A method for monitoring the presence of personnel on two-wheeled vehicles based on millimeter-wave radar

Abstract

The present invention discloses a method for monitoring personnel on-site of a two-wheeled vehicle based on millimeter-wave radar, which belongs to the field of two-wheeled vehicle monitoring technology, and comprises the following steps: radar is started, and target information is output after 1D-FFT processing, 2D-FFT processing, CFAR and angle estimation, clustering, and α-β filtering processing; pre-processing, when the number of targets is 0, the monitoring state is directly updated to an unmanned state; when the target is not 0, the position points (x, y) of all targets are obtained by calculation; after caching multiple frames of data, multiple frames of accumulated data are processed, and the effective area of ​​detection is divided into 6, and the monitoring state is updated after the target is accumulated; the updated state is output, the target data is cleared, and the judgment of the next cycle is executed. The present invention can specifically judge the specific position of the driver: sitting, sitting at the back side, standing in front of the seat, the left side of the vehicle, the right side of the vehicle, the personnel leaving the vehicle, and the unmanned state.

Description

Technical Field

[0001] The present invention relates to the technical field of two-wheeled vehicle monitoring, and in particular to a method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter wave radar. Background Art

[0002] Electric two-wheeled vehicles include electric vehicles and electric assisted vehicles. As electric two-wheeled vehicles develop towards intelligence, there is a strong demand for driver presence detection and personnel location monitoring functions.

[0003] Millimeter-wave radar is a commonly used sensor for human body detection, which can monitor the position of the human body in a non-contact manner.

[0004] However, there are relatively few technologies for monitoring the seat occupancy of two-wheeled vehicles. The driver's body perception mainly relies on the pressure sensor in the seat cushion, which has the problem of single function and serious false alarms, and it is difficult to achieve non-contact detection, and it is impossible to accurately describe the driver's position and seat status.

[0005] Based on this, the present invention designs a method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter-wave radar to solve the above problems. Summary of the invention

[0006] In view of the above-mentioned shortcomings of the prior art, the present invention provides a method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter-wave radar.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0008] A method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter-wave radar comprises the following steps:

[0009] 1. The radar is started, and a frequency-modulated continuous pulse signal is sent through the TX antenna, and the reflected echo signal of the target is received through multiple radar receiving antennas RX. After 1D-FFT processing, 2D-FFT processing, CFAR and angle estimation, the point cloud data of the target is obtained;

[0010] According to the point cloud data, a density-based clustering method is used to obtain the target data, and the target data is processed by α-β filtering to output the target information of this frame;

[0011] Second, pre-process the target data of the frame, remove invalid data, and count the number of targets. When the number of targets is 0, directly update the monitoring status to unmanned status;

[0012] 3. When the target is not 0, the position points (x, y) of all targets are calculated based on the range and angle of all targets in this frame;

[0013] Among them, "range" represents the range, and "angle" represents the angle;

[0014] When the number of targets is 1, the only target in this frame is cached as the core target. When the number of targets > 1, one core target is selected for storage;

[0015] IV. After caching multiple frames of data, process the accumulated data of multiple frames. Divide the detected effective area into the seating area, the rear seating area, the standing area inside the vehicle, the getting-off area, the right area, and the left area. Update the monitoring status after the targets are accumulated;

[0016] V. Output the updated status, clear all target data cached in this judgment, and execute the judgment for the next cycle.

[0017] Furthermore, the target data includes the distance, angle, speed, and moving / static information of the target.

[0018] Furthermore, the calculation method of the position point (x, y) is as follows:

[0019] x = range * sin(angle);

[0020] y = range * cos(angle).

[0021] Furthermore, the screening method in step 3 is as follows:

[0022] 3.1. Judge whether the target is within the seating area of the two-wheeler;

[0023] 3.2. When there is no target within the seating area, calculate the distances from all targets to the center point target_center(x0, y0) of the two-wheeler, and cache the point with the closest distance as the core target.

[0024] Furthermore, in step 4, based on all the core target points (x_val, y_val) cached, judge which effective area each target is in;

[0025] When y1 < y_val < y2 and x1 <= x_val <= x2, judge that this core target is in the seating area, where y1 and y2 are the upper and lower boundaries in the y direction of this area respectively, and x1 and x2 are the left and right boundaries in the x direction respectively;

[0026] When y0 < y_val < y1 and x1 <= x_val <= x2, judge that this core target is in the rear seating area, where y0 and y1 are the upper and lower boundaries in the y direction of this area respectively, and x1 and x2 are the left and right boundaries in the x direction respectively;

[0027] When y2 < y_val < y3 and x1 <= x_val <= x2, it is determined that the core target is the standing area inside the vehicle, where y2 and y3 are the upper and lower boundaries of the area in the y direction respectively, and x1 and x2 are the left and right boundaries in the x direction;

[0028] When y3 < y_val, it is determined that the core target is the area away from the vehicle, where y3 is the lower boundary of the area in the y direction;

[0029] When 0 < y_val < y3 and x_val >= x2, it is determined that the core target is the right area, where 0 and y3 are the upper and lower boundaries of the area in the y direction respectively, and x2 is the left boundary in the x direction;

[0030] When 0 < y_val < y3 and x_val <= x1, it is determined that the core target is the left area, where 0 and y3 are the upper and lower boundaries of the area in the y direction respectively, and x1 is the right boundary in the x direction.

[0031] Furthermore, in step four, the targets appearing in each area are accumulated. When the number of core targets in an area is greater than the threshold value V, the monitoring status is updated to the status represented by the area where it is located. If the number of points in no area is greater than the threshold value, the status is updated to the motion state.

[0032] Furthermore, the threshold value V = 6.

[0033] Furthermore, in step 3.1, the constraint conditions are:

[0034] 0 < y <= y_centaland, x_cental_left <= x <= x_cental_right;

[0035] Among them, y_centaland is the seating boundary value in the y direction, x_cental_left is the minimum seating boundary value in the x direction, and x_cental_right is the maximum seating boundary value in the x direction. When there is a target in the seating area, the target is cached as the core target point and the screening ends.

[0036] Furthermore, the processing process in step one includes: 1D-FFT processing, 2D-FFT processing, CFAR and angle estimation.

[0037] Beneficial effects

[0038] The present invention can realize the detection of the driver's body presence of a two-wheeled vehicle through a millimeter-wave radar; it can detect the presence of people around the vehicle, determine whether there are people around the electric vehicle, whether there are multiple people, whether someone is seated, and whether there may be multiple people seated;

[0039] The present invention can specifically judge the specific position of the driver: seated, seated at the rear side, standing in front of the seat, on the left side of the vehicle, on the right side of the vehicle, when a person leaves the vehicle, and when no one is in the vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0041] Figure 1 The present invention is a flow chart of a method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter-wave radar. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] The present invention will be further described below in conjunction with the embodiments.

[0044] Please refer to the instruction manual Figure 1 A method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter-wave radar comprises the following steps:

[0045] 1. The radar starts, sends a frequency-modulated continuous pulse signal through the TX antenna, and receives the reflected echo signal of the target through multiple radar receiving antennas RX. Then, after 1D-FFT (one-dimensional Fourier transform) processing, 2D-FFT (two-dimensional fast Fourier transform) processing, CFAR (constant false alarm rate) and angle estimation, the point cloud data of the target is obtained;

[0046] According to the point cloud data, a density-based clustering method is used to obtain the target data (including the target's distance, angle, speed, and dynamic and static information), and the target data is processed by α-β filtering to output the target information of this frame;

[0047] Second, pre-process the target data of the frame, remove invalid data, and count the number of targets. When the number of targets is 0, directly update the monitoring status to unmanned status;

[0048] III. When the target is not 0, based on the range and angle of all targets in this frame, calculate the position points (x, y) of all targets:

[0049] x = range * sin(angle);

[0050] y = range * cos(angle);

[0051] When the number of targets is 1, cache the only target in this frame as the core target. When the number of targets > 1, a core target needs to be selected for storage, and the screening logic is as follows:

[0052] 3.1. Determine whether the target is within the two-wheeler seating area, and the constraint conditions are:

[0053] 0 < y <= y_centaland, x_cental_left <= x <= x_cental_right;

[0054] Among them, y_centaland is the seating boundary value in the y direction, x_cental_left is the minimum seating boundary value in the x direction, and x_cental_right is the maximum seating boundary value in the x direction. When there is a target within the seating area, cache this target as the core target point and end the screening;

[0055] 3.2. When there is no target within the seating area, calculate the distances from all targets to the center point target_center(x0, y0) of the two-wheeler, and cache the point with the shortest distance as the core target;

[0056] IV. After caching multi-frame data, process the multi-frame accumulated data. Divide the effective detection area into 6 parts. Taking 10 frames as an example:

[0057] 4.1. Based on all core target points (x_val, y_val) cached in 10 frames, determine which effective area each target is in;

[0058] When y1 < y_val < y2, x1 <= x_val <= x2, determine that this core target is in the seating area, where y1 and y2 are the upper and lower boundaries in the y direction of this area, and x1 and x2 are the left and right boundaries in the x direction respectively;

[0059] When y0 < y_val < y1, x1 <= x_val <= x2, determine that this core target is in the rear seating area, where y0 and y1 are the upper and lower boundaries in the y direction of this area, and x1 and x2 are the left and right boundaries in the x direction respectively;

[0060] When y2 < y_val < y3 and x1 <= x_val <= x2, it is determined that the core target is the standing area inside the vehicle, where y2 and y3 are the upper and lower boundaries in the y-direction of this area respectively, and x1 and x2 are the left and right boundaries in the x-direction respectively;

[0061] When y3 < y_val, it is determined that the core target is the area away from the vehicle, where y3 is the lower boundary in the y-direction of this area;

[0062] When 0 < y_val < y3 and x_val >= x2, it is determined that the core target is the right area, where 0 and y3 are the upper and lower boundaries in the y-direction of this area respectively, and x2 is the left boundary in the x-direction;

[0063] When 0 < y_val < y3 and x_val <= x1, it is determined that the core target is the left area, where 0 and y3 are the upper and lower boundaries in the y-direction of this area respectively, and x1 is the right boundary in the x-direction;

[0064] 4.2. Accumulate the targets that appear in each area. When the number of core targets in an area is greater than the threshold value V (for example, V = 6), update the monitoring status to the status represented by the area where it is located. If the number of points in no area is greater than the threshold value, update the status to the moving state;

[0065] V. Output the updated status, clear all target data cached in this judgment, and execute the judgment for the next cycle.

[0066] The present invention can realize the perception of the driver's body presence on a two-wheeled vehicle through a millimeter-wave radar; it can detect the presence of people around the vehicle, judge whether there are people around the electric vehicle, whether there are multiple people, whether someone is seated, and whether it is possible that multiple people are seated;

[0067] The present invention can specifically judge the specific position of the driver: seated, seated at the rear, standing in front of the seat, on the left side of the vehicle, on the right side of the vehicle, people away from the vehicle, and no one state.

[0068] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements will not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter wave radar, characterized in that: It includes the following steps: First, the radar is started, a frequency-modulated continuous pulse signal is sent through the TX antenna, and the reflected echo signals of the target are received through multiple radar receiving antennas RX. After processing, the point cloud data of the target is obtained; Based on the point cloud data, a density-based clustering method is performed to obtain target data, and the target data is subjected to α-β filtering processing, and the target information of this frame is output; Second, the target data of this frame is preprocessed, invalid data is剔除, and the number of targets is counted at the same time. When the number of targets is 0, the monitoring status is directly updated to the unmanned state; Third, when the target is not 0, based on the range and angle of all targets in this frame, the position points (x, y) of all targets are calculated through calculation; Among them, range represents the range, and angle represents the angle; When the number of targets is 1, the only target in this frame is cached as the core target. When the number of targets > 1, a core target is selected for storage; Fourth, after multiple frames of data are cached, the processed multi-frame accumulated data is divided into an occupancy area, a rear occupancy area, an in-vehicle standing area, an out-of-vehicle area, a right area, and a left area. After the targets are accumulated, the monitoring status is updated; Fifth, the updated status is output, and all target data cached in this judgment is cleared, and the judgment of the next cycle is executed.

2. The method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter-wave radar according to claim 1, characterized in that: The target data includes the distance, angle, speed, and static / dynamic information of the target.

3. The method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter wave radar according to claim 2, characterized in that: The calculation method of the position point (x, y) is as follows: x = range * sin(angle); y = range * cos(angle).

4. The method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter-wave radar according to claim 3, characterized in that: The screening method in step three is: 3.

1. Judge whether the target is within the two-wheeler occupancy area; 3.

2. When no target is within the occupancy area, calculate the distances from all targets to the center point target_center(x0, y0) of the two-wheeler, and cache the point with the closest distance as the core target.

5. The method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter-wave radar according to claim 4, characterized in that: In step four, based on all the cached core target points (x_val, y_val), judge which effective area each target is in; When y1 < y_val < y2 and x1 <= x_val <= x2, judge that the core target is in the occupancy area, where y1 and y2 are the upper and lower boundaries of the y direction of this area respectively, and x1 and x2 are the left and right boundaries in the x direction respectively; When y0 < y_val < y1 and x1 <= x_val <= x2, judge that the core target is in the rear occupancy area, where y0 and y1 are the upper and lower boundaries of the y direction of this area respectively, and x1 and x2 are the left and right boundaries in the x direction respectively; When y2 < y_val < y3 and x1 <= x_val <= x2, judge that the core target is in the in-vehicle standing area, where y2 and y3 are the upper and lower boundaries of the y direction of this area respectively, and x1 and x2 are the left and right boundaries in the x direction respectively; When y3 < y_val, judge that the core target is in the out-of-vehicle area, where y3 is the lower boundary of the y direction of this area; When \(0 < y\_val < y3\) and \(x\_val >= x2\), it is determined that the core target is in the right region, where 0 and y3 are the upper and lower boundaries of the region in the y direction respectively, and x2 is the left boundary in the x direction; When \(0 < y\_val < y3\) and \(x\_val <= x1\), it is determined that the core target is in the left region, where 0 and y3 are the upper and lower boundaries of the region in the y direction respectively, and x1 is the right boundary in the x direction.

6. The method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter wave radar according to claim 5, characterized in that: In step four, the targets appearing in each region are accumulated. When the number of core targets in a region is greater than the threshold value V, the monitoring state is updated to the state represented by the region where it is located. If the number of points in no region is greater than the threshold value, the state is updated to the motion state.

7. The method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter-wave radar according to claim 6, characterized in that: The threshold value V = 6.

8. The method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter-wave radar according to claim 7, characterized in that: In step 3.1, the constraint conditions are: \(0 < y <= y\_centaland\) and \(x\_cental\_left <= x <= x\_cental\_right\); Among them, y\_centaland is the seating boundary value in the y direction, x\_cental\_left is the minimum seating boundary value in the x direction, and x\_cental\_right is the maximum seating boundary value in the x direction. When there is a target in the seating area, the target is cached as a core target point and the screening ends.

9. The method for monitoring the presence of personnel on a two-wheeled vehicle based on millimeter-wave radar according to claim 8, characterized in that: The processing process in step one includes: 1D-FFT processing, 2D-FFT processing, CFAR and angle estimation.

Images