A 4D millimeter wave radar-based river surface flow velocity measurement method

Abstract

The application discloses a river surface flow velocity measurement method based on a 4D millimeter wave radar, belongs to the technical field of environmental monitoring and remote sensing, and obtains distance, speed, azimuth and pitch information of a river surface scattering point through a radar, and converts radial velocity into surface flow velocity along a flow direction. Through statistical flow velocity data into three-dimensional space grid voxels, a high spatial resolution river surface two-dimensional flow velocity distribution map is generated. The application realizes non-contact, high-resolution and flexible measurement of a river surface flow velocity field, is particularly suitable for river monitoring of complex terrains and hard-to-reach areas, and provides a new technical means for hydrological measurement and flood warning.

Description

Technical Field

[0001] This invention belongs to the fields of environmental monitoring, hydrological measurement and remote sensing technology, specifically a method for measuring river surface velocity based on 4D millimeter-wave radar. Background Technology

[0002] River flow monitoring is crucial for water resource management, flood warning, irrigation, and hydropower. Existing non-contact river surface velocity measurement technologies, such as handheld or fixed single-point surface velocity radar (SVR), are typically based on single-input single-output Doppler radar designs, which can only measure the velocity at a single point at the radar beam focus. To estimate flow rate from surface velocity distribution using entropy-based models, measurements need to be taken at multiple points across the river cross-section. Handheld devices require operator movement and aiming, while fixed flow meters require the installation of multiple sensors, increasing infrastructure complexity and cost.

[0003] Traditional radar methods, such as some frequency-modulated continuous wave radars, can measure both distance and velocity simultaneously, but they are usually used in conjunction with separate flow velocity measurement sensors or can only process the range-velocity map envelope, and cannot provide high spatial resolution two-dimensional flow velocity distribution maps. Summary of the Invention

[0004] This invention proposes a method for measuring river surface velocity based on 4D millimeter-wave radar.

[0005] The technical solution to achieve the objective of this invention is: a spatially resolved method for measuring river surface velocity based on 4D imaging radar, comprising the following steps:

[0006] The radar is fixed at a predetermined position above the target river area to acquire 4D point cloud data of the target river area; the first set of data is used as the baseline set, the next ten sets are used as the verification set, and eleven sets of data constitute one frame.

[0007] Simultaneously acquire radar attitude information;

[0008] The 4D point cloud data is filtered, rotation compensation is performed based on radar attitude information, and the three-dimensional coordinates of the target river area in the geodetic coordinate system are calculated by combining the position information of the radar coordinate system in which the target river area is located.

[0009] Based on the preset radar heading angle and river flow direction angle, the radial velocity of each target is converted into a surface velocity component along the river flow direction;

[0010] The three-dimensional space of the river surface is divided into grid voxels, and the transformed surface velocity components are assigned to the corresponding grid voxels according to their corresponding three-dimensional coordinates.

[0011] By setting a time window, the surface velocity components within each voxel are statistically analyzed to generate a two-dimensional spatially resolved velocity distribution map of the river surface.

[0012] Compared with the prior art, the significant advantages of this invention are: the invention deploys 4D radar at a predetermined position above the river, realizing velocity survey of the river surface; the 4D radar of this invention uses real-time dynamic positioning and attitude data from a nine-axis sensor to perform position compensation and georegistration of the radar detection target, and finally generates a georeferenced synthetic velocity image, which can reveal complex non-laminar flow patterns.

[0013] The present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0014] Figure 1 This invention provides a method for measuring river surface velocity based on 4D millimeter-wave radar.

[0015] Figure 2 This is a diagram of radar illumination.

[0016] Figure 3 This is a diagram of a time window.

[0017] Figure 4 This is the final generated flow rate map. Detailed Implementation

[0018] like Figure 1 As shown, a method for measuring river surface velocity based on 4D millimeter-wave radar interprets radar-detected targets as fundamental backscatterers on the river surface, with each target representing a velocity contribution with a specific location and radial velocity. By combining a target list from multiple radar frames and utilizing the known radar-to-river heading angle, the algorithm converts the radial velocity of the targets into a surface velocity component along the river's flow direction. Further, the algorithm divides the three-dimensional space near the river surface into a voxel grid and statistically analyzes the velocity contribution within each voxel (e.g., calculating the mean and standard deviation), thereby generating a two-dimensional velocity distribution map of the river surface. The specific steps are as follows: A 4D millimeter-wave measurement method for river surface velocity, the specific steps are as follows:

[0019] Step 1: As Figure 2 As shown, the radar is positioned at a fixed location above the target river area. The 4D millimeter-wave radar is activated, relevant configuration initialization is completed, and scanning begins to acquire 4D point cloud data of the target river area. The first set of data is used as the baseline set, the following ten sets are used as the verification set, and eleven sets of data constitute one frame.

[0020] Specifically, the 4D point cloud data includes the distance from the river to the radar, the radial velocity of the river, the azimuth angle of the river from the radar center point, and the elevation angle. In this embodiment, the azimuth angle of the scanning area is -70° to 70°, and the elevation angle is -12° to 12°.

[0021] Step 2: Obtain radar attitude information from the attitude sensor ( , , ), This is the roll angle. The pitch angle, This is the yaw angle.

[0022] Specifically, the attitude sensor is a nine-axis sensor, which is a miniature inertial measurement unit that fuses data from an accelerometer, gyroscope, and magnetometer using a complex algorithm. It can measure the three-dimensional attitude and motion of a river in real time and accurately, and is one of the core components of modern intelligent devices for motion sensing, attitude control, and navigation positioning.

[0023] Step 3: Filter the 4D point cloud data, perform rotation compensation based on radar attitude information, and calculate the target's three-dimensional coordinates in the geodetic coordinate system by combining the target river area's position information (x, y, z) in the radar coordinate system. Furthermore, while the 4D radar continuously acquires the river surface velocity point cloud data in real time, river velocity can be affected by environmental factors (such as gusts of wind and floating debris), hydrological factors (such as local eddies and minor water level fluctuations), and equipment factors (instantaneous interference with radar signals), resulting in abrupt changes in temporal values. Simultaneously, as a continuous fluid, the river's velocity exhibits stability and correlation over a short period, without any physically meaningless drastic fluctuations.

[0024] Therefore, this invention employs a time series processing scheme using a fixed-duration time window and a sliding t-test, such as... Figure 3 As shown, the core objective is to use historical time-series data as a benchmark to verify the rationality of current real-time data, identify and correct abnormal flow velocity points in the time-series dimension, and ensure the temporal continuity and physical rationality of flow velocity data.

[0025] Step 3.1: Single-frame data preprocessing. For the point cloud data collected by the radar in real time, first complete the grid division and noise point removal in the spatial dimension, and the 3σ correction and flow velocity mean assignment in the velocity dimension to obtain a single frame of effective point cloud data (1 minute / frame).

[0026] Step 3.2: Concatenate the time-series data. Concatenate all valid data from each single frame into a time-series point cloud dataset according to the acquisition time, retaining the core information of flow velocity and spatial coordinates of each frame.

[0027] Step 3.3: Perform an independent samples t-test on the mean flow velocity data of the baseline group and the test group, and calculate two core statistics—the t-statistic (reflecting the degree of difference between the means of the two groups) and the p-value (reflecting the probability of significance of the difference between the means).

[0028] Step 3.4: Determine outliers and set a significance level threshold. If the calculated p-value is less than the significance level threshold, the last frame (test group) within the time window is determined to have an outlier in the flow rate. If the p-value is greater than or equal to the threshold, the data in that frame is determined to be reasonable and is directly retained.

[0029] Step 3.5: Slide the initial time window backward by 3 to 5 frames, and repeat steps 3.3 to 3.5 until the time window covers the entire time series dataset, completing the temporal dimension verification and correction of all frames.

[0030] Step 3.6: Integrate all frame data that have been checked / corrected by time windows to obtain a valid point cloud dataset with no noise in the spatial-velocity-temporal dimensions, providing a foundation for the subsequent adaptive correction model.

[0031] Calculate the three-dimensional coordinates of the target in the geodetic coordinate system. Specifically:

[0032]

[0033]

[0034]

[0035]

[0036] In the formula, α is the angle from the x-axis of the radar rotating coordinate system to the geodetic coordinate system, β is the angle from the y-axis of the radar rotating coordinate system to the geodetic coordinate system, and γ is the angle from the z-axis of the radar rotating coordinate system to the geodetic coordinate system.

[0037] Step 4: Based on the preset radar heading angle and river flow direction angle, convert the radial velocity of each target into a surface velocity component along the river flow direction.

[0038]

[0039] Where v is the surface velocity component along the river's flow direction. For the radial velocity of the target measured by radar, and These are the compensated target elevation and azimuth angles, respectively. and These are the preset river flow direction angle and radar heading angle, respectively.

[0040] Step 5: Divide the three-dimensional space of the river surface into grid voxels, and assign the surface velocity component converted from the radial velocity measured by radar to the corresponding grid voxels according to its three-dimensional coordinates.

[0041] In a further embodiment, the three-dimensional space above and near the water surface of the river channel is divided into a series of regular three-dimensional grid voxels according to a preset resolution rule, and each voxel has a unique spatial index and spatial range.

[0042] The surface velocity components of all valid target points are matched and assigned to their respective three-dimensional voxels according to their corresponding three-dimensional spatial coordinates, thus completing the aggregation and mapping of velocity data within the spatial grid and providing a standardized data foundation for subsequent velocity visualization.

[0043] Step 6: Set a time window, perform statistical calculations on the surface velocity components within each voxel, and generate a two-dimensional spatially resolved velocity distribution map of the river surface, such as... Figure 4 As shown.

[0044] Specifically, the time window algorithm divides the real-time radar data into frames per minute, uses a sliding T-test to compare the current frame data with historical data (the previous ten frames), identifies significant changes in flow velocity, and performs weighted correction on outliers, thereby achieving dynamic change tracking and temporal filtering.

[0045] In some embodiments, the millimeter-wave radar used consists of a radar radio frequency front-end, a radar data processing board, and a communication module. The collection and processing of radar data are integrated into a host computer process, so that the data can be processed and analyzed and flow velocity measured immediately after the millimeter-wave radar has collected the data.

Claims

1. A spatially resolved method for measuring river surface velocity based on 4D imaging radar, characterized in that, Includes the following steps: The radar is fixed at a predetermined position above the target river area to acquire 4D point cloud data of the target river area; the first set of data is used as the baseline set, the next ten sets are used as the verification set, and eleven sets of data constitute one frame. Simultaneously acquire radar attitude information; The 4D point cloud data is filtered, rotation compensation is performed based on radar attitude information, and the three-dimensional coordinates of the target river area in the geodetic coordinate system are calculated by combining the position information of the radar coordinate system in which the target river area is located. Based on the preset radar heading angle and river flow direction angle, the radial velocity of each target is converted into a surface velocity component along the river flow direction; The three-dimensional space of the river surface is divided into grid voxels, and the transformed surface velocity components are assigned to the corresponding grid voxels according to their corresponding three-dimensional coordinates. By setting a time window, the surface velocity components within each voxel are statistically analyzed to generate a two-dimensional spatially resolved velocity distribution map of the river surface.

2. The spatial resolution measurement method for river surface velocity based on 4D imaging radar according to claim 1, characterized in that, The 4D imaging radar unit is a frequency-modulated continuous wave multiple-transmitter multiple-receiver radar, operating at a frequency in the 77GHz band.

3. The spatially resolved measurement method for river surface velocity based on 4D imaging radar according to claim 1, characterized in that, The specific method for filtering 4D point cloud data is as follows: Step 3.1: For the point cloud data collected by the radar in real time, complete the grid division and noise point removal in the spatial dimension, as well as the 3σ correction and flow velocity mean value assignment in the velocity dimension to obtain a single frame of effective point cloud data; Step 3.2: Concatenate the time-series data, concatenating all single-frame valid point cloud data into a time-series point cloud dataset according to the acquisition time sequence; Step 3.3: Perform an independent samples t-test on the mean flow rate data of the baseline group and the test group, and calculate the t-statistic and p-value; Step 3.4: Set a significance level threshold. If the calculated p-value is less than the significance level threshold, it is determined that there is an abnormal flow rate point in the last frame within the time window; if the p-value is greater than or equal to the threshold, the data is determined to be reasonable and is directly retained. Step 3.5: Slide the initial time window backward by 3 to 5 frames, and repeat steps 3.3 to 3.5 until the time window covers the entire time series dataset, completing the temporal dimension verification and correction of all frames; Step 3.6: Integrate all frame data that have been checked / corrected by time windows to obtain a valid point cloud dataset with no noise in the spatial-velocity-temporal dimensions.

4. The spatially resolved measurement method for river surface velocity based on 4D imaging radar according to claim 1, characterized in that, The three-dimensional coordinates of the target river area in the geodetic coordinate system ( Specifically: In the formula, ( This refers to the location information of the target river area in the radar coordinate system. In the formula, ( , , (This refers to radar attitude information.) This is the roll angle. The pitch angle, This is the yaw angle.

5. The spatially resolved measurement method for river surface velocity based on 4D imaging radar according to claim 1, characterized in that, The specific method for converting the radial velocity of each target into a surface velocity component along the river flow direction is as follows: Where v is the surface velocity component along the river's flow direction. For the radial velocity of the target measured by radar, and These are the compensated target elevation and azimuth angles, respectively. and These are the preset river flow direction angle and radar heading angle, respectively.

Images