Cable trench toxic gas detection method, system, equipment and medium

Abstract

The invention discloses a cable trench toxic gas detection method, system and device and a medium, and relates to the technical field of gas detection, and the method comprises the following steps: collecting data and forming a time sequence monitoring data set; inputting the time sequence monitoring data set into the prediction model to obtain a predicted gas concentration result; comparing the predicted gas concentration result with a corresponding safety reference parameter, and judging whether a toxic gas abnormal condition exists in the cable trench or not; when it is judged that the toxic gas is abnormal, triggering an inspection device to perform cooperative inspection on a target area in the cable trench; and performing fusion processing on the monitoring data set, the predicted gas concentration result and the inspection result to generate a cable trench toxic gas detection result. According to the method, a traditional passive alarm mode depending on a fixed threshold value is converted into an active early warning mode based on trend analysis, single-point detection is upgraded into a comprehensive judgment mechanism combining multi-point collaborative analysis and spatial modeling, and the timeliness of toxic gas leakage discovery, the accuracy of positioning and the reliability of disposal are improved.

Description

Technical Field

[0001] This invention relates to the field of gas detection technology, and in particular to a method, system, equipment, and medium for detecting toxic gases in cable trenches. Background Technology

[0002] Cable trenches, as underground infrastructure used for laying cables in power systems, are characterized by enclosed spaces, limited ventilation, and complex environments. During long-term operation, issues such as cable insulation aging, loose joints, external environmental infiltration, or the accumulation of flammable and toxic gases can lead to the accumulation of toxic and flammable gases within the trenches, threatening the safety of maintenance personnel and the stable operation of the power system. To mitigate these risks, existing technologies commonly employ gas sensors within cable trenches to monitor gas concentrations at critical locations and trigger alarms when detected values ​​exceed preset thresholds. However, these solutions often rely on real-time data from fixed monitoring points, lacking comprehensive utilization of historical data and making it difficult to analyze and predict gas concentration trends, leading to delayed warnings or false alarms. Furthermore, existing detection methods often employ single-point or limited monitoring point deployment, failing to reflect the overall gas distribution within the cable trench and making it difficult to determine the extent and direction of abnormal areas. Leakage locations typically depend on manual inspections or even experience-based judgment, resulting in low efficiency and accuracy. In actual operation and maintenance, inspection work still mainly relies on manual entry into cable trenches. This involves harsh working environments and high risks, and is limited by space constraints and work cycles, making it difficult to achieve high-frequency, full-coverage inspections. Furthermore, different monitoring devices often operate independently, lacking unified management and fusion analysis methods for data. Inspection results are mostly presented as numerical values ​​or simple alarms, lacking intuitive visualization methods. This makes it difficult for management personnel to grasp the safety status within the cable trenches in a timely and comprehensive manner, failing to meet the current power system's demand for intelligent and refined management of safe operation. Summary of the Invention

[0003] In view of the above-mentioned problems, the present invention provides a method, system, equipment and medium for detecting toxic gases in cable trenches.

[0004] Therefore, the problem that this invention aims to solve is that existing technical solutions are mostly based on real-time numerical judgments from fixed monitoring points, lack comprehensive utilization of historical data, make it difficult to analyze and predict gas concentration change trends, and are prone to problems such as delayed early warnings or false alarms.

[0005] To address the aforementioned technical problems, this invention provides the following technical solution: a method for detecting toxic gases in cable trenches, comprising: collecting toxic gas concentration data and environmental parameter data related to gas diffusion at various monitoring locations within the cable trench to form a time-series monitoring dataset; inputting the time-series monitoring dataset into a prediction model to extract the temporal characteristics of gas concentration changes over time and the spatial correlation characteristics between monitoring locations, performing trend prediction on the gas concentration, and obtaining a predicted gas concentration result; comparing the predicted gas concentration result with corresponding safety benchmark parameters to determine whether there is an abnormal toxic gas situation within the cable trench, and generating an abnormal risk assessment result; when an abnormal toxic gas situation is determined, triggering an inspection device to perform collaborative inspection of the target area within the cable trench, verifying the abnormal area based on the gas concentration information obtained during the inspection, and determining the leakage location; fusing the monitoring dataset, the predicted gas concentration result, and the inspection result to generate a toxic gas detection result for the cable trench, and outputting the corresponding detection information.

[0006] As a preferred embodiment of the toxic gas detection method in cable trenches according to the present invention, the formation of the time-series monitoring dataset includes: synchronously collecting toxic gas concentration and environmental parameter data at each monitoring point by deploying monitoring nodes at different locations in the cable trench, and attaching unified time reference information to each collected data; performing unified data processing on the collected monitoring data; rearranging the processed monitoring data according to time order, and aligning, labeling, and structurally encapsulating the data from different monitoring locations to form a time-series monitoring dataset containing time dimension, spatial location identifier, and multi-dimensional environmental parameters.

[0007] As a preferred embodiment of the toxic gas detection method in cable trenches according to the present invention, the method for obtaining the predicted gas concentration includes: extracting temporal feature information characterizing the evolution of gas concentration over time based on historical gas concentration change data of each monitoring location in the time-series monitoring dataset; extracting spatial correlation feature information characterizing the diffusion relationship of gas inside the cable trench based on the data correlation relationship between different monitoring locations; fusing the temporal feature information and the spatial correlation feature information; and outputting the predicted gas concentration result for the target time period based on the fused feature result.

[0008] As a preferred embodiment of the method for detecting toxic gases in cable trenches according to the present invention, the generation of abnormal risk assessment results includes: comparing the predicted gas concentration result with the corresponding safety benchmark parameters item by item to determine whether the predicted gas concentration exceeds the preset safety limit; when the predicted gas concentration exceeds the safety limit, classifying the degree of abnormality to generate a corresponding risk level; and based on the risk level, generating an abnormal risk assessment result for toxic gases in cable trenches, indicating the current degree of abnormality and potential risk status.

[0009] As a preferred embodiment of the toxic gas detection method for cable trenches described in this invention, the collaborative inspection of the target area within the cable trench includes: constructing a navigation map for the target area based on the spatial structure information of the cable trench and the gas distribution information obtained through real-time monitoring; characterizing the spatial structure status and risk area distribution within the cable trench; and generating an inspection path covering the target area according to the navigation map, guiding the inspection device to scan abnormal areas while meeting spatial obstacle avoidance requirements.

[0010] The beneficial effects of this preferred technical solution are as follows: by combining cable trench spatial structure information and gas distribution information to construct a navigation map, and planning inspection routes on this basis, the inspection process is transformed from experience-based blind inspection to intelligent guidance based on risk situation, improving the coverage efficiency and hit rate of abnormal areas; at the same time, the path planning function reduces inspection blind spots and repeated scanning, improves the overall efficiency of collaborative inspection, and reduces the equipment operating burden.

[0011] As a preferred embodiment of the toxic gas detection method for cable trenches described in this invention, the method for determining the leak location includes: when the airborne gas sensor of the inspection device detects an abnormal concentration, immediately switching to a fine scanning mode to re-check the target area using a spiral progressive path; simultaneously initiating a collaborative verification program to retrieve historical data from nearby fixed monitoring points and cross-compare it with the current status; if the verification result remains abnormal, triggering a three-dimensional coordinate positioning command to control the inspection device to stop at the leak point, and determining the leak source location by combining laser ranging and visual marking.

[0012] The beneficial effects of this preferred technical solution are as follows: by cross-comparison, the leak point is identified based on the high concentration area, so that the determination of the leak location is changed from manual experience judgment to spatial model analysis, reducing the probability of false alarms and misjudgments and improving the positioning accuracy.

[0013] As a preferred embodiment of the toxic gas detection method for cable trenches described in this invention, the output of corresponding detection information includes: encapsulating monitoring data, prediction results, and inspection results to generate a detection result dataset; constructing a visualization interface for cable trench detection results based on the detection result dataset to present the detection information graphically; and outputting the visualized detection information to a monitoring terminal for maintenance personnel to view and make decisions.

[0014] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a toxic gas detection system for cable trenches, comprising: a data acquisition module, a prediction module, an evaluation module, a re-inspection module, and a detection information generation module; the data acquisition module acquires toxic gas concentration data and environmental parameter data related to gas diffusion at various monitoring locations within the cable trench, forming a time-series monitoring dataset; the prediction module inputs the time-series monitoring dataset into a prediction model, extracts the temporal characteristics of gas concentration changes over time and the spatial correlation characteristics between monitoring locations, performs trend prediction of gas concentration, and obtains the predicted gas concentration result. The assessment module compares the predicted gas concentration results with the corresponding safety benchmark parameters to determine whether there is an abnormal toxic gas situation in the cable trench and generates an abnormal risk assessment result. The re-inspection module, when an abnormal toxic gas situation is determined, triggers the inspection device to perform a collaborative inspection of the target area in the cable trench, verifies the abnormal area based on the gas concentration information obtained during the inspection, and determines the leak location. The detection information generation module fuses the monitoring dataset, the predicted gas concentration results, and the inspection results to generate the toxic gas detection results of the cable trench and outputs the corresponding detection information.

[0015] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the method for detecting toxic gases in cable trenches as described above.

[0016] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for detecting toxic gases in a cable trench as described above.

[0017] The beneficial effects of this invention are as follows: This invention transforms the traditional passive alarm method that relies on fixed thresholds into an active early warning mode based on trend analysis, upgrades single-point detection into a comprehensive judgment mechanism that combines multi-point collaborative analysis and spatial modeling, and transforms the inefficient inspection method based on manual inspection into an intelligent inspection mode based on navigation maps and path planning. This improves the timeliness of toxic gas leak detection, the accuracy of location, and the reliability of handling, reduces safety hazards during cable trench operation, reduces the probability of maintenance personnel directly contacting hazardous environments, and improves the overall safety of cable trench operation and the level of intelligent management. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of a method for detecting toxic gases in cable trenches in Example 1. Detailed Implementation

[0020] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0021] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0022] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides a method for detecting toxic gases in cable trenches, comprising: S1: Collect toxic gas concentration data and environmental parameter data related to gas diffusion at each monitoring location in the cable trench to form a time-series monitoring dataset.

[0023] S2: Input the time-series monitoring dataset into the prediction model, extract the time-series characteristics of gas concentration changes over time and the spatial correlation characteristics between monitoring locations, perform trend prediction of gas concentration, and obtain the predicted gas concentration result.

[0024] S3: Compare the predicted gas concentration results with the corresponding safety benchmark parameters to determine whether there is an abnormal situation of toxic gas in the cable trench, and generate an abnormal risk assessment result.

[0025] S4: When an abnormality of toxic gas is detected, the inspection device is triggered to perform a collaborative inspection of the target area in the cable trench. The abnormal area is verified based on the gas concentration information obtained during the inspection, and the location of the leak is determined.

[0026] S5: The monitoring dataset, predicted gas concentration results, and inspection results are fused and processed to generate the detection results of toxic gases in the cable trench and output the corresponding detection information.

[0027] It should be noted that existing toxic gas detection technologies in cable trenches generally rely on fixed threshold alarms, primarily depending on single-point sensor data. Alarms are only triggered when the gas concentration exceeds a preset threshold, making it difficult to identify the slow accumulation process in the early stages of a leak, resulting in delayed warnings. Furthermore, most systems lack the ability to comprehensively analyze historical data, failing to model and predict gas concentration trends, thus hindering timely detection of potential risks. In addition, existing technologies typically rely on static monitoring points, lacking comprehensive analysis of gas diffusion characteristics, making it impossible to determine the extent and direction of abnormal areas. Leak locations often depend on manual investigation or even experience-based judgment, resulting in low location efficiency and large errors.

[0028] Therefore, in response to the above problems, such as Figure 1 As shown, through steps S1-S5, the gas concentration and environmental parameters collected by multiple monitoring nodes deployed in the cable trench are first collected and organized to build a monitoring data foundation with temporal continuity. On this basis, the gas change patterns and spatial diffusion characteristics are comprehensively analyzed to form a judgment on the concentration change trend. When the analysis results indicate that there is a potential risk, the degree of anomaly is graded and assessed, and a collaborative inspection of the target area is automatically initiated accordingly. The inspection device is guided by a navigation map to focus on scanning high-risk areas. During the inspection, high-density monitoring data of the abnormal area is acquired, and the local gas distribution characteristics are further analyzed to pinpoint possible leakage sources. Finally, the risk level, inspection results, and leakage location information are integrated and output to the monitoring terminal in a visual manner, providing managers with intuitive and reliable decision support.

[0029] Example 2, a second embodiment of the present invention, differs from the first embodiment in that: a method for detecting toxic gases in cable trenches further includes, in step S1, forming a time-series monitoring dataset, comprising the following steps A1-A3: A1: By deploying monitoring nodes at different locations in the cable trench, the concentration of toxic gases and environmental parameters at each monitoring point are collected synchronously, and a unified time reference information is added to each piece of collected data.

[0030] A2: Perform unified data processing on the collected monitoring data to eliminate the impact of differences in environmental conditions, collection errors, or data fluctuations on the detection results, so that the monitoring data meets the input requirements for subsequent time series organization and predictive analysis.

[0031] A3: The processed monitoring data is rearranged according to time sequence, and the data from different monitoring locations are aligned, labeled, and structured to form a time-series monitoring dataset containing time dimension, spatial location identifier, and multi-dimensional environmental parameters.

[0032] In this embodiment of the application, step A2 involves data processing using a data normalization method based on environmental compensation and consistency correction, including the following steps A211-A213: A211: Based on the preset normal working range threshold, the monitoring data is quickly screened, and data that deviates significantly from the normal range is marked as abnormal candidate data. Based on the occurrence time and the working status of the associated monitoring nodes, the data is initially removed or corrected to reduce the impact of occasional anomalies on the overall data quality.

[0033] A212: By utilizing synchronously collected temperature, humidity, and air pressure parameters, environmental factors are eliminated from toxic gas concentration data. The original concentration values ​​are adjusted according to the influence of environmental parameter changes on the detection results, making the data obtained under different working conditions comparable.

[0034] A213: Cross-compare and adjust the consistency of gas concentration data from different monitoring locations to eliminate systematic biases introduced by differences in the sensitivity of different sensors or differences in installation location, and output the corrected unified standard monitoring data.

[0035] In an optional implementation, data processing may also employ a data normalization method based on statistical stabilization, including the following steps A221-A223: A221: Perform time-segmented statistical analysis on the monitoring data, identify data samples that significantly deviate from the overall distribution characteristics, and perform replacement, smoothing, or weight reduction processing on such samples to reduce the interference of local outliers on the overall time series characteristics.

[0036] A222: By stabilizing the raw monitoring data within a time window, the impact of short-term drastic fluctuations is weakened, making the processed data show a smoother and more continuous trend in the time dimension.

[0037] A223: Perform unified scale mapping processing on gas concentration data and environmental parameter data to enable data of different dimensions to have a consistent data expression form in the same numerical space, and enhance the fusion of data of different dimensions.

[0038] In another alternative implementation, data processing may also employ a model-based data normalization method, including the following steps A231-A233: A231: Construct corresponding data deviation models for each monitoring node to characterize the systematic error characteristics introduced by factors such as sensor aging and deployment differences.

[0039] A232: By using a deviation model to dynamically correct the collected data, the gas concentration data can be made closer to the real environmental conditions, thereby improving the reliability of the data.

[0040] A233: Output the multi-source data after model correction in a unified format to meet the input requirements for data continuity and consistency in subsequent steps.

[0041] It should be noted that by aligning and standardizing the data from multiple monitoring nodes within the cable trench using a unified time reference, the originally discrete and asynchronous multi-source monitoring data is integrated into a structured time-series monitoring dataset. This provides a consistent data foundation for subsequent gas concentration prediction, thereby improving the stability and accuracy of the prediction model. At the same time, by eliminating the impact of changes in environmental conditions and sensor errors on the original detection values, the reliability of the monitoring data is improved, providing a high-quality input data foundation for trend analysis.

[0042] Furthermore, in step S2, obtaining the predicted gas concentration result includes the following steps B1-B3: B1: Based on the historical gas concentration change data of each monitoring location in the time-series monitoring dataset, extract time-series feature information that characterizes the relationship between gas concentration and time evolution.

[0043] B2: Based on the data correlation between different monitoring locations, extract spatial correlation feature information that characterizes the diffusion relationship of gas inside the cable trench.

[0044] B3: The temporal feature information and spatial correlation feature information are fused together, and the predicted gas concentration results for the target time period are output based on the fused feature results.

[0045] Specifically, a dual-channel neural network architecture is used to process spatiotemporal features. The time series channel uses LSTM (Long Short-Term Memory) to capture the dynamic evolution of gas concentration and uses a gating mechanism to filter key time node features. The spatial correlation channel uses a convolutional neural network to extract the spatial distribution features among multiple sensors and establish a mapping relationship between gas diffusion paths and leakage source locations.

[0046] In this embodiment of the application, step B3 employs a dual-channel feature weighted fusion prediction method, including the following steps B311-B313: B311: Vectorize the temporal and spatial correlation features and construct a fusion feature input by weighted concatenation to fully characterize the temporal and spatial diffusion attributes of gas concentration changes.

[0047] B312: Input the fused features into the prediction network, and establish the correspondence between the fused features and the prediction parameters through fusion mapping processing to achieve unified modeling of the coupling relationship between features.

[0048] B313: Based on the fusion mapping processing results, output the predicted gas concentration results for the target time period.

[0049] In an optional implementation, the fusion process may also employ a fusion prediction method based on confidence weight adjustment, including the following steps B321-B323: B321: Perform reliability assessments on temporal features and spatial correlation features respectively, and generate corresponding feature confidence parameters.

[0050] B322: Based on the feature confidence parameter, assign corresponding fusion weights to the two types of features to construct a weight-adaptive fusion feature input.

[0051] B323: Input the weighted adaptive fusion features into the prediction module to obtain the predicted gas concentration results.

[0052] In another alternative implementation, the fusion processing may also employ a fusion prediction method based on phased collaborative prediction and online correction, including the following steps B331-B333: B331: Input the temporal features and spatial correlation features into the corresponding prediction sub-models to generate multiple stage prediction results, including short-term prediction results based on the time dimension and regional prediction results based on spatial diffusion characteristics, which are used to characterize the changing trend of gas in the temporal evolution direction and the diffusion status in the spatial distribution direction, respectively.

[0053] B332: Perform consistency assessment on the interim prediction results, identify the degree of deviation between different prediction results, and correct prediction results with large deviations according to the preset collaborative rules, so that the prediction results of each stage are consistent in terms of change trend and numerical range, so as to generate the fusion prediction intermediate results after collaborative correction.

[0054] B333: The intermediate prediction results after collaborative correction are integrated and processed to generate the final predicted gas concentration result, which serves as the predicted output of the gas concentration level within the target time period.

[0055] It should be noted that by simultaneously extracting the temporal evolution characteristics and spatial correlation characteristics of gas concentration, single-point detection is upgraded to multi-point collaborative analysis, avoiding the problem of misjudgment based solely on local monitoring results. This enables the prediction results to reflect the diffusion trend and spatial behavior of gas within the cable trench, effectively enhancing the ability to detect potential leakage risks in advance. Thus, the traditional "post-event alarm" is transformed into "pre-event warning," significantly improving the initiative and foresight of toxic gas detection.

[0056] Furthermore, in step S3, generating the anomaly risk assessment result includes the following steps C1-C3: C1: Compare the predicted gas concentration results with the corresponding safety benchmark parameters item by item to determine whether the predicted gas concentration exceeds the preset safety limit.

[0057] C2: When the predicted gas concentration exceeds the safety limit, the degree of anomaly is classified and a corresponding risk level is generated.

[0058] C3: Based on the risk level, generate an abnormal risk assessment result for toxic gases in cable trenches, indicating the current degree of abnormality and potential risk status.

[0059] In this embodiment of the application, step C2, the grading process adopts a grading evaluation method based on "degree of exceeding the limit + trend of change", including the following steps C211-C213: C211: Based on the difference between the predicted gas concentration and the corresponding safety benchmark parameter, calculate the extent of gas concentration exceeding the limit, and convert the extent of exceeding the limit into an amplitude index value for risk assessment. The amplitude index value can be discretized into intervals according to different gas types or risk level requirements to form a graded label for distinguishing between "minor exceedance", "moderate exceedance" and "severe exceedance", which is used to initially characterize the severity of the abnormality.

[0060] C212: Analyze the direction and rate of change of predicted gas concentration in multiple consecutive time slices, extract the trend characteristics of concentration change, and statistically analyze the duration of abnormal states to distinguish between instantaneous fluctuation-type abnormal states and continuous accumulation-type abnormal states; by analyzing the growth rate, stability level and consistency of changes of predicted concentration in the time dimension, obtain trend indicators reflecting the evolution of risk.

[0061] C213: Jointly determine the over-limit amplitude indicator and trend indicator, and match them in a predefined risk level mapping table to output the corresponding abnormal risk level; wherein, the risk level includes at least multiple levels such as safety, warning and danger, which are used to trigger different emergency strategies at the system level.

[0062] In an optional implementation, the tiered processing may also employ a risk assessment method based on a comprehensive score, including the following steps C221-C223: C221: The deviation, rate of change, and duration of the predicted gas concentration results from the safety benchmark parameters are mapped to comparable quantitative index values, and the indices with different physical meanings are numerically normalized to eliminate the influence of dimensional differences on the evaluation results, so that each index has a unified evaluation basis.

[0063] C222: Based on the set indicator weighting rules, the quantitative indicator values ​​are linearly or non-linearly combined to generate a comprehensive risk score; the weight ratio of each indicator in the comprehensive risk score can be dynamically adjusted according to historical operating experience, the importance of the detection point or risk preference, so as to improve the degree of matching of the score results with the actual working conditions.

[0064] C223: Match the comprehensive risk score with the risk level range to determine the risk range where the score value is located and output the corresponding risk level; among them, each risk level range can be configured and revised according to the actual operating environment to adapt to different cable trench structure conditions and environmental differences.

[0065] In another alternative implementation, the tiered processing may also employ a rule-based risk assessment method, including the following steps C231-C233: C231: Construct a risk assessment rule set covering multiple abnormal scenarios, and use the predicted gas concentration results and their change characteristics as input parameters for rule matching. By matching and comparing the prediction results with the triggering conditions of each risk rule, target rules that meet the triggering conditions are selected.

[0066] C232: When the predicted concentration result meets the threshold condition, combination condition or trend condition set by a certain risk rule, execute the trigger judgment logic of the corresponding rule and identify the risk type or risk level corresponding to the current abnormal state.

[0067] C233: Based on the level attribute of the triggering rule, output the final abnormal risk level and simultaneously generate risk identification information for recording abnormal status to support the triggering decision of subsequent collaborative inspection and handling steps.

[0068] It should be noted that by comparing the prediction results with safety benchmark parameters at each level and assessing the risk level, the abnormal state can be quantitatively described and managed hierarchically. This enables the system to distinguish between risk situations of different severity levels, thereby avoiding the coarse-grained judgment of "single alarm scale" in traditional detection and improving the accuracy of decision-making. At the same time, it provides a basis for the formulation of subsequent inspection strategies, allowing inspection resources to be prioritized for high-risk areas and improving overall operational efficiency.

[0069] Furthermore, in step S4, the collaborative inspection of the target area within the cable trench includes the following steps D1-D2: D1: Based on the spatial structure information of the cable trench and the gas distribution information obtained from real-time monitoring, a navigation map is constructed for the target area to characterize the spatial structure status and risk area distribution inside the cable trench.

[0070] D2: Based on the navigation map, generate an inspection path covering the target area, guiding the inspection device to scan abnormal areas while meeting spatial obstacle avoidance requirements.

[0071] Specifically, this embodiment employs UAVs for collaborative patrols. Multi-source sensor fusion technology is used to spatially register the 3D point cloud data of the cable trench with real-time gas concentration distribution maps, constructing a navigation map with obstacle markers and hazard area classifications. An ant colony algorithm is used to achieve autonomous patrol path planning, and a concentration gradient guiding factor is introduced into the traditional pheromone update mechanism, enabling the UAV to prioritize the exploration of high-risk areas while avoiding structural obstacles.

[0072] While meeting the requirements of full coverage scanning, the algorithm minimizes flight energy consumption and maximizes proximity to suspected leak points. It employs a segmented search strategy, dividing the cable trench into several sub-units based on functional areas, and dynamically adjusting the inspection priority of each unit according to historical detection data. Multiple standard flight attitude template libraries are pre-set, and the optimal passage method is automatically selected through real-time point cloud matching to ensure stable flight in confined spaces.

[0073] To further explain, determining the location of the leak includes the following steps D3-D5: D3: When the airborne gas sensor of the inspection device detects an abnormal concentration, it immediately switches to fine scanning mode and uses a spiral progressive path to re-check the target area.

[0074] D4: Simultaneously initiate the collaborative verification process to retrieve historical data from nearby fixed monitoring points and cross-compare it with the current status.

[0075] D5: If the verification results continue to be abnormal, a three-dimensional coordinate positioning command will be triggered to control the inspection device to stop at the leak point and determine the location of the leak source by combining laser ranging and visual marking.

[0076] Specifically, when the airborne gas sensor detects an abnormal concentration, the UAV immediately switches to a fine scanning mode, using a spiral-progressive path to perform multi-angle verification and detection of the target area. Simultaneously, a collaborative verification program is initiated, retrieving historical data from nearby fixed monitoring points and cross-referencing it with the current status. If the verification results remain abnormal, the system automatically triggers a three-dimensional coordinate positioning command, controlling the UAV to hover above the leak point and determining the precise location of the leak source through a combination of laser ranging and visual marking. To improve communication reliability in complex environments, a multi-link redundant transmission scheme was designed. The main control signal uses a 5.8GHz high-frequency channel to ensure real-time performance, while auxiliary positioning data is transmitted redundantly through a 433MHz low-frequency channel. The ground control station is equipped with an augmented reality (AR) display interface that overlays real-time UAV imagery with a digital twin model of the cable trench, providing operators with intuitive spatial situational awareness.

[0077] Furthermore, in step S5, outputting the corresponding detection information includes the following steps E1-E3: E1: Encapsulate monitoring data, prediction results, and inspection results to generate a dataset of inspection results.

[0078] E2: Based on the detection result dataset, a visualization interface for cable trench detection results is constructed to present the detection information in a graphical manner.

[0079] E3: Outputs visual detection information to the monitoring terminal for maintenance personnel to view and make decisions.

[0080] In this embodiment of the application, in step E2, the visualization interface adopts a visualization display method based on a three-dimensional model, including the following steps E211-E213: E211: Construct a three-dimensional display model based on the structural data of the cable trench, and overlay the distribution information of display devices and spatial structure information on the three-dimensional model.

[0081] E212: Maps the leak location, risk level, and prediction results to the corresponding spatial location in the 3D display model, and intuitively displays the abnormal state through color gradients or labels.

[0082] E213: Supports scaling, rotation, and layered display of 3D display models, enabling managers to understand the safety situation inside cable trenches from multiple perspectives.

[0083] In an optional implementation, the visualization interface may also employ a visualization method based on a heat map, including the following steps E221-E223: E221: Divide the cable trench space into several display units, each display unit corresponding to a unique space range.

[0084] E222: Different colors or color levels are used to identify each display unit according to its corresponding risk level or gas concentration value in order to form a gas risk thermal distribution map.

[0085] E223: The thermal distribution map is updated in real time as the detection results change to show the gas diffusion trend.

[0086] In another alternative implementation, the visualization interface may also employ a visualization method based on a combination of line graphs and lists, including the following steps E231-E233: E231: Classify and organize various types of data in the detection result dataset to form time series data and state data.

[0087] E232: Generates gas concentration trend curves based on time series data and displays the corresponding risk levels and leak locations in a list format.

[0088] E233: Enables linked display between trend curves and status lists, providing a visual representation of gas change processes and abnormal states.

[0089] It should be noted that by packaging, visualizing, and publishing the test results, complex test data can be presented to managers in an intuitive form, improving the efficiency of information acquisition and the accuracy of understanding. At the same time, by visually displaying the risk level, leakage location, and changing trends, managers can quickly grasp the on-site situation, improve emergency response speed and decision reliability, and enhance the overall availability of the system.

[0090] Example 3, the third embodiment of the present invention, differs from the previous two embodiments in that: a toxic gas detection system for cable trenches includes a data acquisition module, a prediction module, an evaluation module, a re-inspection module, and a detection information generation module; the data acquisition module collects toxic gas concentration data and environmental parameter data related to gas diffusion at various monitoring locations within the cable trench, forming a time-series monitoring dataset; the prediction module inputs the time-series monitoring dataset into a prediction model, extracts the time-series characteristics of gas concentration changes over time and the spatial correlation characteristics between monitoring locations, performs trend prediction of gas concentration, and obtains the predicted gas concentration result; the evaluation module compares the predicted gas concentration result with the corresponding safety benchmark parameters, determines whether there is an abnormal toxic gas situation in the cable trench, and generates an abnormal risk assessment result; when an abnormal toxic gas situation is determined, the re-inspection module triggers an inspection device to perform a collaborative inspection of the target area within the cable trench, verifies the abnormal area based on the gas concentration information obtained during the inspection, and determines the leak location; the detection information generation module fuses the monitoring dataset, the predicted gas concentration result, and the inspection result to generate a toxic gas detection result for the cable trench and outputs the corresponding detection information.

[0091] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0092] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0093] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0094] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented in combination with any of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0095] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for detecting toxic gases in cable trenches, characterized in that: include, Data on toxic gas concentrations and environmental parameters related to gas diffusion were collected at various monitoring locations within the cable trench to form a time-series monitoring dataset. The time-series monitoring dataset is input into the prediction model to extract the time-series characteristics of gas concentration changes over time and the spatial correlation characteristics between monitoring locations. The gas concentration is then used to predict the trend and obtain the predicted gas concentration result. The predicted gas concentration results are compared with the corresponding safety benchmark parameters to determine whether there is an abnormal situation of toxic gas in the cable trench, and an abnormal risk assessment result is generated. When an anomaly of toxic gas is detected, the inspection device is triggered to perform a coordinated inspection of the target area in the cable trench. The gas concentration information obtained during the inspection is used to verify the abnormal area and determine the location of the leak. The monitoring dataset, predicted gas concentration results, and inspection results are fused together to generate toxic gas detection results for cable trenches and output the corresponding detection information.

2. The method for detecting toxic gases in cable trenches as described in claim 1, characterized in that: The time-series monitoring dataset includes, By deploying monitoring nodes at different locations in the cable trench, the concentration of toxic gases and environmental parameters at each monitoring point are collected synchronously, and a unified time reference information is added to each piece of collected data. The collected monitoring data undergoes unified data processing; The processed monitoring data is rearranged according to time sequence, and data from different monitoring locations are aligned, labeled, and structured to form a time-series monitoring dataset containing time dimension, spatial location identifier, and multi-dimensional environmental parameters.

3. The method for detecting toxic gases in cable trenches as described in claim 2, characterized in that: The obtained predicted gas concentration results include, Based on the historical gas concentration change data of each monitoring location in the time-series monitoring dataset, time-series feature information characterizing the evolution of gas concentration over time is extracted; Based on the data correlation between different monitoring locations, spatial correlation feature information characterizing the diffusion relationship of gas inside the cable trench is extracted; The temporal feature information and spatial correlation feature information are fused together, and the predicted gas concentration results for the target time period are output based on the fused feature results.

4. The method for detecting toxic gases in cable trenches as described in claim 3, characterized in that: The generated anomaly risk assessment results include, The predicted gas concentration results are compared with the corresponding safety benchmark parameters one by one to determine whether the predicted gas concentration exceeds the preset safety limit. When the predicted gas concentration exceeds the safety limit, the degree of anomaly is classified and a corresponding risk level is generated. Based on the risk level, an abnormal risk assessment result for toxic gases in the cable trench is generated, indicating the current degree of abnormality and potential risk status.

5. The method for detecting toxic gases in cable trenches as described in claim 4, characterized in that: The coordinated inspection of the target area within the cable trench includes, Based on the spatial structure information of the cable trench and the gas distribution information obtained from real-time monitoring, a navigation map is constructed for the target area to characterize the spatial structure status and risk area distribution inside the cable trench. Based on the navigation map, an inspection path covering the target area is generated, guiding the inspection device to scan abnormal areas while meeting spatial obstacle avoidance requirements.

6. The method for detecting toxic gases in cable trenches as described in claim 5, characterized in that: Determining the location of the leak includes, When the airborne gas sensor of the inspection device detects an abnormal concentration, it immediately switches to fine scanning mode and uses a spiral progressive path to re-check and inspect the target area. Simultaneously, a collaborative verification procedure is initiated to retrieve historical data from nearby fixed monitoring points and cross-compare it with the current status. If the verification results remain abnormal, a three-dimensional coordinate positioning command is triggered, controlling the inspection device to stop at the leak point and determine the location of the leak source by combining laser ranging and visual marking.

7. The method for detecting toxic gases in cable trenches as described in claim 6, characterized in that: The detection information corresponding to the output includes: The monitoring data, prediction results, and inspection results are encapsulated to generate a dataset of inspection results. A visualization interface for cable trench inspection results is constructed based on the aforementioned inspection result dataset to present the inspection information in a graphical manner. Visualized detection information is output to the monitoring terminal for maintenance personnel to view and make decisions.

8. A toxic gas detection system for cable trenches, employing the toxic gas detection method for cable trenches as described in any one of claims 1 to 7, characterized in that: It includes a data acquisition module, a prediction module, an evaluation module, a re-inspection module, and a test information generation module; The data acquisition module collects toxic gas concentration data and environmental parameter data related to gas diffusion at each monitoring location in the cable trench, forming a time-series monitoring dataset. The prediction module inputs the time-series monitoring dataset into the prediction model, extracts the time-series characteristics of gas concentration changes over time and the spatial correlation characteristics between monitoring locations, performs trend prediction of gas concentration, and obtains the predicted gas concentration result. The assessment module compares the predicted gas concentration results with the corresponding safety benchmark parameters to determine whether there is an abnormal situation of toxic gas in the cable trench, and generates an abnormal risk assessment result. When the re-inspection module detects an abnormality in toxic gas, it triggers the inspection device to perform a collaborative inspection of the target area in the cable trench. Based on the gas concentration information obtained during the inspection, the abnormal area is re-verified, and the location of the leak is determined. The detection information generation module integrates the monitoring dataset, predicted gas concentration results, and inspection results to generate toxic gas detection results for cable trenches and outputs the corresponding detection information.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the method for detecting toxic gases in cable trenches according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the method for detecting toxic gases in cable trenches according to any one of claims 1 to 7.

Images