CN-122022099-A - Hydraulic engineering inspection management and protection method and system

CN122022099ACN 122022099 ACN122022099 ACN 122022099ACN-122022099-A

Abstract

The invention is suitable for the technical field of hydraulic engineering informatization, and provides a hydraulic engineering inspection management and protection method and system, wherein the method comprises the steps of collecting shoreline parameter data, water area parameter data and peri-base environment data; the method comprises the steps of preprocessing shore line parameter data, water area parameter data and peri-base environment data to obtain a shore line characteristic data set, a water area characteristic data set and a peri-base environment characteristic data set, carrying out weighted fusion on the shore line characteristic data set, the water area characteristic data set and the peri-base environment characteristic data set to obtain a comprehensive characteristic data set, constructing a potential safety hazard prediction model, inputting the comprehensive characteristic data set into the potential safety hazard prediction model, predicting to obtain a potential safety hazard coefficient of hydraulic engineering, comparing the predicted potential safety hazard coefficient of the hydraulic engineering with a preset potential safety hazard coefficient threshold value, and judging whether the hydraulic engineering has potential safety hazards.

Inventors

YU YU
LIU HAO
DU LIJIAO
ZHANG LIMIN
LIU XIAOBIN
WANG SHUAI
YAN HONGYAN
ZHANG WEN
LI BINGXING
LIU MIN

Assignees

中水智禹(天津)信息科技发展有限公司

Dates

Publication Date: 20260512
Application Date: 20260129

Claims (10)

1. The hydraulic engineering inspection management and protection method is characterized by comprising the following steps of: s1, acquiring shoreline parameter data, water area parameter data, peri-library environment data and video monitoring data; S2, preprocessing the shoreline parameter data, the water area parameter data and the peri-base environment data to obtain a shoreline characteristic data set, a water area characteristic data set and a peri-base environment characteristic data set, and carrying out weighted fusion on the shoreline characteristic data set, the water area characteristic data set and the peri-base environment characteristic data set to obtain a comprehensive characteristic data set; s3, constructing a potential safety hazard prediction model, inputting the comprehensive characteristic data set into the potential safety hazard prediction model, and predicting to obtain a potential safety hazard coefficient of the hydraulic engineering; s4, comparing the predicted potential safety hazard coefficient of the hydraulic engineering with a preset potential safety hazard coefficient threshold value, and judging whether the potential safety hazard exists in the hydraulic engineering; S5, if the hydraulic engineering has no hidden trouble, a patrol terminal sends a monitoring instruction, and a worker carries out patrol on a preset patrol point through video monitoring data; if hidden danger exists in the hydraulic engineering, associating the shoreline characteristic data set with the peri-library environment characteristic data set to construct a GIS-BIM three-dimensional space information model; S6, constructing a hidden danger type prediction model, inputting a hydraulic engineering potential safety hazard coefficient into the hidden danger type prediction model, predicting to obtain hidden danger type grades, making a routing inspection route and an overhaul scheme in the GIS-BIM three-dimensional space information model according to the hidden danger type grades, and selecting an optimal routing inspection route and an optimal overhaul scheme through an A-type algorithm.
2. The hydraulic engineering inspection management method according to claim 1, wherein the shoreline parameter data comprises the length, gradient, material of the shoreline and distribution position of hydraulic engineering facilities; The water area parameter data comprise water level, water flow speed, water flow shape and water area of the water area; The peri-warehouse environment data comprise peri-warehouse meteorological data, peri-warehouse vegetation coverage rate and peri-warehouse geological disaster occurrence probability; The video surveillance data includes continuous video frame data, geographic coordinates of the camera, a rotation angle of the camera, and a time stamp of each frame of video data.
3. The hydraulic engineering inspection management method according to claim 2, wherein in S2, the specific steps of obtaining the comprehensive feature data set are as follows: Identifying and removing abnormal data in the shoreline parameter data, the water area parameter data and the peri-base environment data through an LOF algorithm to obtain a processed shoreline characteristic data set, a water area characteristic data set and a peri-base environment characteristic data set; Normalizing the shoreline characteristic dataset, the water area characteristic dataset and the peri-warehouse environment characteristic dataset, and converting the normalized shoreline characteristic dataset, the water area characteristic dataset and the peri-warehouse environment characteristic dataset into standard normal distribution to obtain a normalized shoreline characteristic dataset, a normalized water area characteristic dataset and a normalized peri-warehouse environment characteristic dataset; and fusing the shoreline characteristic data set, the water area characteristic data set and the peri-base environment characteristic data set through a weighting formula to form a comprehensive characteristic data set.
4. The hydraulic engineering inspection management method according to claim 3, wherein in S3, the training method of the potential safety hazard prediction model includes: The method comprises the steps of dividing a data set into a training set, a verification set and a test set, training a model and evaluating the performance of the model, constructing a fully-connected neural network model as a potential safety hazard prediction model, wherein the potential safety hazard prediction model comprises an input layer, a hidden layer and an output layer; the neuron number of the input layer is matched with the characteristic number of the history comprehensive characteristic data set and is used for inputting the history comprehensive characteristic data set; the number of the neurons of the output layer corresponds to the number of the prediction targets, and the potential safety hazard coefficient of the hydraulic engineering is output through one neuron; The method comprises the steps of using a mean square error as a loss function, using training set data to train a potential safety hazard prediction model, minimizing the loss function through an SGD optimizer, using a verification set to evaluate the performance of the potential safety hazard prediction model and perform tuning, and using a test set to evaluate the performance of a final potential safety hazard prediction model.
5. The hydraulic engineering inspection and management method according to claim 4, wherein in S4, the method for judging whether the hydraulic engineering has the potential safety hazard comprises the following steps: If the predicted potential safety hazard coefficient of the hydraulic engineering is greater than or equal to a preset potential safety hazard coefficient threshold value of the hydraulic engineering, judging that the potential safety hazard exists in the hydraulic engineering; if the predicted potential safety hazard coefficient of the hydraulic engineering is smaller than a preset potential safety hazard coefficient threshold value, judging that the hydraulic engineering has no potential safety hazard.
6. The hydraulic engineering inspection management method according to claim 5, wherein in S5, the method for constructing the GIS-BIM three-dimensional space information model includes: Quantifying the spatial correlation between the shoreline characteristic data set and different data point positions in the peri-library environmental characteristic data set by constructing a variation function; the variation function is: ; Wherein, the Is the distance Is a function value of variation of (a); distance between data points in the shoreline characteristic data set and the surrounding environment characteristic data set; Is the space position Is a measurement of the observed value of (2); Is the space position Is a measurement of the observed value of (2); is a distance of Is the number of pairs of points; One observation point in the pair of data points; A spherical model is selected as a suitable variation function model, the spherical model being: ; Wherein, the Is a base station value; Is a change course; Estimating the value of the target point by using a known observation value through a kriging estimation formula; limiting the weight coefficient in the Kriging estimation formula through a weight coefficient adjustment model; estimating a prediction error of the kriging estimate by the kriging variance; the method comprises the steps of correlating a shoreline characteristic data set with a perikurarinary environment characteristic data set by using a kriging interpolation method, and verifying the accuracy of a kriging estimation formula by using a cross verification method; the space estimation value obtained by the Kriging estimation is recorded as The shoreline characteristic data set is as follows The characteristic data set of the surrounding environment is as follows The integrated data set is Integrating the data set by using GIS and BIM software And constructing a GIS-BIM three-dimensional space information model.
7. The hydraulic engineering inspection management and protection method according to claim 6, wherein in S5, the method for inspecting the preset inspection point through the simulation scene by linking the video monitoring data with the GIS-BIM three-dimensional space information model comprises: extracting GIS data and BIM data from the GIS-BIM three-dimensional space information model through a GIS and BIM software tool; Extracting each frame of image from the video monitoring data; using camera position information and shooting angle information in video monitoring data, and carrying out camera positioning and attitude estimation through a PnP algorithm; Matching the feature points in the video frame with the geographic feature points in the GIS data, and finding out the corresponding relation between the video frame and the GIS data through an ORB algorithm; Matching building or facility features in the video frame with corresponding features in the BIM model, and finding out the corresponding relation between the video frame and BIM data through an ICP algorithm; The video monitoring data are superimposed on a GIS-BIM three-dimensional model, and the fused data are displayed in a three-dimensional view through a rendering technology; Collecting real-time video monitoring data through a sensor, and updating a formula through a GIS-BIM three-dimensional space information model to continuously update state information and environment information in the GIS-BIM three-dimensional space information model; limiting the influence factors in the GIS-BIM three-dimensional space information model updating formula by using an influence factor limiting formula; And detecting and tracking the target from the video monitoring data by using a target detection and tracking algorithm, combining the video monitoring data and a GIS-BIM three-dimensional space information model, realizing the three-dimensional positioning of the target, and carrying out inspection on a preset inspection point.
8. The hydraulic engineering inspection management method according to claim 7, wherein in S6, the training method of the hidden danger type prediction model includes: The method comprises the steps of dividing a data set into a training set, a verification set and a test set, constructing a convolutional neural network model as a hidden danger type prediction model, wherein the hidden danger type prediction model comprises an input layer, a convolutional layer, a pooling layer, a full-connection layer and an output layer; The input layer is used for inputting potential safety hazard coefficients of historical hydraulic engineering, the output layer is used for outputting potential hazard type grades, and the output layer uses a softmax activation function; The method comprises the steps of using multi-classification cross entropy as a loss function of a model, training a hidden danger type prediction model by using a training set, updating model parameters by using a back propagation algorithm to minimize the loss function; And evaluating the performance of the model in a prediction task by using a test set, and predicting the potential safety hazard coefficient of the current hydraulic engineering by using a trained potential safety hazard type prediction model to obtain a potential safety hazard type grade.
9. The hydraulic engineering inspection management method according to claim 8, wherein in S6, the method of selecting the optimal inspection route and the optimal inspection scheme includes: classifying the hidden danger type class into a hidden danger type and a hidden danger class; marking key areas and inspection points to be inspected in the GIS-BIM three-dimensional space information model according to the hidden danger type and the hidden danger level to form an inspection point set; according to the inspection result and the hidden danger level, an inspection scheme associated with the inspection point is formulated to form an inspection scheme set; presetting a routing inspection starting point and a routing inspection terminal point in the routing inspection point set, and finding a shortest path from the starting point to the terminal point as an optimal routing inspection path through an A-algorithm; Defining a cost set required by an overhaul scheme, presetting an overhaul starting point and an end point in the cost set required by the overhaul scheme, and finding out the minimum path cost from the overhaul starting point to the overhaul end point through an A-type algorithm, wherein the corresponding overhaul scheme is the optimal overhaul scheme.
10. A hydraulic engineering inspection and management system for implementing the hydraulic engineering inspection and management method according to any one of claims 1 to 9, characterized by comprising: the data acquisition module is used for acquiring shoreline parameter data, water area parameter data, peri-base environment data and video monitoring data; The data processing module is used for preprocessing the shoreline parameter data, the water area parameter data and the peri-base environment data to obtain a shoreline characteristic data set, a water area characteristic data set and a peri-base environment characteristic data set; The potential safety hazard prediction module is used for constructing a potential safety hazard prediction model, inputting the comprehensive characteristic data set into the potential safety hazard prediction model, and predicting to obtain a potential safety hazard coefficient of the hydraulic engineering; the potential safety hazard assessment module is used for comparing the predicted potential safety hazard coefficient of the hydraulic engineering with a preset potential safety hazard coefficient threshold value and judging whether the hydraulic engineering has potential safety hazards or not; The system comprises a patrol management module, a GIS-BIM three-dimensional space information model, a video monitoring module, a storage module and a storage module, wherein if the hydraulic engineering has no hidden danger, a monitoring instruction is sent by the patrol terminal, and a worker carries out patrol on a preset patrol point through video monitoring data; The system comprises a hidden danger type prediction model, a routing inspection route planning module, a GIS-BIM three-dimensional space information model, a routing inspection route and an inspection scheme, wherein the hidden danger type prediction model is constructed, a potential safety hazard coefficient of hydraulic engineering is input into the hidden danger type prediction model to be predicted to obtain hidden danger type grades, the routing inspection route and the inspection scheme are formulated according to the hidden danger type grades in the GIS-BIM three-dimensional space information model, an optimal routing inspection route and an optimal inspection scheme are selected through an A-algorithm, and the modules are connected in a wired and/or wireless mode.

Description

Hydraulic engineering inspection management and protection method and system Technical Field The invention relates to the technical field of hydraulic engineering informatization, in particular to a hydraulic engineering inspection management and protection method and system. Background The hydraulic engineering is used as an important national infrastructure, and the safe and stable operation of the hydraulic engineering is directly related to public benefits such as flood control, disaster relief, water supply guarantee and the like. The daily inspection management and protection of the hydraulic engineering is enhanced, and the method is a key link for ensuring the long-term safe operation of the hydraulic engineering. The China patent CN117636357A discloses a hydraulic engineering information management system based on big data, which belongs to the technical field of management systems and comprises an acquisition module, a processing module, an extraction module, a detection module and a storage module, wherein the acquisition module is used for acquiring image data information of paper data, the processing module is used for carrying out standardized processing on the acquired image data information, the extraction module is used for extracting characteristics of the image data information, the detection module is used for detecting whether the surface of the paper data has a pollution condition or not, the display module is used for displaying detection results, the storage module is used for storing the data information, the characteristics of the surface of the paper data can be utilized for accurately detecting whether the surface of the paper data has a pollution problem or not, and the characteristic representation of paper pollution can be learned from a large amount of data, so that automatic pollution detection is realized, and the management quality of the information data is further improved. The traditional hydraulic engineering inspection management and protection mainly relies on manual periodic inspection and manual recording, and has the following main problems: The spatial correlation and variability between different data points cannot be effectively quantified without constructing a variation function or using a spherical model and other models correctly, the description of the spatial characteristics becomes fuzzy and inaccurate, the accuracy of subsequent analysis and application is affected, more errors are introduced when a GIS-BIM three-dimensional spatial information model is constructed due to the lack of accurate spatial characteristic description, the reliability and application value of the model are affected, the weight coefficient cannot be limited by a weight coefficient adjustment model, the observation point with stronger spatial correlation with a target point cannot be ensured to obtain larger weight in the estimation process, the interpolation result deviates from the actual condition, and the accuracy is reduced; The method is characterized in that a GIS-BIM three-dimensional space information model updating formula is not considered to be combined with real-time monitoring data and sensor data for dynamic updating, so that the state information and environment information in the model cannot reflect actual changes in time, the timeliness of the data is reduced, the virtual inspection technology is not considered to simulate inspection, problems and difficulties possibly encountered in the inspection process cannot be found in advance, the inspection accuracy and effect are reduced, and the problems of low timeliness and accuracy of the data, low inspection efficiency and accuracy, insufficient potential risk identification and evaluation capability and the like can lead to incomplete information or distortion required by decision making; The hidden danger type and hidden danger level are not subdivided, so that the inspection work lacks pertinence, inspection personnel can not accurately identify the areas and inspection points which need to be focused on, and further blind inspection is performed, the non-optimal inspection path can increase the labor, material resources and time cost in the inspection process, reduce the overall inspection efficiency, and the potential hidden danger can not be found and eliminated in time, so that the overall safety management level is reduced, and the safety of personnel is threatened. Therefore, in view of the above situation, there is an urgent need to provide a hydraulic engineering inspection management and protection method and system, so as to overcome the shortcomings in the current practical application. Disclosure of Invention The invention aims to provide a hydraulic engineering inspection management and protection method and system, which effectively solve the problems in the background technology. The invention discloses a hydraulic engineering inspection management and protection method, which comprises the follo