Limited space gas diffusion source analysis method and system based on multi-sensor fusion

Abstract

The invention discloses a finite space gas diffusion source analysis method and system based on multi-sensor fusion, and relates to the technical field of finite space gas diffusion source analysis, and the method comprises the following steps: analyzing the gas leakage risk degree in a finite space; performing model construction on the risk space to obtain a target space model; performing correlation analysis on data of different sensors in the risk space, identifying a gas propagation path and direction in the risk space, constructing a gas diffusion directed graph of the risk space, determining the irregularity degree of gas flow in the risk space, and analyzing the true probability that different nodes in the gas diffusion directed graph are gas diffusion sources; according to the diffusion position data in the risk space, the troubleshooting priority condition of different suspected position coordinates of the gas diffusion source in the diffusion position data is analyzed, workers are informed of troubleshooting, the analysis efficiency is high, and the analysis accuracy of the diffusion source in the limited space gas is greatly improved compared with a traditional method.

Description

Technical Field

[0001] This invention relates to the field of confined space gas diffusion source analysis technology, specifically to a confined space gas diffusion source analysis method and system based on multi-sensor fusion. Background Technology

[0002] Confined spaces are generally enclosed or semi-enclosed spaces, typically containing pipes, storage tanks, underground pipe corridors, ship cabins, and reactors. Gas diffusion in confined spaces can easily cause significant hazards. For example, once flammable gas mixes with air in a confined space and reaches a critical limit, it can easily explode due to open flames, static electricity, or high-temperature surfaces, destroying equipment and structures. Therefore, identifying gas diffusion sources in confined spaces is particularly important. Traditional analysis of gas diffusion sources in confined spaces mainly relies on manual inspections and sparsely deployed single-type sensors. Manual inspections require professionals to carry portable equipment into hazardous environments, which is time-consuming and laborious, resulting in low efficiency and high personnel risks. On the other hand, sparsely deployed single-type sensors for analyzing gas diffusion sources in confined spaces only measure concentration and cannot simultaneously monitor other key influencing factors, which not only easily leads to blind spots but also results in a high false alarm rate.

[0003] Deploying multiple sensors in a confined space and using multi-sensor fusion technology to analyze gas diffusion sources in that space can not only solve the problem of blind spots in traditional analysis methods, but also improve the efficiency of investigation and avoid personnel safety issues caused by manual inspections. However, in the current process of using multiple sensors to analyze gas diffusion sources in confined spaces, there are irregular and complex airflows generated by ventilation, equipment operation, and personnel movement, which make gas diffusion highly disordered and changeable in an instant. This makes it impossible to accurately analyze gas diffusion sources in confined spaces by simply using classical diffusion models. It is easy to cause the actual gas diffusion source to be completely different from the gas diffusion source obtained by analysis. This not only wastes time, but also pollutes the air and may even lead to catastrophic chain reactions. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for analyzing gas diffusion sources in confined spaces based on multi-sensor fusion, so as to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for analyzing gas diffusion sources in confined spaces based on multi-sensor fusion, the method comprising:

[0006] Step S1: Obtain gas sampling data in a confined space, analyze the degree of gas leakage risk in the confined space, and obtain the risk space; Step S2: Acquire the three-dimensional spatial data and multi-sensor data of the risk space, construct a model of the risk space, and obtain the target space model; Step S3: Based on the target space model, perform relevant analysis on the data from different sensors in the risk space, identify the gas propagation path and direction in the risk space, construct a directed gas diffusion graph in the risk space, determine the degree of airflow irregularity in the risk space, analyze the true probability that different nodes in the directed gas diffusion graph are gas diffusion sources, and determine the diffusion location data in the risk space. Step S4: Based on the diffusion location data in the risk space, analyze the priority of investigating different suspected location coordinates of gas diffusion sources in the diffusion location data, and notify staff to conduct the investigation.

[0007] Furthermore, step S3 includes: Calculate the cross-correlation function p of the key gas monitoring parameters between the d-th gas sensor and the f-th gas sensor in the risk space. (d,f) [k], obtain the leading state and correlation of changes between the d-th gas sensor and the f-th gas sensor; Obtain several gas sensors with stable correlation in the risk space, and treat each of the gas sensors as a node. When the correlation between a gas sensor and another gas sensor is stable and the change of the gas sensor is ahead of that of the other gas sensor, then create an edge from the node of the gas sensor to the node of the other gas sensor. When several gas sensors are obtained as nodes, the edges between the nodes corresponding to the several gas sensors are used to construct a directed graph of gas diffusion in the risk space. Obtain each node at the starting point in the directed graph of gas diffusion, and obtain the total number of pointing edges in each node; Calculate the degree of airflow irregularity E in the risk space, set the airflow irregularity threshold e, when E≤e, the airflow in the risk space is determined to be irregular, and set the pointing edge threshold. When the total number of pointing edges of a node at the starting point in the gas diffusion directed graph is less than the pointing edge threshold, the node is determined to be a suspected gas diffusion source and recorded as the target node. When E>e, the airflow in the risk space is determined to be irregular, and the node at the starting point in the gas diffusion directed graph is recorded as the critical node. Calculate the diffusion probability value of each key node in the directed graph of gas diffusion, set the diffusion probability threshold W´, obtain the diffusion probability value W of a key node in the directed graph of gas diffusion, if W>W´, then the key node is determined to be a suspected gas diffusion source and recorded as the target node, otherwise the key node is determined not to be a suspected gas diffusion source. Obtain the position coordinates of the gas sensors corresponding to each target node in the gas diffusion directed graph in the target space model, and record them as the suspected position coordinates of the gas diffusion source in the risk space; The coordinates of each suspected gas diffusion source in the risk space are collected to obtain diffusion location data in the risk space; The above method not only enables accurate analysis of gas diffusion sources in confined spaces, but also solves the problem that traditional methods cannot address during the analysis of gas diffusion sources in confined spaces. This is because the analysis process is subject to irregular and complex airflow caused by ventilation, equipment operation, and personnel movement, which makes gas diffusion highly disordered and changeable in an instant. Even under complex airflow conditions, it is possible to quickly and accurately analyze and locate gas diffusion sources in confined spaces, thus effectively ensuring the safety of confined spaces.

[0008] Furthermore, step S2 includes: Acquire the three-dimensional spatial data of the risk space, including a three-dimensional scan of the risk space. Use a preset three-dimensional modeling software to perform three-dimensional modeling of the risk space to obtain the three-dimensional spatial model. Verify the mesh independence of the three-dimensional spatial model. The risk space is acquired by each sensor, and each sensor collects data on the risk space at preset unit time intervals within the current period to obtain multi-sensor data. The multi-sensor data includes the values ​​of the monitoring parameters of each sensor within each unit time interval of the current period. The timestamps of the monitoring parameters of each sensor in the current period are obtained from the multi-sensor data. The position coordinates of each sensor are obtained from the spatial three-dimensional model. The position coordinates of the sensors in the risk space and the values ​​of the monitoring parameters in a certain unit of time in the current period are collected to obtain the state vector of the sensor in a certain unit of time. The state vectors of the sensors are obtained within each unit of time in the current period, the state set of the sensors in the current period is obtained, the state set of each sensor in the risk space and the spatial 3D model are obtained, and the target space model of the risk space is constructed.

[0009] Furthermore, step S1 includes: Set a unit time interval. Within the current cycle, the sensor will collect gas monitoring parameters in the confined space at each unit time interval to obtain gas collection data of the confined space within the current cycle. The gas collection data includes the values ​​of various gas monitoring parameters in the confined space within each unit time interval of the current cycle. Obtain the rate of change of various gas monitoring parameters in a confined space within each unit of time in the current period from the gas acquisition data; Set the parameter threshold range and change rate threshold for each gas monitoring parameter. When the maximum or minimum value of a certain gas monitoring parameter in the gas collection data is not within the corresponding parameter threshold range in each unit time period, it is determined that there is a risk of gas leakage in the confined space, and the confined space is recorded as a risk space. Conversely, the maximum value of the rate of change of each gas monitoring parameter in the gas acquisition data within each unit of time is obtained. If the maximum value of the rate of change of a gas monitoring parameter in the gas acquisition data within each unit of time is greater than the rate of change threshold, it is determined that there is a risk of gas leakage in the confined space, and the confined space is recorded as a risk space. Otherwise, it is determined that there is no risk of gas leakage in the confined space.

[0010] Furthermore, step S4 includes: Acquire diffusion location data in the risk space, extract the coordinates of each suspected gas diffusion source from the diffusion location data, calculate the straight-line distance between each suspected location coordinate, and set a distance threshold. When the straight-line distance between one suspected location coordinate and another suspected location coordinate in the diffusion location data is less than the distance threshold, the other suspected location coordinate is designated as the nearest suspected location coordinate of one suspected location coordinate. Obtain the total number of similar suspected location coordinates among each suspected location coordinate. Set the investigation priority of the suspected location coordinate corresponding to the maximum total number of similar suspected location coordinates to the highest. Then, sort the suspected location coordinates in descending order of the total number of similar suspected location coordinates. The coordinates of each suspected location in the risk space are obtained and prioritized and aggregated to obtain priority data for investigation. This priority data is then sent to staff through the platform to notify them to investigate gas diffusion sources in the risk space.

[0011] To better implement the above methods, a confined space gas diffusion source analysis system based on multi-sensor fusion is also proposed. The system includes a risk analysis module, a model building module, a gas diffusion source analysis module, and an investigation module. The risk analysis module is used to analyze the degree of gas leakage risk in a confined space and obtain the risk space; The model building module is used to acquire spatial three-dimensional data and multi-sensor data of the risk space, build a model of the risk space, and obtain the target space model. The gas diffusion source analysis module is used to analyze the true probability that different nodes in the directed gas diffusion graph are gas diffusion sources and to determine the diffusion location data in the risk space. The investigation module is used to analyze the investigation priority of different suspected gas diffusion source coordinates in the diffusion location data based on the diffusion location data in the risk space, and notify staff to carry out the investigation.

[0012] Furthermore, the risk analysis module includes a data acquisition unit and a risk analysis unit; The data acquisition unit is used to collect gas monitoring parameters of the confined space by the sensor at unit intervals within the current cycle, so as to obtain the gas acquisition data of the confined space within the current cycle. The risk analysis unit is used to analyze the degree of gas leakage risk in a confined space based on gas collection data within the current period, and to obtain the risk space.

[0013] Furthermore, the model building module includes 3D modeling units and model building units; The 3D modeling unit is used to acquire the spatial 3D data of the risk space and use preset 3D modeling software to perform 3D modeling of the risk space to obtain the spatial 3D model. The model building unit is used to acquire the state set and three-dimensional model of each sensor in the risk space, and to construct the target space model of the risk space.

[0014] Furthermore, the gas diffusion source analysis module includes a gas diffusion directed graph construction unit and a gas diffusion source analysis unit; The gas diffusion directed graph construction unit is used to perform correlation analysis on the data of different sensors in the risk space based on the target space model, identify the gas propagation path and direction in the risk space, and construct the gas diffusion directed graph of the risk space. The gas diffusion source analysis unit is used to determine the degree of airflow irregularity in the risk space, analyze the true probability that different nodes in the directed gas diffusion graph are gas diffusion sources, and determine the diffusion location data in the risk space.

[0015] Furthermore, the investigation module includes investigation units; The investigation unit is used to analyze the investigation priority of different suspected gas diffusion source coordinates in the diffusion location data based on the diffusion location data in the risk space, obtain investigation priority data, and send the investigation priority data to the staff through the platform to notify the staff to investigate the gas diffusion sources in the risk space.

[0016] Compared with existing technologies, the beneficial effects of this invention are: by using multi-sensor fusion analysis to determine the risk level of gas leakage in a confined space, it can quickly assess the risk of gas leakage in a confined space. Furthermore, by constructing a model of the risk space, a target space model is obtained, and the data in the risk space is displayed more intuitively. In addition, the influence of airflow irregularity on the analysis of gas diffusion sources in the confined space is considered during the analysis process, thus solving the error caused by airflow irregularity in the analysis of gas diffusion sources. This achieves a high degree of consistency between the actual gas diffusion source and the analyzed gas diffusion source location. Not only is the analysis efficiency fast, but the accuracy of diffusion source analysis in confined space gas is also significantly improved compared with traditional methods. Attached Figure Description

[0017] Figure 1 This is a flowchart of the gas diffusion source analysis method of the present invention based on multi-sensor fusion for gas diffusion source analysis in confined space; Figure 2 This is a schematic diagram of the modules of the confined space gas diffusion source analysis system based on multi-sensor fusion of the present invention. Detailed Implementation

[0018] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Example: Figures 1-2 As shown, this invention provides a technical solution: a method for analyzing gas diffusion sources in confined spaces based on multi-sensor fusion, the method comprising: Step S1: Obtain gas sampling data in a confined space, analyze the degree of gas leakage risk in the confined space, and obtain the risk space; Step S1 includes: Set a unit time interval. Within the current cycle, the sensor will collect gas monitoring parameters in the confined space at each unit time interval to obtain gas collection data of the confined space within the current cycle. The gas collection data includes the values ​​of various gas monitoring parameters in the confined space within each unit time interval of the current cycle. For example, gas monitoring parameters include methane, carbon monoxide, etc. Obtain the rate of change of various gas monitoring parameters in a confined space within each unit of time in the current period from the gas acquisition data; For example, the rate of change va = (L...) of gas monitoring parameters in a confined space within the current period in the a-th unit of time can be obtained from gas acquisition data. (a+1) -L a ) / L a , where L (a+1) L represents the value of the gas monitoring parameter in the gas acquisition data within the (a+1)th unit of time in the current period. a To obtain the values ​​of gas monitoring parameters in a confined space within the current period's a-th unit of time from the gas acquisition data; Set the parameter threshold range and change rate threshold for each gas monitoring parameter. When the maximum or minimum value of a certain gas monitoring parameter in the gas collection data is not within the corresponding parameter threshold range in each unit time period, it is determined that there is a risk of gas leakage in the confined space, and the confined space is recorded as a risk space. Conversely, the maximum value of the rate of change of each gas monitoring parameter in the gas acquisition data within each unit of time is obtained. If the maximum value of the rate of change of a gas monitoring parameter in the gas acquisition data within each unit of time is greater than the rate of change threshold, it is determined that there is a risk of gas leakage in the confined space, and the confined space is recorded as a risk space. Otherwise, it is determined that there is no risk of gas leakage in the confined space.

[0020] Step S2: Acquire the three-dimensional spatial data and multi-sensor data of the risk space, construct a model of the risk space, and obtain the target space model; Step S2 includes: Acquire the three-dimensional spatial data of the risk space, including a three-dimensional scan of the risk space. Use a preset three-dimensional modeling software to perform three-dimensional modeling of the risk space to obtain the three-dimensional spatial model. Verify the mesh independence of the three-dimensional spatial model. For example, 3D modeling software includes SpaceClaim, etc. For example, the specific steps for mesh independence verification are as follows: 1. Select key monitoring parameters and assessment locations: Select target physical quantities, such as the values ​​of gas monitoring parameters and wall shear force; Assessment locations: Select locations such as sensor positions and ventilation openings within the hazardous space as the locations to be assessed. 2. Systematically encrypt the grid: The entire computational domain's grid size is refined at a fixed ratio (e.g., reduced to half its original size); 3. Quantitative assessment: Calculate the rate of change of the target physical quantity at each evaluation location after each encryption and before encryption. If the rate of change of the target physical quantity at each evaluation location after encryption and before encryption is less than a threshold, the solution is determined to have network independence. The risk space is acquired by each sensor, and each sensor collects data on the risk space at preset unit time intervals within the current period to obtain multi-sensor data. The multi-sensor data includes the values ​​of the monitoring parameters of each sensor within each unit time interval of the current period. The timestamps of the monitoring parameters of each sensor in the current period are obtained from the multi-sensor data. The position coordinates of each sensor are obtained from the spatial three-dimensional model. The position coordinates of the sensors in the risk space and the values ​​of the monitoring parameters in a certain unit of time in the current period are collected to obtain the state vector of the sensor in a certain unit of time. For example, the monitoring parameters include carbon monoxide, hydrogen sulfide, and wind speed; Obtain the state vector of the sensor in each unit of time within the current period, obtain the state set of the sensor in the current period, obtain the state set of each sensor in the risk space and the spatial three-dimensional model, and construct the target space model of the risk space. For example, the target space model of the risk space is constructed, and the specific construction process is as follows: The position coordinates and state sets of each sensor in the three-dimensional spatial modeling are obtained and aggregated to construct the target space model of the risk space. Among them, the information contained in the sensor in the target space model includes the sensor's position coordinates and state set.

[0021] Step S3: Based on the target space model, perform relevant analysis on the data from different sensors in the risk space, identify the gas propagation path and direction in the risk space, construct a directed gas diffusion graph in the risk space, determine the degree of airflow irregularity in the risk space, analyze the true probability that different nodes in the directed gas diffusion graph are gas diffusion sources, and determine the diffusion location data in the risk space. Step S3 includes: Calculate the cross-correlation function p of the key gas monitoring parameters between the d-th gas sensor and the f-th gas sensor in the risk space. (d,f) [k], obtain the leading state and correlation of changes between the d-th gas sensor and the f-th gas sensor; For example, calculating the cross-correlation function p (d,f) The specific process for [k] is as follows: Obtain the target space model of the risk space within the current period, obtain the gas leakage monitoring parameters that are determined to exist in the risk space within the current period, and record them as key gas monitoring parameters; Obtain the time series of parameters of key gases monitored by each gas sensor in the risk space; For example, the specific process for obtaining the time series of key gas monitoring parameters in a gas sensor is as follows: The state set of the gas sensor is obtained from the target space model. The values ​​of the key gas monitoring parameters of the gas sensor in each unit of time in the current period are obtained from the state set and aggregated to obtain the parameter time series of the key gas monitoring parameters in the gas sensor. Calculate the cross-correlation function of the key gas monitoring parameters between each gas sensor, where p is the cross-correlation function of the key gas monitoring parameters between the d-th gas sensor and the f-th gas sensor in the risk space. (d,f) [k], where k is the cross-correlation function p (d,f) The time-delay index of [k]; For example, the cross-correlation function p (d,f) The specific calculation process for [k] is as follows: Calculate the discrete cross-correlation function Rd of the key gas monitoring parameters between the d-th gas sensor and the f-th gas sensor. (d,f) [k]; , Where M is the maximum time delay; D[n] is the parameter time series of the key gas monitoring parameters in the d-th gas sensor; F[n] is the parameter time series of the key gas monitoring parameters in the f-th gas sensor; N is the total number of elements in D[n]; and k is the cross-correlation function p. (d,f) The time-delay index of [k]; For discrete cross-correlation function R (d,f) [k] is normalized to obtain the cross-correlation function p. (d,f) [k]; Cross-correlation function p (d,f) The formula for calculating [k] is: , Among them, R (d,d) [0] and R (f,f) [0] represents the autocorrelation values ​​of the d-th gas sensor and the f-th gas sensor at zero time delay; k is the cross-correlation function p (d,f) The time-delay index of [k]; For example, to obtain the leading state and correlation between the d-th gas sensor and the f-th gas sensor, the specific acquisition process is as follows: Obtain the cross-correlation function p (d,f) The peak value p with the largest absolute value in the [k] curve max Set the kurtosis threshold p', when pmax When p < p´, it is determined that the cross-correlation between the d-th gas sensor and the f-th gas sensor is not significant. When p max ≥ p´, it is determined that the cross-correlation between the d-th gas sensor and the f-th gas sensor is significant, and the peak value p max corresponding time delay τ max is obtained; When τ max > 0, it is determined that the change of the d-th gas sensor leads that of the f-th gas sensor. When τ max < 0, it is determined that the change of the d-th gas sensor leads that of the f-th gas sensor; Obtain the time delay τ´ corresponding to the peak value in the cross-correlation function between the f-th gas sensor and the d-th gas sensor. When the absolute value of the sum of τ max and τ´ is less than the preset absolute threshold, it is determined that the correlation between the d-th gas sensor and the f-th gas sensor is stable. Otherwise, it is determined that the correlation between the d-th gas sensor and the f-th gas sensor is unstable; Obtain several gas sensors with stable correlations in the risk space. Take each sensor in the several gas sensors as a node. When the correlation between a certain gas sensor and another gas sensor in the several gas sensors is stable, and the change of a certain gas sensor leads that of another gas sensor, an edge is created from the node of a certain gas sensor to the node of another gas sensor; When several gas sensors are taken as nodes, obtain the edges between the corresponding nodes of the several gas sensors, and construct a gas diffusion directed graph in the risk space; Obtain each node at the starting point in the gas diffusion directed graph, and obtain the total number of incoming edges of each node; For example, the specific way to obtain the total number of incoming edges of a node is as follows: If there is an edge from the node of a certain gas sensor to the node of another gas sensor, then there is one incoming edge for the node corresponding to the other gas sensor; As Figure 1 shown: Calculate the airflow irregularity degree E in the risk space, set the airflow irregularity threshold e. When E ≤ e, it is determined that the airflow in the risk space does not have irregularity, and set the incoming edge threshold. When the total number of incoming edges of a certain node at the starting point in the gas diffusion directed graph is less than the incoming edge threshold, it is determined that a certain node is a suspected gas diffusion source and is recorded as the target node; For example, the specific calculation process of calculating the airflow irregularity degree E in the risk space is as follows: The wind speeds measured by each wind speed sensor for monitoring the wind speed in the target space model within each unit time period in the current cycle; Calculate the wind speed fluctuation value C=σ / μ for a certain wind speed sensor in the risk space, where σ is the standard deviation of the wind speed of a certain wind speed sensor in each unit of time in the current period, and μ is the average value of the wind speed of a certain wind speed sensor in each unit of time in the current period. Calculate the degree of airflow irregularity E in the risk space: , Where j is the total number of wind speed sensors in the risk space; C i Let be the wind speed fluctuation value of the i-th wind speed sensor in the risk space; When E>e, the airflow in the risk space is determined to be irregular, and the node at the starting point in the gas diffusion directed graph is recorded as the critical node. Calculate the diffusion probability value of each key node in the directed graph of gas diffusion, set the diffusion probability threshold W´, obtain the diffusion probability value W of a key node in the directed graph of gas diffusion, if W>W´, then the key node is determined to be a suspected gas diffusion source and recorded as the target node, otherwise the key node is determined not to be a suspected gas diffusion source. The diffusion probability value of each key node in the directed graph of gas diffusion is calculated using the following formula: Obtain the prior probabilities of key nodes in the directed graph of gas diffusion, and use the Gaussian plume model to predict the predicted values ​​of key gas monitoring parameters for each key node in the directed graph of gas diffusion, and aggregate them to obtain the gas prediction vector H. For example, the prior probability of a critical node in a directed graph of gas diffusion is set based on experience. When the gas sensor corresponding to a critical node is near a valve, its prior probability is higher than that of a gas sensor near a straight pipe section. The values ​​of key gas monitoring parameters detected by the gas sensors corresponding to each key node in the directed graph of gas diffusion are obtained and collected in the current period. The gas observation vector G is constructed, the likelihood function of each key node is defined, and the posterior probability of each key node being a gas diffusion source is calculated according to Bayes' theorem and normalized to obtain the diffusion probability value of each key node. For example, the specific calculation process for determining the posterior probability that each key node is a gas diffusion source is as follows: Based on the gas prediction vector H and the gas observation vector G, define the likelihood function P(G|s) for the key node s, and calculate the posterior probability P(s|G) that the key node s is a gas diffusion source: , Where P(s) is the prior probability of key node s; P(d) is the marginal likelihood of the gas observation vector G of key node s. Obtain the position coordinates of the gas sensors corresponding to each target node in the gas diffusion directed graph in the target space model, and record them as the suspected position coordinates of the gas diffusion source in the risk space; The coordinates of various suspected gas diffusion sources in the risk space are collected and aggregated to obtain diffusion location data in the risk space.

[0022] Step S4: Based on the diffusion location data in the risk space, analyze the priority of investigating different suspected location coordinates of gas diffusion sources in the diffusion location data, and notify staff to conduct the investigation.

[0023] Step S4 includes: Acquire diffusion location data in the risk space, extract the coordinates of each suspected gas diffusion source from the diffusion location data, calculate the straight-line distance between each suspected location coordinate, and set a distance threshold. When the straight-line distance between one suspected location coordinate and another suspected location coordinate in the diffusion location data is less than the distance threshold, the other suspected location coordinate is designated as the nearest suspected location coordinate of one suspected location coordinate. Obtain the total number of similar suspected location coordinates among each suspected location coordinate. Set the investigation priority of the suspected location coordinate corresponding to the maximum total number of similar suspected location coordinates to the highest. Then, sort the suspected location coordinates in descending order of the total number of similar suspected location coordinates. The coordinates of each suspected location in the risk space are obtained and prioritized and aggregated to obtain priority data for investigation. This priority data is then sent to staff through the platform to notify them to investigate gas diffusion sources in the risk space.

[0024] To better implement the above methods, a confined space gas diffusion source analysis system based on multi-sensor fusion is also proposed. The system includes a risk analysis module, a model building module, a gas diffusion source analysis module, and an investigation module. The risk analysis module is used to analyze the degree of gas leakage risk in a confined space and obtain the risk space; The model building module is used to acquire spatial three-dimensional data and multi-sensor data of the risk space, build a model of the risk space, and obtain the target space model. The gas diffusion source analysis module is used to analyze the true probability that different nodes in the directed gas diffusion graph are gas diffusion sources and to determine the diffusion location data in the risk space. The investigation module is used to analyze the investigation priority of different suspected gas diffusion source coordinates in the diffusion location data based on the diffusion location data in the risk space, and notify staff to carry out the investigation.

[0025] The risk analysis module includes a data acquisition unit and a risk analysis unit. The data acquisition unit is used to collect gas monitoring parameters of the confined space by the sensor at unit intervals within the current cycle, so as to obtain the gas acquisition data of the confined space within the current cycle. The risk analysis unit is used to analyze the degree of gas leakage risk in a confined space based on gas collection data within the current period, and to obtain the risk space.

[0026] The model building module includes a 3D modeling unit and a model building unit. The 3D modeling unit is used to acquire the spatial 3D data of the risk space and use preset 3D modeling software to perform 3D modeling of the risk space to obtain the spatial 3D model. The model building unit is used to acquire the state set and three-dimensional model of each sensor in the risk space, and to construct the target space model of the risk space.

[0027] The gas diffusion source analysis module includes a gas diffusion directed graph construction unit and a gas diffusion source analysis unit. The gas diffusion directed graph construction unit is used to perform correlation analysis on the data of different sensors in the risk space based on the target space model, identify the gas propagation path and direction in the risk space, and construct the gas diffusion directed graph of the risk space. The gas diffusion source analysis unit is used to determine the degree of airflow irregularity in the risk space, analyze the true probability that different nodes in the directed gas diffusion graph are gas diffusion sources, and determine the diffusion location data in the risk space.

[0028] The investigation module includes investigation units; The investigation unit is used to analyze the investigation priority of different suspected gas diffusion source coordinates in the diffusion location data based on the diffusion location data in the risk space, obtain investigation priority data, and send the investigation priority data to the staff through the platform to notify the staff to investigate the gas diffusion sources in the risk space.

[0029] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for analyzing gas diffusion sources in confined spaces based on multi-sensor fusion, characterized in that, The method includes: Step S1: Obtain gas sampling data in a confined space, analyze the degree of gas leakage risk in the confined space, and obtain the risk space; Step S2: Acquire the three-dimensional spatial data and multi-sensor data of the risk space, construct a model of the risk space, and obtain the target space model; Step S3: Based on the target space model, perform relevant analysis on the data from different sensors in the risk space, identify the gas propagation path and direction in the risk space, construct a directed gas diffusion graph in the risk space, determine the degree of airflow irregularity in the risk space, analyze the true probability that different nodes in the directed gas diffusion graph are gas diffusion sources, and determine the diffusion location data in the risk space. Step S4: Based on the diffusion location data in the risk space, analyze the priority of investigating different suspected location coordinates of gas diffusion sources in the diffusion location data, and notify staff to conduct the investigation.

2. The method for analyzing gas diffusion sources in a confined space based on multi-sensor fusion according to claim 1, characterized in that, Step S3 includes: Calculate the cross-correlation function p of the key gas monitoring parameters between the d-th gas sensor and the f-th gas sensor in the risk space. (d,f) [k], obtain the leading state and correlation of changes between the d-th gas sensor and the f-th gas sensor; Obtain several gas sensors with stable correlation in the risk space, and treat each of the gas sensors as a node. When the correlation between a gas sensor and another gas sensor is stable and the change of the gas sensor is ahead of that of the other gas sensor, then create an edge from the node of the gas sensor to the node of the other gas sensor. When several gas sensors are obtained as nodes, the edges between the nodes corresponding to the several gas sensors are used to construct a directed graph of gas diffusion in the risk space. Obtain each node at the starting point in the directed graph of gas diffusion, and obtain the total number of pointing edges in each node; Calculate the degree of airflow irregularity E in the risk space, set the airflow irregularity threshold e, when E≤e, the airflow in the risk space is determined to be irregular, and set the pointing edge threshold. When the total number of pointing edges of a node at the starting point in the gas diffusion directed graph is less than the pointing edge threshold, the node is determined to be a suspected gas diffusion source and recorded as the target node. When E>e, the airflow in the risk space is determined to be irregular, and the node at the starting point in the gas diffusion directed graph is recorded as the critical node. Calculate the diffusion probability value of each key node in the directed graph of gas diffusion, set the diffusion probability threshold W´, obtain the diffusion probability value W of a key node in the directed graph of gas diffusion, if W>W´, then the key node is determined to be a suspected gas diffusion source and recorded as the target node, otherwise the key node is determined not to be a suspected gas diffusion source. Obtain the position coordinates of the gas sensors corresponding to each target node in the gas diffusion directed graph in the target space model, and record them as the suspected position coordinates of the gas diffusion source in the risk space; The coordinates of various suspected gas diffusion sources in the risk space are collected and aggregated to obtain diffusion location data in the risk space.

3. The method for analyzing gas diffusion sources in confined spaces based on multi-sensor fusion according to claim 1, characterized in that, Step S2 includes: Acquire the three-dimensional spatial data of the risk space, including a three-dimensional scan of the risk space. Use a preset three-dimensional modeling software to perform three-dimensional modeling of the risk space to obtain the three-dimensional spatial model. Verify the mesh independence of the three-dimensional spatial model. The risk space is acquired by each sensor, and each sensor collects data on the risk space at preset unit time intervals within the current period to obtain multi-sensor data. The multi-sensor data includes the values ​​of the monitoring parameters of each sensor within each unit time interval of the current period. The timestamps of the monitoring parameters of each sensor in the current period are obtained from the multi-sensor data. The position coordinates of each sensor are obtained from the spatial three-dimensional model. The position coordinates of the sensors in the risk space and the values ​​of the monitoring parameters in a certain unit of time in the current period are collected to obtain the state vector of the sensor in a certain unit of time. The state vectors of the sensors are obtained within each unit of time in the current period, the state set of the sensors in the current period is obtained, the state set of each sensor in the risk space and the spatial 3D model are obtained, and the target space model of the risk space is constructed.

4. The method for analyzing gas diffusion sources in confined spaces based on multi-sensor fusion according to claim 1, characterized in that, Step S1 includes: Set a unit time interval. Within the current cycle, the sensor will collect gas monitoring parameters in the confined space at each unit time interval to obtain gas collection data of the confined space within the current cycle. The gas collection data includes the values ​​of various gas monitoring parameters in the confined space within each unit time interval of the current cycle. Obtain the rate of change of various gas monitoring parameters in a confined space within each unit of time in the current period from the gas acquisition data; Set the parameter threshold range and change rate threshold for each gas monitoring parameter. When the maximum or minimum value of a certain gas monitoring parameter in the gas collection data is not within the corresponding parameter threshold range in each unit time period, it is determined that there is a risk of gas leakage in the confined space, and the confined space is recorded as a risk space. Conversely, the maximum value of the rate of change of each gas monitoring parameter in the gas acquisition data within each unit of time is obtained. If the maximum value of the rate of change of a gas monitoring parameter in the gas acquisition data within each unit of time is greater than the rate of change threshold, it is determined that there is a risk of gas leakage in the confined space, and the confined space is recorded as a risk space. Otherwise, it is determined that there is no risk of gas leakage in the confined space.

5. The method for analyzing gas diffusion sources in a confined space based on multi-sensor fusion according to claim 1, characterized in that, Step S4 includes: Acquire diffusion location data in the risk space, extract the coordinates of each suspected gas diffusion source from the diffusion location data, calculate the straight-line distance between each suspected location coordinate, and set a distance threshold. When the straight-line distance between one suspected location coordinate and another suspected location coordinate in the diffusion location data is less than the distance threshold, the other suspected location coordinate is designated as the nearest suspected location coordinate of one suspected location coordinate. Obtain the total number of similar suspected location coordinates among each suspected location coordinate. Set the investigation priority of the suspected location coordinate corresponding to the maximum total number of similar suspected location coordinates to the highest. Then, sort the suspected location coordinates in descending order of the total number of similar suspected location coordinates. The coordinates of each suspected location in the risk space are obtained and prioritized and aggregated to obtain priority data for investigation. This priority data is then sent to staff through the platform to notify them to investigate gas diffusion sources in the risk space.

6. A confined space gas diffusion source analysis system based on multi-sensor fusion, used to execute the confined space gas diffusion source analysis method based on multi-sensor fusion as described in any one of claims 1-5, characterized in that, The system includes a risk analysis module, a model building module, a gas diffusion source analysis module, and an investigation module; The risk analysis module is used to analyze the degree of gas leakage risk in a confined space to obtain the risk space; The model building module is used to acquire spatial three-dimensional data and multi-sensor data of the risk space, build a model of the risk space, and obtain a target space model. The gas diffusion source analysis module is used to analyze the true probability that different nodes in the directed gas diffusion graph are gas diffusion sources and determine the diffusion location data in the risk space. The investigation module is used to analyze the investigation priority of different suspected location coordinates of gas diffusion sources in the diffusion location data based on the diffusion location data in the risk space, and notify staff to carry out the investigation.

7. The confined space gas diffusion source analysis system based on multi-sensor fusion according to claim 6, characterized in that, The risk analysis module includes a data acquisition unit and a risk analysis unit; The data acquisition unit is used to collect gas monitoring parameters of the confined space by the sensor at unit intervals within the current period, so as to obtain gas acquisition data of the confined space within the current period. The risk analysis unit is used to analyze the degree of gas leakage risk in the confined space based on the gas collection data of the confined space in the current period, and to obtain the risk space.

8. The confined space gas diffusion source analysis system based on multi-sensor fusion according to claim 6, characterized in that, The model building module includes a 3D modeling unit and a model building unit; The three-dimensional modeling unit is used to acquire the spatial three-dimensional data of the risk space and use preset three-dimensional modeling software to perform three-dimensional modeling of the risk space to obtain a spatial three-dimensional model. The model building unit is used to acquire the state set and three-dimensional model of each sensor in the risk space, and to construct the target space model of the risk space.

9. The confined space gas diffusion source analysis system based on multi-sensor fusion according to claim 6, characterized in that, The gas diffusion source analysis module includes a gas diffusion directed graph construction unit and a gas diffusion source analysis unit; The gas diffusion directed graph construction unit is used to perform correlation analysis on the data of different sensors in the risk space according to the target space model, identify the gas propagation path and direction in the risk space, and construct the gas diffusion directed graph of the risk space. The gas diffusion source analysis unit is used to determine the degree of airflow irregularity in the risk space, analyze the true probability that different nodes in the directed gas diffusion graph are gas diffusion sources, and determine the diffusion location data in the risk space.

10. The confined space gas diffusion source analysis system based on multi-sensor fusion according to claim 6, characterized in that, The investigation module includes an investigation unit; The investigation unit is used to analyze the investigation priority of different suspected gas diffusion source coordinates in the diffusion location data based on the diffusion location data in the risk space, obtain investigation priority data, and send the investigation priority data to the staff through the platform to notify the staff to investigate the gas diffusion sources in the risk space.

Images