Underwater detection robot system and method based on water conservancy facility hidden danger

Abstract

The present application relates to the technical field of water conservancy safety detection, and relates to an underwater detection robot system and method based on water conservancy hazards. According to the underwater structure diagram of the dam body and the historical hazard record, the present application generates an underwater monitoring point set, plans a robot travel detection trajectory based on the distribution position of the monitoring points, adjusts the shooting parameters in real time during the movement of the robot along the travel detection trajectory, identifies the dam body surface attachment area according to the collected underwater images, scans and detects the biological coverage and the dam body corrosion grade of the attachment area, quantifies the structure strength attenuation coefficient based on the dam body structure strength attenuation model, detects the scour defects of the non-attachment area, obtains the position and geometric size of the scour defects, analyzes the dam body structure failure probability in combination with the historical detection record of the dam body, comprehensively determines the dam body safety hazard risk, provides support for timely and effective maintenance decision-making, and guarantees the safe operation of water conservancy facilities.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of safety detection of water conservancy facilities, and relates to an underwater detection robot system and method based on water conservancy facility hidden dangers. BACKGROUND

[0002] With the increase of the operation life of water conservancy facilities, the underwater dam body structure faces multiple hidden dangers such as biological attachment, corrosion, and scouring defects, which seriously threaten the structural safety of the dam body. With the development of robot technology, underwater detection robots have been gradually applied to the detection of water conservancy facilities. However, the existing systems still have deficiencies in environmental adaptability and comprehensive risk assessment, and there is an urgent need for an intelligent detection system that can accurately identify hidden dangers and comprehensively assess risks.

[0003] The prior art, such as Chinese Patent Publication No. CN113567503B, discloses a deep water environment hydraulic structure defect detection method based on a piezoelectric sensing array. This method installs a flexible piezoelectric sensor array on an ROV robot and combines a sonar positioning system to achieve unmanned and non-destructive detection in a deep water environment. The flexible power generation layer can self-supply power, reducing the number of maintenance times and improving the service life of the equipment. It has the characteristics of high flexibility and strong sensitivity.

[0004] Chinese Patent Publication No. CN111007071A discloses an underwater inspection control method and an underwater inspection system. By receiving underwater environment images to form a work graph, planning a double-path inspection and collecting images, and merging the images to obtain a defect condition graph, the underwater inspection difficulty is reduced, and clear underwater building image data can be generated.

[0005] However, the existing technology has the following problems: 1. The existing technology does not dynamically adjust the shooting parameters according to the real-time underwater environmental light intensity and water turbidity. The underwater environment is complex and variable, and the images collected by fixed parameters are prone to be blurred and distorted, which leads to a decrease in the accuracy of subsequent defect recognition and affects the accuracy of hidden danger judgment.

[0006] 2. The existing technology focuses on the identification and positioning of defects, but does not analyze the correlation between biological attachment and dam body corrosion and dam body structure strength decay. Biological attachment and corrosion can directly weaken the carrying capacity of the dam body, and the lack of quantitative evaluation can lead to one-sided judgment of structural safety, making it impossible to early warn of insufficient strength risks.

[0007] 3. The existing technology only presents the defect situation through image stitching, and fails to effectively integrate the geometric data of scouring defects and the historical detection records of the dam body, resulting in insufficient reliability of structure failure probability evaluation, inability to accurately determine the safety hidden danger risk of the dam body, and difficulty in supporting scientific maintenance decisions, which may delay the opportunity to deal with hidden dangers. SUMMARY

[0008] In order to overcome the deficiencies of the prior art, the application provides a water conservancy facility hidden danger underwater detection robot system and method, which realizes efficient and accurate detection and risk assessment of dam underwater structure hidden dangers through multi-dimensional information fusion and intelligent analysis.

[0009] The technical solution adopted by the application to solve its technical problems is: a water conservancy facility hidden danger underwater detection robot system, comprising a detection trajectory planning module, an underwater image acquisition module, an attachment area identification module, a structure strength evaluation module, an erosion defect analysis module and a risk evaluation module.

[0010] The connection relationship between the modules is: the detection trajectory planning module is in communication connection with the underwater image acquisition module, the attachment area identification module is in communication connection with the underwater image acquisition module, the structure strength evaluation module is in communication connection with the attachment area identification module, the erosion defect analysis module is in communication connection with the attachment area identification module, and the risk evaluation module is in communication connection with the structure strength evaluation module and the erosion defect analysis module respectively.

[0011] The detection trajectory planning module forms an underwater monitoring point set on the dam underwater surface according to the dam underwater structure diagram and the historical hidden danger record, and plans the robot's advancing detection trajectory based on the distribution position of the underwater monitoring point set;

[0012] The underwater image acquisition module acquires images of the dam underwater surface according to the adjusted shooting parameters during the movement of the robot along the advancing detection trajectory, and forms an underwater image sequence;

[0013] The attachment area identification module presents the attachment area in the surface attachment state through feature recognition of the underwater image sequence, and detects the biological coverage rate and the dam corrosion grade through scanning of the robot on the attachment area;

[0014] The structure strength evaluation module generates a structure strength attenuation coefficient according to the biological coverage rate and the dam corrosion grade, combined with the dam material mechanics characteristics;

[0015] The erosion defect analysis module detects the edge of the erosion defect in the remaining area other than the attachment area through the robot, and obtains the center position and geometric size data of the erosion defect in the remaining area;

[0016] The risk evaluation module analyzes the probability of structural failure of the dam based on the center position and geometric size data of the erosion defect, combined with the historical detection record of the dam, and determines the dam safety hidden danger risk together with the structure strength attenuation coefficient.

[0017] On the other hand, the application provides a water conservancy facility hidden danger underwater detection method, comprising: forming an underwater monitoring point set on the dam underwater surface according to the dam underwater structure diagram and the historical hidden danger record, and planning the robot's advancing detection trajectory based on the distribution position of the underwater monitoring point set.

[0018] During the movement of the robot along the detection track, the underwater surface of the dam body is imaged according to the adjusted shooting parameters to form an underwater image sequence.

[0019] The underwater image sequence is subjected to feature recognition to present an attached area of an attached state, and the robot is used to scan and detect biological coverage and dam body corrosion grade of the attached area.

[0020] According to the biological coverage and the dam body corrosion grade, a structure strength attenuation coefficient is generated in combination with the mechanical properties of the dam body material.

[0021] The robot is used to detect the edge of the scour defect in the remaining area other than the attached area to obtain the center position and geometric size data of the scour defect in the remaining area.

[0022] Based on the center position and geometric size data of the scour defect, the dam body structure failure probability is analyzed in combination with the historical detection records of the dam body, and the dam body safety hidden danger risk is determined in combination with the structure strength attenuation coefficient.

[0023] Compared with the prior art, the present application has the following beneficial effects: (1) the present application generates an underwater monitoring point set according to the underwater structure diagram of the dam body and the historical hidden danger records, plans a detection track of the robot based on the distribution position of the monitoring points, plans the track by integrating the initial monitoring points and the marked monitoring points, realizes the key coverage of the historical hidden danger area, and thus improves the comprehensiveness and accuracy of the detection and reduces the omission of high-risk areas.

[0024] (2) The present application dynamically adjusts the shooting parameters by real-time acquisition of underwater environmental light intensity and water turbidity, images the underwater surface of the dam body according to the adjusted shooting parameters, solves the problem of poor image quality caused by dynamic underwater environment, ensures the clarity of the underwater image sequence, provides a high-quality data basis for subsequent attached area recognition and defect analysis, and improves the accuracy of hidden danger identification.

[0025] (3) The present application improves the calculation accuracy of the structure strength attenuation coefficient by feature recognition of the attached area of the underwater image sequence, scanning and detecting the biological coverage and the dam body corrosion grade of the attached area, and quantifying the structure strength attenuation coefficient based on the dam body structure strength attenuation model, realizes scientific evaluation of the actual bearing capacity of the dam body, and thus can early warn the strength deficiency risk and provide a quantitative basis for maintenance and reinforcement.

[0026] (4) The application integrates the center position and geometric size data of the scour defect and the historical detection record of the dam body, analyzes the structural failure probability of the dam body, and determines the dam body safety hidden danger risk in combination with the structural strength attenuation coefficient, realizes comprehensive quantitative evaluation on the safety state of the dam body, can accurately judge the hidden danger severity, provides support for timely and effective maintenance decision, and guarantees the operation safety of water conservancy facilities. BRIEF DESCRIPTION OF DRAWINGS

[0027] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the description of the embodiments. Obviously, the drawings in the following description only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0028] Figure 1 It is a schematic diagram of system module connection of the present application.

[0029] Figure 2 It is a schematic diagram of setting process of underwater image shooting parameter adjustment mapping table in the present application.

[0030] Figure 3 It is a schematic diagram of specific process flow of forming multiple groups of categories in the present application.

[0031] Figure 4 It is a schematic diagram of method step flow of the present application. DETAILED DESCRIPTION

[0032] Various exemplary embodiments of the present application will now be described in detail with reference to the accompanying drawings. Note that the relative arrangement, numerical expressions, and values of components and steps set forth in these embodiments are not limiting to the scope of the present application unless otherwise specifically stated. Also, it should be understood that the dimensions of the various parts shown in the drawings are not drawn to scale for the sake of convenience of description.

[0033] The following description of at least one example embodiment is merely illustrative in nature and is in no way limiting to the scope of the application and its applications or uses. Techniques, methods, and devices known to those of ordinary skill in the relevant art can not be discussed in detail, but should be considered part of the specification, where appropriate.

[0034] In all examples shown and discussed herein, any specific values should be interpreted as merely illustrative and not as a limitation. Thus, other examples of the example embodiments can have different values.

[0035] Reference should be made to Figure 1As shown, the application provides a water detection robot system based on water conservancy hidden dangers, which comprises a detection trajectory planning module, an underwater image acquisition module, an attachment area identification module, a structural strength evaluation module, a scour defect analysis module and a risk assessment module.

[0036] The connection relationship between each module is: the detection trajectory planning module is in communication connection with the underwater image acquisition module, the attachment area identification module is in communication connection with the underwater image acquisition module, the structural strength evaluation module is in communication connection with the attachment area identification module, the scour defect analysis module is in communication connection with the attachment area identification module, and the risk assessment module is in communication connection with the structural strength evaluation module and the scour defect analysis module respectively.

[0037] The detection trajectory planning module forms a set of underwater monitoring points on the surface of the dam body based on the dam body underwater structure diagram and the historical hidden danger record, and plans the robot's advancing detection trajectory based on the distribution position of the set of underwater monitoring points.

[0038] It should be noted that the specific content of the detection trajectory planning module is as follows: the dam body underwater structure diagram is obtained from the dam body standard design structure diagram according to the current water level of the dam body, and the surface area of the dam body underwater structure diagram is divided into a plurality of initial monitoring points according to the preset meshing.

[0039] The hidden danger area positions corresponding to each hidden danger record are retrieved from the dam body historical detection record, and the different hidden danger area positions are marked on the surface area of the dam body underwater structure diagram as a plurality of marked monitoring points.

[0040] The plurality of initial monitoring points and the plurality of marked monitoring points are integrated to form a set of underwater monitoring points.

[0041] The advancing detection trajectory connecting all the monitoring points in the set of underwater monitoring points is generated according to the distribution position of all the monitoring points in the set of underwater monitoring points.

[0042] In a specific embodiment, the current water level of the dam body is obtained in real time by a water level sensor, and the maximum inscribed cube area of a single grid in the preset meshing is determined by the shooting range of the underwater detection robot from the dam body surface area set distance.

[0043] The initial monitoring points and the marked monitoring points are both position points from the dam body surface area set distance, and the initial monitoring points can be the center points of the single grid.

[0044] The hidden danger area corresponding to each hidden danger record includes but is not limited to the area of cracks, corrosion, biological attachment or scour, and the hidden danger area position is the center position of the hidden danger area.

[0045] The optimal travel detection track is the closest travel detection track selected from different travel detection tracks planned according to the distribution positions of all monitoring points in the underwater monitoring point set.

[0046] The application generates an underwater monitoring point set according to a dam underwater structure diagram and historical hidden danger records, plans a robot travel detection track based on the distribution positions of the monitoring points, and plans a track by integrating the initial monitoring points and the marked monitoring points to achieve key coverage of the historical hidden danger area, thereby improving the comprehensiveness and accuracy of detection and reducing the omission of high-risk areas.

[0047] The underwater image acquisition module acquires images of the underwater surface of the dam according to the adjusted shooting parameters during the movement of the robot along the travel detection track, and forms an underwater image sequence.

[0048] It should be noted that if the robot encounters a sudden obstacle during movement along the travel detection track, the robot will automatically bypass the sudden obstacle and re-plan a shortest path to connect the subsequent travel detection track to ensure the continuity of detection.

[0049] In a specific embodiment, the adjustment method of the shooting parameters is as follows: when the robot moves to an underwater monitoring point along the travel detection track, the underwater environmental illumination intensity and water turbidity of the underwater monitoring point are acquired in real time by the robot, and the underwater environmental illumination intensity and water turbidity are substituted into a set underwater image shooting parameter adjustment mapping table to obtain the exposure time and white balance parameters during image shooting.

[0050] As shown in Figure 2 A1, historical shooting data is retrieved from the underwater image shooting history record, and the underwater environmental illumination intensity and water turbidity in the historical shooting data are clustered to form multiple groups of categories.

[0051] A2, historical shooting data with an image definition higher than a set image definition threshold is screened from the multiple groups of categories, and the optimal values of the exposure time and white balance parameters during image shooting are counted therefrom. The optimal values can be median values, which have stronger anti-exceptional value capability.

[0052] A3, a data interval is formed based on the maximum and minimum values of the underwater environmental illumination intensity and water turbidity in the multiple groups of categories, and an adjustment mapping table of the underwater environmental illumination intensity interval and water turbidity interval and the exposure time and white balance parameters is established.

[0053] As shown in Figure 3 The specific process of step A1 includes: A11, the underwater environmental illumination intensity and water turbidity in the multiple historical shooting data are standardized, and K centroids are randomly initialized after the standardized environmental illumination intensity and water turbidity.

[0054] A12, calculate the Euclidean distance between the multiple historical shooting data and the K centroids, distribute the multiple historical shooting data to the cluster where the nearest centroid is located, and form K temporary clusters.

[0055] A13, recalculate the mean of all historical shooting data in each temporary cluster, and take it as the centroid of the temporary cluster.

[0056] A14, repeat steps A12-A13 until the centroid position is stable, and finally form K stable clusters, that is, multiple groups of categories.

[0057] The application adjusts the shooting parameters dynamically according to the real-time collection of underwater environmental light intensity and water turbidity, collects the image of the underwater surface of the dam according to the adjusted shooting parameters, solves the problem of poor image quality caused by dynamic underwater environment, ensures the clarity of the underwater image sequence, provides high-quality data basis for subsequent attachment area recognition and defect analysis, and improves the accuracy of hidden danger identification.

[0058] The attachment area recognition module performs feature recognition on the underwater image sequence to present the attachment area in the surface attachment state, detects the biological coverage and dam corrosion grade of the attachment area through the robot, thereby improving the recognition accuracy and significantly improving the calculation accuracy of the structural strength attenuation coefficient.

[0059] It should be noted that the specific content of the attachment area recognition module is as follows: the underwater complete image of the dam is obtained by splicing and integrating the underwater image sequence, and the underwater gray image of the dam is obtained by performing gray processing on the underwater complete image of the dam.

[0060] The pixel separation of the attachment area in the surface attachment state in the underwater gray image of the dam is performed through an image segmentation algorithm, the contour boundary area of the attachment area is determined, and the coverage is obtained by comparing the contour boundary area with the total area of the underwater gray image of the dam. The image splicing and integration method and the image segmentation algorithm are both prior art, and the application will not be described in detail.

[0061] A high-resolution ultrasonic scanner is used to scan the surface of the attachment area point by point, three-dimensional coordinate information of all points in the surface of the attachment area is obtained, a micro three-dimensional model of the surface of the attachment area is formed, and a region with a height lower than the surrounding surface in the model is determined as a corrosion pit region.

[0062] The contour extraction of all corrosion pit regions in the attachment area is performed to determine the corrosion pit feature parameters, the corrosion pit feature parameters are compared with the corresponding feature parameter ranges of each corrosion grade in the preset corrosion grade evaluation standard library, the matching degree is analyzed through weighted scoring, and the corrosion grade with the highest matching degree is taken as the dam corrosion grade of the attachment area.

[0063] In a specific embodiment, the corrosion pit characteristic parameters include, but are not limited to, the number of corrosion pits, the maximum depth, the distribution density, and the average diameter. The number of corrosion pits is the total number of independent corrosion pits, the maximum depth is the maximum height difference between the lowest point in the corrosion pit and the surrounding normal surface, the distribution density is the number of corrosion pits per unit area, and the average diameter is the average value of the equivalent circle diameters of all corrosion pit boundary profiles.

[0064] The matching degree analysis method is as follows: if a certain corrosion pit characteristic parameter is within the range of the characteristic parameters corresponding to a certain corrosion grade, the matching value of the corrosion resistance characteristic parameter and the corrosion grade is recorded as 1, otherwise the matching value of the corrosion resistance characteristic parameter and the corrosion grade is recorded as 0. The matching values of all corrosion pit characteristic parameters and each corrosion grade are summed to obtain the matching degree of the corrosion pit characteristic parameters and each corrosion grade.

[0065] It needs to be explained that the weights of the number of corrosion pits, the maximum depth, the distribution density, and the average diameter can be set according to industry experience or obtained through a limited number of test data. For example, a large number of dam body samples with known corrosion grades are first collected, the corrosion pit characteristic parameters of each sample are extracted, each corrosion pit characteristic parameter of each sample is normalized, the information entropy of each corrosion pit characteristic parameter is calculated based on the normalized each corrosion pit characteristic parameter, the information entropy difference coefficient is evaluated, and the proportion of the information entropy difference coefficient is converted into the weights of the number of corrosion pits, the maximum depth, the distribution density, and the average diameter, and the total weight is 1.

[0066] The structural strength evaluation module generates a structural strength decay coefficient based on the biological coverage and the dam body corrosion grade, in combination with the dam body material mechanical properties.

[0067] It needs to be explained that the structural strength decay coefficient is generated as follows: the dam body corrosion grade of the attachment area is used to query a predefined corrosion decay ratio table to obtain a basic decay coefficient corresponding to the dam body corrosion grade, the basic decay coefficient corresponding to the dam body corrosion grade is substituted into an established dam body structural strength decay model with the biological coverage to output the decayed dam body structural strength. The dam body structural strength can be the compressive strength.

[0068] The initial structural strength corresponding to the dam body material mechanical properties is selected from a material property parameter library, and the structural strength decay coefficient is obtained based on the deviation ratio analysis of the decayed dam body structural strength and the initial structural strength.

[0069] The structural strength decay coefficient is the ratio of the deviation value of the initial structural strength and the decayed dam body structural strength to the initial structural strength.

[0070] In a specific implementation, the predefined corrosion decay ratio table can simulate different corrosion levels by performing accelerated corrosion tests on commonly used materials of the dam body, determine the structural strength of the materials under different corrosion levels, and calculate the ratio of the structural strength to the initial structural strength of the materials as the basic decay coefficient of each corrosion level.

[0071] The dam structure strength decay model is established in the following manner: collecting biological coverage, basic decay coefficients corresponding to dam corrosion levels, and dam structure strength of multiple groups of historical records in dam structure strength decay detection data to form a data set.

[0072] The data set is divided into a training set and a test set according to a set proportion, the biological coverage and the basic decay coefficients in the training set are used as input features, and the dam structure strength is used as a target variable to be substituted into a set linear regression equation for training. The regression coefficients and constant terms of the biological coverage and the basic decay coefficients are obtained by least squares fitting, and an initial dam structure strength decay model is constructed.

[0073] The dam structure strength of the test set is predicted by the initial dam structure strength decay model, and the initial dam structure strength decay model is adjusted and optimized using accuracy and mean square error, and the final dam structure strength decay model is output.

[0074] The accuracy indicates the proportion of samples whose deviation between the predicted dam structure strength and the actual dam structure strength is within the allowable range, and the mean square error indicates the overall deviation degree of the predicted dam structure strength and the actual dam structure strength.

[0075] The adjustment and optimization of the initial dam structure strength decay model can be adjusting the regression coefficients and constant terms of the biological coverage and the basic decay coefficients when the accuracy is less than the set accuracy or the mean square error is greater than the set mean square error, reevaluating the accuracy and the mean square error of the initial dam structure strength decay model, and outputting the final dam structure strength decay model when the accuracy is greater than the set accuracy and the mean square error is less than the set mean square error.

[0076] The present application significantly improves the calculation accuracy of the structural strength decay coefficient by identifying the attachment area of the underwater image sequence, scanning and detecting the biological coverage and the dam corrosion level of the attachment area, and quantifying the structural strength decay coefficient based on the dam structure strength decay model. The scientific evaluation of the actual carrying capacity of the dam is realized, so as to provide a quantitative basis for maintenance and reinforcement by early warning of the risk of insufficient strength.

[0077] The scour defect analysis module detects the center position and geometric size data of the scour defects in the remaining area outside the attachment area by the robot. The geometric size data includes maximum width, extension length, area, and perimeter, etc.

[0078] The risk assessment module analyzes the probability of structural failure of the dam based on the center position and geometric size data of the scour defect, in combination with the historical detection record of the dam body, and determines the dam safety risk by combining the structural strength attenuation coefficient.

[0079] It should be noted that the analysis process of the probability of structural failure of the dam is as follows: the geometric size data of the scour defect in the remaining area is integrated to establish a comprehensive evaluation data set.

[0080] The same center position of the scour defect is retrieved from the historical detection record of the dam, and the geometric size data of the scour defect corresponding to each reference record is established to form a reference evaluation data set.

[0081] The comprehensive evaluation data set and the reference evaluation data set of each reference record are compared in similarity, the reference records with a similarity higher than a set similarity threshold are screened out, the number of reference records of the dam structure hidden danger is counted, and the proportion of the number of reference records in the total number of screened reference records is recorded as the probability of structural failure of the dam.

[0082] In a specific embodiment, the safety risk judgment rule of the dam is: when the structural strength attenuation coefficient of the attached area is greater than a set structural strength attenuation coefficient threshold or the probability of structural failure of the dam is greater than a set failure probability threshold, it is determined that the dam has a safety risk, and a warning signal is triggered by the robot.

[0083] The present application integrates the center position and geometric size data of the scour defect and the historical detection record of the dam, analyzes the probability of structural failure of the dam, and determines the safety risk of the dam by combining the structural strength attenuation coefficient, realizes the comprehensive quantitative evaluation of the safety state of the dam, and can accurately judge the severity of the hidden danger, provides support for timely and effective maintenance decision, and ensures the safe operation of water conservancy facilities.

[0084] In another aspect as shown in Figure 4 The present application provides an underwater detection method based on water conservancy facility hidden danger, which comprises: forming an underwater monitoring point set on the underwater surface of the dam according to the underwater structure diagram of the dam and the historical hidden danger record, and planning the detection trajectory of the robot based on the distribution position of the underwater monitoring point set.

[0085] During the movement of the robot along the detection trajectory, the underwater surface of the dam is imaged according to the adjusted imaging parameters to form an underwater image sequence.

[0086] The underwater image sequence is subjected to feature recognition to present the attached area of the surface attachment state, and the biological coverage and dam corrosion grade of the attached area are scanned and detected by the robot.

[0087] According to the biological coverage and the dam body corrosion grade, the structure strength attenuation coefficient is generated in combination with the mechanical properties of the dam body material.

[0088] The remaining area outside the attachment area is subjected to edge detection of the scour defect by the robot, and the center position and geometric size data of the scour defect in the remaining area are obtained.

[0089] Based on the center position and geometric size data of the scour defect, the dam body structural failure probability is analyzed in combination with the dam body historical detection record, and the dam body safety hidden danger risk is determined in combination with the structure strength attenuation coefficient.

[0090] The above formulas are all dimensionless numerical calculations, the formulas are obtained by software simulation of a large amount of data to obtain the latest real situation, and the preset parameters in the formulas are set by the person skilled in the art according to the actual situation.

[0091] The above embodiments can be realized wholly or partially by software, hardware, firmware or any combination thereof. When realized by software, the above embodiments can be realized wholly or partially in the form of a computer program product.

[0092] Those skilled in the art can realize that the modules and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. The person skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0093] In addition, each functional module in each embodiment of the present application can be integrated in one processing module, or each module can exist physically alone, or two or more modules can be integrated in one module.

[0094] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0095] Finally, the above is only a preferred embodiment of the present application and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A water detection robot system based on water conservancy hidden danger, characterized in that, The method comprises the following steps: a detection trajectory planning module forms an underwater monitoring point set on the underwater surface of the dam body according to the underwater structure diagram of the dam body and historical hidden danger records, and plans a detection trajectory of the robot based on the distribution position of the underwater monitoring point set; an underwater image acquisition module acquires images of the underwater surface of the dam body according to the adjusted shooting parameters during the movement of the robot along the detection trajectory, and forms an underwater image sequence; an attached area identification module identifies the attached area in the underwater image sequence to present the surface attachment state, and detects the biological coverage and the dam body corrosion grade of the attached area by the robot; a structure strength evaluation module generates a structure strength attenuation coefficient according to the biological coverage and the dam body corrosion grade and in combination with the mechanical properties of the dam body material; a scour defect analysis module detects the edge of the scour defect in the remaining area other than the attached area by the robot, and obtains the center position and geometric size data of the scour defect in the remaining area; a risk evaluation module analyzes the probability of structural failure of the dam body based on the center position and geometric size data of the scour defect in combination with the historical detection records of the dam body, and determines the safety hidden danger risk of the dam body in combination with the structure strength attenuation coefficient; the generation mode of the structure strength attenuation coefficient is that a basic attenuation coefficient corresponding to the dam body corrosion grade is obtained by querying a predefined corrosion attenuation proportion table according to the dam body corrosion grade of the attached area, the basic attenuation coefficient corresponding to the dam body corrosion grade and the biological coverage are substituted into an established dam body structure strength attenuation model, and the attenuated dam body structure strength is output; an initial structure strength corresponding to the mechanical properties of the dam body material is selected from a material property parameter library, and a structure strength attenuation coefficient is obtained by analyzing the deviation ratio between the attenuated dam body structure strength and the initial structure strength; the establishment mode of the dam body structure strength attenuation model is that biological coverage, basic attenuation coefficients corresponding to dam body corrosion grades, and dam body structure strengths of multiple historical records in dam body structure strength attenuation detection data are collected to form a data set; the data set is divided into a training set and a test set according to a set proportion, the biological coverage and the basic attenuation coefficients in the training set are taken as input features, the dam body structure strength is taken as a target variable, and a set linear regression equation is substituted for training, the regression coefficients of the biological coverage and the basic attenuation coefficients and the constant term are solved by least squares fitting, an initial dam body structure strength attenuation model is constructed; the dam body structure strength of the test set is predicted by the initial dam body structure strength attenuation model, the initial dam body structure strength attenuation model is adjusted and optimized by using the accuracy and the mean square error, and a final dam body structure strength attenuation model is output. The dam structure failure probability analysis process is as follows: the geometric size data of the remaining area erosion defects are integrated to establish a comprehensive evaluation data set; the same center position of the erosion defect is called from the dam historical detection record, and the geometric size data of each reference record corresponding to the erosion defect is established to form a reference evaluation data set; the comprehensive evaluation data set and the reference evaluation data set of each reference record are compared in similarity, the reference records with similarity higher than the set similarity threshold are screened out, the number of dam structure hidden danger reference records is counted, and the proportion of the number in the total number of screened reference records is recorded as the dam structure failure probability.

2. The underwater inspection robot system based on water conservancy hidden danger according to claim 1, characterized in that: The specific content of the detection trajectory planning module is as follows: According to the current water level of the dam, the underwater structure diagram of the dam is obtained from the standard design structure diagram of the dam, and the surface area of the underwater structure diagram of the dam is divided into a plurality of initial monitoring points according to the preset meshing; The position of each hidden danger area corresponding to the hidden danger record is called from the dam historical detection record, and the different hidden danger area positions are marked on the surface area of the underwater structure diagram of the dam, which are used as a plurality of marked monitoring points; The initial monitoring points and the marked monitoring points are integrated to form an underwater monitoring point set; According to the distribution position of all monitoring points in the underwater monitoring point set, the best travel detection trajectory connecting all monitoring points is generated.

3. The underwater inspection robot system based on water conservancy hidden danger according to claim 1, characterized in that: The adjustment mode of the shooting parameter is as follows: When the robot moves to the underwater monitoring point along the travel detection trajectory, the underwater environment light intensity and water turbidity of the underwater monitoring point are collected in real time by the robot, and the underwater environment light intensity and water turbidity are substituted into the set underwater image shooting parameter adjustment mapping table to obtain the exposure time and white balance parameter during image shooting.

4. The underwater detection robot system based on water conservancy hidden danger according to claim 3, characterized in that: The setting mode of the underwater image shooting parameter adjustment mapping table is as follows: The underwater environment light intensity and water turbidity in the multiple historical shooting data are clustered to form multiple groups of categories; The historical shooting data with image definition higher than the set image definition threshold in the multiple groups of categories are screened out, and the optimal values of the exposure time and white balance parameter during image shooting are counted; Based on the maximum and minimum values of the underwater environment light intensity and water turbidity in the multiple groups of categories, a data interval is formed, and an adjustment mapping table of the underwater environment light intensity interval and water turbidity interval and the exposure time and white balance parameter is established.

5. The underwater inspection robot system based on water conservancy hidden danger according to claim 1, characterized in that: The specific content of the attachment area identification module is as follows: The underwater image sequence is spliced and integrated to obtain a complete underwater image of the dam, and the complete underwater image of the dam is subjected to gray processing to obtain a dam underwater gray image; The attachment area in the dam underwater gray image is subjected to pixel separation by an image segmentation algorithm, the contour boundary area of the attachment area is determined, and the coverage is obtained by comparing the contour boundary area with the total area of the dam underwater gray image; The surface of the attachment area is scanned point by point, the three-dimensional coordinate information of all points in the surface of the attachment area is obtained, a micro three-dimensional model of the surface of the attachment area is formed, and the area lower than the surrounding surface by a set height in the model is determined as a corrosion pit area; The profile extraction is performed on all corrosion pit areas in the attached area to determine the corrosion pit characteristic parameters, the corrosion pit characteristic parameters are compared with the corresponding characteristic parameter ranges of each corrosion grade in the preset corrosion grade evaluation standard library, the matching degree is analyzed through weighted scoring, and the corrosion grade with the highest matching degree is taken as the dam body corrosion grade of the attached area.

6. The underwater inspection robot system based on water conservancy hidden danger according to claim 1, characterized in that: The safety hidden danger risk judgment rule of the dam body is: When the structure strength attenuation coefficient of the attached area is greater than the set structure strength attenuation coefficient threshold or the structure failure probability of the dam body is greater than the set failure probability threshold, it is determined that the dam body has a safety hidden danger risk, and a warning signal is sent by the robot.

7. The method of underwater detection of water facilities hazards, by the underwater detection robot system of water facilities hazards based on the steps of claims 1-6, characterized in that, It comprises: An underwater monitoring point set is formed on the underwater surface of the dam body according to the underwater structure diagram of the dam body and the historical hidden danger record, and the detection trajectory of the robot is planned based on the distribution position of the underwater monitoring point set; During the movement of the robot along the detection trajectory, the underwater surface of the dam body is imaged according to the adjusted imaging parameters to form an underwater image sequence; The feature recognition is performed on the underwater image sequence to present the attached area in the surface attached state, the biological coverage and the dam body corrosion grade are detected by the robot through scanning the attached area; The structure strength attenuation coefficient is generated according to the biological coverage and the dam body corrosion grade in combination with the mechanical properties of the dam body material; The edge detection of the scour defect is performed on the remaining area outside the attached area by the robot to obtain the center position and geometric size data of the scour defect in the remaining area; Based on the center position and geometric size data of the scour defect, the structure failure probability of the dam body is analyzed in combination with the historical detection record of the dam body, and the safety hidden danger risk of the dam body is determined in combination with the structure strength attenuation coefficient.

Images