local gas molecule free path and thermal motion coefficient are determined through the temperature data, the humidity data and the air pressure data, so as to calculate the diffusion coefficient of each grid node, solve the local airflow field according to the airflow velocity data, obtain the flow velocity vector in each spatial unit, input the diffusion coefficient and the flow velocity vector into the partial differential equation, and obtain the first evolution sequence of the theoretical concentration field through numerical solution.

4. The method for locating a leak source by fusion of multi-source gas data according to claim 3, characterized in that: The method of determining the location of the leakage source based on the leakage process and mapping it back to the physical structure of the gas pipeline network specifically includes: Calculating the difference between the gas concentration change data and the first evolution sequence to construct a residual vector sequence; Using a gradient descent method based on a back-propagation mechanism, multiple rounds of gradient updates are performed on the parameters of the source term function of the partial differential equation, wherein the source term function is used to describe the time-varying estimated value of the position and intensity of the leakage source, and the residual vector sequence serves as the gradient source of the objective function optimized by the source term function; After each round of gradient update is completed, the partial differential equation is numerically solved again to update the second evolution sequence of the theoretical concentration field, and the target residual vector is calculated again until the residual mean is lower than the threshold, and the target residual vector is output; Outputting a target source term function corresponding to the target residual vector, and determining a non-zero distribution interval of the target source term function as the leakage source position; The leakage source location is mapped to the gas structure topology map to determine the physical location of the leakage source.

5. The method for locating a leak source by fusion of multi-source gas data according to claim 1, characterized in that: The acquiring of multi-source data collected by the sensor system deployed in the target area specifically includes: Deploy a plurality of gas sensors according to a gas structure topology diagram, wherein nodes of the gas structure topology diagram correspond to gas source locations, gas transmission paths, gas connection locations, and gas use terminal locations; determining a first time when the gas first arrives at a first sensor based on a set concentration mutation threshold, and determining a second time when the gas first arrives at each second sensor, wherein the first sensor is any one of the plurality of gas sensors, and the second sensor is any one of the plurality of gas sensors except the first sensor; A plurality of environmental sensors are deployed at each gas sensor position, and the environmental sensors include a temperature sensor, a humidity sensor, an air pressure sensor and an air flow velocity sensor, which are used to collect temperature data, humidity data, air pressure data and gas flow velocity data respectively.

6. The method for locating a leak source by fusion of multi-source gas data according to claim 5, characterized in that: The fusing of the gas concentration data and the physical environment data to perform inverse calculation of the leakage process of the leaked gas specifically includes: defining a gas propagation velocity function based on the temperature data, the humidity data, the air pressure data, and the gas flow velocity data; constructing an arrival time constraint equation based on the gas propagation velocity function, wherein the arrival time constraint equation is used to represent the path propagation time of the gas from the leakage source location to each second sensor, which is equal to the time difference between the second time and the first time; Establishing a set of simultaneous constraint equations for the plurality of gas sensors using the arrival time constraint equations, and constructing a residual function for each of the gas sensors; A nonlinear least squares optimization objective function in vector form is constructed through the residual functions of all the gas sensors, wherein the nonlinear least squares optimization objective function takes the estimated leakage source position and the first time as parameters to be determined, and minimizes the sum of the squares of the differences between the path propagation time of each node and the actual arrival time.

7. The method for locating a leak source by fusion of multi-source gas data according to claim 6, characterized in that: The method of determining the location of the leakage source based on the leakage process and mapping it back to the physical structure of the gas pipeline network specifically includes: Iteratively solving the nonlinear least squares optimization objective function, recalculating the residual value of the residual function according to the current estimated value of the parameter to be solved during each iterative solution process, and calculating the derivative matrix of the residual function with respect to the parameter to be solved, and adjusting the update step size by constructing a Jacobian matrix and a damping term; During the multiple iterative solving processes, if it is determined that the change of the parameter to be solved is less than a set threshold, the estimated leakage source position corresponding to the last parameter to be solved is output to obtain the leakage source position; The leakage source location is mapped to the gas structure topology map to determine the physical location of the leakage source.

8. A leakage source locating device based on multi-source gas data fusion, characterized in that: The device is used to execute a leakage source positioning method based on multi-source gas data fusion as claimed in any one of claims 1 to 7, and the device comprises an acquisition module (201), a processing module (202) and an output module (203), wherein: The acquisition module (201) is used to acquire multi-source data collected by a sensor system deployed in a target area, wherein the sensor system includes a plurality of gas sensors and a plurality of environmental sensors, the gas sensors are used to collect gas concentration data, and the environmental sensors are used to collect physical environment data; The processing module (202) is used to fuse the gas concentration data with the physical environment data and perform inverse calculation on the leakage process of the leaked gas; The output module (203) is used to determine the location of the leakage source based on the leakage process, and map it back to the physical structure of the gas pipeline network.

9. An electronic device, characterized in that: The electronic device comprises a processor (301), a communication bus (302), a user interface (303), a network interface (304) and a memory (305), wherein the memory (305) is used to store instructions, the user interface (303) and the network interface (304) are both used to communicate with other devices, the communication bus (302) is used to realize connection and communication between components in the electronic device, and the processor (301) is used to execute the instructions stored in the memory (305) so that the electronic device executes the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores instructions, and when the instructions are executed, the method according to any one of claims 1 to 7 is executed.

Images