CN-121977512-A - Lake underwater topography measurement method

CN121977512ACN 121977512 ACN121977512 ACN 121977512ACN-121977512-A

Abstract

The invention discloses a lake underwater topography measurement method which comprises the following steps of firstly obtaining preliminary topography, namely carrying out low-altitude aerial survey by using an unmanned aerial vehicle carrying an optical camera to obtain the form of a lake water surface, the outline of a shoreline and the preliminary topography characteristics of a shallow water area, secondly dividing the lake water area into a large-area main area and a refined compensation area and generating a cooperative measurement scheme, thirdly controlling an unmanned ship to measure the large-area underwater topography according to a preset aerial line in the large-area main area, fourthly controlling the unmanned ship to measure the fine measurement and the local compensation of the underwater robot according to the preset aerial line, carrying out fine scanning by the underwater robot and carrying out positioning and fine scanning measurement by an SLAM algorithm, and fifthly, carrying out real-time processing and strategy adjustment by an AI intelligent optimization algorithm, namely dynamically adjusting the measurement paths of the unmanned ship and the underwater robot, and constructing a multi-source data fusion and high-precision topography model.

Inventors

HAO YI
LI RUIJIA
LI RONG
MAO CHEN
LIU XIAOFENG
DU YINLONG

Assignees

中国地质调查局呼和浩特自然资源综合调查中心

Dates

Publication Date: 20260505
Application Date: 20260121

Claims (8)

1. The method for measuring the underwater topography of the lake is characterized by comprising the following steps of: The method comprises the steps of firstly, obtaining preliminary topography, namely performing low-altitude aerial survey by using an unmanned aerial vehicle carrying an optical camera, generating a digital surface model by multi-view image stitching and a three-dimensional matching algorithm, and obtaining the water surface form, the shoreline outline and the preliminary topography characteristics of a shallow water area of a lake; Secondly, dividing a measurement area, namely fusing the digital surface model with historical hydrological data, inputting the fused digital surface model into a terrain feature prediction model, identifying a terrain complex area and a key monitoring area through a machine learning algorithm, dividing a lake water area into a large-area main area and a fine compensation area, and generating a cooperative measurement scheme; controlling a multi-beam depth finder and a side-scan sonar carried by the unmanned ship, measuring in the large-area main area according to a preset route and a preset route, collecting water depth and substrate type data, and generating underwater topography basic data; In the fine compensation area defined in the second step, carrying out fine scanning by the underwater robot with a high-precision sonar and a water quality sensor, and carrying out positioning and fine scanning measurement by an SLAM algorithm; Performing space-time alignment on the underwater robot data and the unmanned ship data by using an automatic mapping neural network, detecting and removing abnormal values of the aligned data, dynamically adjusting the measurement paths of the unmanned ship and the underwater robot based on a dynamic path planning algorithm and real-time topographic data, and simultaneously, predicting topographic changes by combining a topographic feature prediction model to guide the measurement strategy adjustment, wherein an optimized path instruction is fed back to the unmanned ship and the underwater robot; And step six, multi-source data fusion and high-precision terrain model construction, wherein a data fusion algorithm is adopted to integrate the digital surface model acquired by the unmanned aerial vehicle, the underwater terrain basic data acquired by the unmanned ship and the fine scanning measurement data acquired by the underwater robot, and the influence of tide level is corrected through a water level change compensation algorithm to construct a lake underwater three-dimensional terrain model.
2. The method for measuring underwater topography of a lake according to claim 1, wherein the digital surface model is used for identifying a boundary of a lake water area and a potential complex area, the potential complex area comprises a shoal and a submerged reef, input data is provided for dividing the measuring area in the second step, and meanwhile, water surface image data acquired by the unmanned aerial vehicle is used for assisting navigation and positioning of an unmanned ship.
3. The method for measuring underwater topography in a lake according to claim 1, wherein the complex topography area identified by the topography feature prediction model comprises a steep slope, a reef cluster and a vegetation coverage area, the key monitoring area comprises a hydraulic engineering influence area, the topography feature prediction model is based on a convolutional neural network, the complex nonlinear relationship between topography fluctuation, gradient change and hydrologic features is learned by training historical lake topography data, when the current digital surface model and the historical hydrologic data obtained in the first step are input, a topography complexity distribution map is output, and the lake water area is automatically divided into a large-area main area and a refined compensation area according to a preset threshold value.
4. The method for measuring underwater topography in lakes of claim 1, wherein the unmanned ship adopts a self-adaptive equal-depth line or a zigzag preset route for measurement, and the density of sounding data collected by the unmanned ship is not lower than 10 points per square meter.
5. The method for measuring the underwater topography of the lake according to claim 1, wherein the SLAM algorithm adopted by the underwater robot is a point cloud SLAM algorithm based on acoustic characteristics.
6. The method for measuring the underwater topography of the lake according to claim 1, wherein the dynamic path planning algorithm is an a-algorithm and a fast random exploration tree algorithm, the optimization targets comprise shortest path, highest measurement coverage rate and obstacle avoidance, the optimized path instructions are fed back to a control module of the unmanned ship and the underwater robot, and the tasks of the unmanned ship and the underwater robot are dynamically allocated and scheduled according to the equipment electric quantity, the task priority and the real-time hydrologic conditions.
7. The method of claim 1, wherein the data fusion algorithm comprises a kalman filter algorithm for eliminating systematic errors and random noise between different sensor data.
8. The method for measuring the underwater topography of the lake according to claim 1, wherein the water level change compensation algorithm is adopted to correct the influence of the tide level on the measured data before constructing the three-dimensional topography model of the lake.

Description

Lake underwater topography measurement method Technical Field The invention relates to the technical field of underwater topography measurement, in particular to a lake underwater topography measurement method. Background The method has irreplaceable effects on regional ecological balance, water resource allocation, flood control and disaster reduction, shipping safety and biodiversity maintenance in the aspect of taking lakes as important surface water resource carriers, realizes rapid, accurate and full-coverage measurement and dynamic monitoring of underwater topography of the lakes by integrating multi-source heterogeneous data acquisition technologies such as multi-beam sounding, side-scan sonar, underwater laser scanning, satellite remote sensing, unmanned ship/unmanned airborne sensors and the like and combining intelligent processing means such as data fusion, three-dimensional modeling, machine learning algorithm and the like, provides high-precision topography data support for ecological restoration of the lakes, water area capacity calculation, flood disaster risk assessment, shipping channel planning, fishery resource management, water quality evolution analysis and the like, further promotes optimization of health diagnosis and protection restoration strategies of ecological systems of the lakes, supports reasonable development and utilization of water resources and shipping safety guarantee in the aspect of economic significance, promotes innovation and multi-disciplinary cross application of underwater topography measurement technology in the aspect of scientific significance, and finally serves global targets of comprehensive treatment and sustainable development of the lakes. The traditional lake underwater topography measurement method has the problems of low efficiency, insufficient precision, difficult adaptation to complex water area environments and the like, the prior art depends on single equipment, and cannot realize large-area rapid measurement and local fine measurement, and especially under special environments such as plateau lakes, shallow water areas and the like, the measurement effect is not ideal. Therefore, a lake underwater topography measurement method is provided. Disclosure of Invention The invention aims to solve the defects in the prior art, and provides a lake underwater topography measuring method. In order to achieve the above purpose, the present invention adopts the following technical scheme: A lake underwater topography measurement method comprises the following steps: The method comprises the steps of firstly, obtaining preliminary topography, namely performing low-altitude aerial survey by using an unmanned aerial vehicle carrying an optical camera, generating a digital surface model by multi-view image stitching and a three-dimensional matching algorithm, and obtaining the water surface form, the shoreline outline and the preliminary topography characteristics of a shallow water area of a lake; Secondly, dividing a measurement area, namely fusing the digital surface model with historical hydrological data, inputting the fused digital surface model into a terrain feature prediction model, identifying a terrain complex area and a key monitoring area through a machine learning algorithm, dividing a lake water area into a large-area main area and a fine compensation area, and generating a cooperative measurement scheme; controlling a multi-beam depth finder and a side-scan sonar carried by the unmanned ship, measuring in the large-area main area according to a preset route and a preset route, collecting water depth and substrate type data, and generating underwater topography basic data; In the fine compensation area defined in the second step, the underwater robot carries high-precision sonar and a water quality sensor to execute fine scanning, and positioning and fine scanning measurement are performed through SLAM algorithm Performing space-time alignment on the underwater robot data and the unmanned ship data by using an intelligent automatic guided vehicle (AI) intelligent optimization algorithm, performing outlier detection and rejection on the aligned data by using an SOM neural network, dynamically adjusting the measurement paths of the unmanned ship and the underwater robot based on a dynamic path planning algorithm and real-time topographic data, and simultaneously, predicting topographic variation by combining a topographic feature prediction model to guide the measurement strategy adjustment, wherein an optimized path instruction is fed back to the unmanned ship and the underwater robot; And step six, multi-source data fusion and high-precision terrain model construction, wherein a data fusion algorithm is adopted to integrate the digital surface model acquired by the unmanned aerial vehicle, the underwater terrain basic data acquired by the unmanned ship and the fine scanning measurement data acquired by the underwater robot, and the influence of t