Monte Carlo simulation-based laser radar target detection performance evaluation method

Abstract

The invention discloses a laser radar target detection performance evaluation method based on Monte Carlo simulation. The method comprises the following steps: configuring hardware system parameters and light field structure parameters of a laser radar in advance; generating a spherical equidistant spiral scanning path, a wave beam pointing unit vector group at each pulse moment under the three-dimensional coordinate system and a range of a detection allowable elevation angle; randomly generating an initial position and an initial movement speed of the target, and ensuring that a speed vector points to the interior of the detection hemisphere; calculating a target position vector at the moment, setting a corresponding threshold judgment method and a detection threshold according to different light field structures, comparing with the beam vector at the corresponding moment, and judging whether the laser radar can successfully detect the target at the current moment; accumulatively counting corresponding capture results or not; and dividing the number of successful detection times by the total number of simulation times to obtain the detection probability. According to the method, the detection performance of the laser radar using different structured light fields on the random target can be accurately, quickly and extendibly evaluated.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of laser radar detection, and particularly relates to a laser radar target detection performance evaluation method based on Monte Carlo simulation. BACKGROUND

[0002] As a new generation of high-precision active detection means, laser radar (LiDAR, Light Detection and Ranging) has significant advantages in autonomous driving, space-based / sea-based surveillance, fine mapping and regional security warning, etc. due to its centimeter-level ranging accuracy, sub-milliradian spatial angle resolution, strong anti-environmental light interference capability, and stable working performance in day and night and multi-weather conditions. Especially in complex environment and high-speed target detection, laser radar can realize fast positioning and identification of unknown targets by actively emitting and receiving specific spatio-temporal modulated structured light field in the scanning process, providing important sensing support for target detection tasks.

[0003] The existing laser radar target detection performance evaluation mainly relies on field test or simplified analytical model. Although the field test is real, it is high in cost, long in period, and limited by weather and scene conditions, and it is difficult to systematically cover all operating states. Although the analytical model is fast in operation, it is often based on the simplified assumptions of ideal axisymmetric conical beam and fixed parameter scene, and cannot accurately describe the characteristics of laser radar using point, two-dimensional line, hollow ring and other structured light fields for detection. It is also difficult to introduce the comprehensive influence of target random spatial motion vector distribution, beam scanning strategy, system parameters (scanning frequency, field of view angle, angle resolution, etc.), resulting in a large deviation between the evaluation results and the real performance.

[0004] Chinese patent documents with publication numbers CN118244287A and CN118243960A respectively propose methods and devices for target detection using structured light field in atmospheric and ocean scattering media, but they only explain that the large spatial distribution characteristics of the structured light field can increase the possibility of detecting targets, and do not perform strict detection performance analysis, which is not practical enough.

[0005] Chinese patent document with publication number CN120254876A proposes a detailed and accurate method for target velocity vector detection using structured light field, but it only analyzes the quasi-static process in a very short time interval in a certain direction, without considering the random target detection process before velocity vector solving, which is not comprehensive enough.

[0006] Therefore, there is an urgent need for a method that can comprehensively consider the multi-parameter coupling effect of laser radar, truly reflect the space-time scanning characteristics of complex structured light field, support random motion target scene and fast convergence of probability statistics analysis, so as to accurately, quickly and repeatedly quantitatively evaluate the target detection performance of laser radar under different system configurations, light field forms and operating conditions. SUMMARY

[0007] The application provides a laser radar target detection performance evaluation method based on Monte Carlo simulation, which can comprehensively consider the random initial position of the target, the random speed, the scanning mode and trajectory of the laser radar, the system parameters and the detection threshold, support various structured light field beams, and realize fast, expandable and high-precision laser radar target detection performance evaluation.

[0008] A laser radar target detection performance evaluation method based on Monte Carlo simulation, comprising the following steps: (1) Pre-configure the hardware system parameters and light field structure parameters of the laser radar, and establish a three-dimensional coordinate system based on the position of the laser radar; (2) Based on the hardware system parameters and light field structure parameters of the laser radar, generate a spherical equidistant spiral scanning path, and obtain a set of beam pointing unit vectors at each pulse moment in the three-dimensional coordinate system and the range of the detection allowed elevation ; (3) Randomly generate the initial position and initial motion speed of the target based on the pre-configured maximum detection distance and the target motion parameters; the initial position of the target is located at any position on the hemispherical surface with as the radius and within the range of the detection allowed elevation ; the initial motion speed of the target is randomly distributed between , and is the preset upper limit of the target motion speed; (4) Detect the relationship between the generated initial position vector and the initial motion speed vector , if the dot product , discard the sample and repeatedly step (3) and detect until the dot product , to ensure that the speed vector points to the inside of the detection hemisphere; (5) Calculate the target position vector at moment , set the corresponding threshold judgment method and detection threshold according to different light field structures, and compare with the corresponding beam vector at moment to determine whether the laser radar can successfully detect the target at the current moment; (6) Repeat step (5) until the detection is successful or when the target does not return the flying-out detection area, end the single Monte Carlo simulation, and accumulate the count of the corresponding capture results; (7) Repeat steps (3)-(6) until the preset requirement is reached, and the laser radar detection target performance is defined as the number of successful detections divided by the total number of simulations, i.e., the detection probability.

[0009] In step (1), the origin O of the three-dimensional coordinate system is the position of the laser radar, the XOY plane is the horizontal ground, and the z-axis direction is vertically upward.

[0010] In step (1), the light field structure includes a point light source, a two-dimensional linear light source, and a hollow ring light source.

[0011] In the present application, the hardware system parameters include the laser radar scanning rotation speed, the laser pulse repetition frequency, the maximum detection distance and the maximum detection height of the laser radar, and the upper limit of the target random motion speed; the light field structure parameters include the maximum size of the beam divergence full angle and the beam angle spot divergence full angle.

[0012] In step (2), in the spherical equidistant spiral scanning path, the corresponding laser trajectory coordinates are as follows: In the formula, is the maximum detection distance, is the maximum detection height, is the maximum size of the light field at the position, is the parameter equation independent variable, corresponding to the polar angle of the laser radar beam pointing vector.

[0013] In step (2), the beam pointing unit vector group of each pulse moment is corresponding to a complete scanning trajectory from the minimum elevation angle to the maximum elevation angle. When used, it needs to be spliced in the order of first and last to simulate the continuous reciprocating scanning process of the laser radar in actual work.

[0014] In step (2), the detection allowed elevation angle is in the range of: wherein, is the maximum detection height, is the maximum detection distance, is the maximum size of the light field at the position.

[0015] ​​​​​Preferably, in step (3), the minimum initial motion speed of the target is limited to 0.1 m / s, so as to keep close to the minimum drift speed of the actual state target while optimizing the extremely long time consumption caused by random speed.

[0016] In step (5), time Based on the laser pulse repetition frequency The minimum sampling time interval is , to avoid unnecessary system time and space resource overhead caused by too short time division scale.

[0017] In step (5), calculate the time Lower target position vector , the formula is as follows: ; Among them, the target initial motion velocity vector After determination, it moves in a straight line at a uniform speed in three-dimensional space.

[0018] In step (5), the corresponding threshold judgment method and detection threshold are set according to different light field structures, and the corresponding t Beam vector at time Compare and judge whether the laser radar can successfully detect the target at the current moment. The specific formula is: Point light: ; Two-dimensional linear light source: ; Hollow ring light source: ; in, is the vector modulo operator, They are The minimum and maximum polar angles of a two-dimensional linear light source when it is transmitted in space at this moment, They are The polar angle and elevation angle of the target position at the moment, They are The minimum and maximum elevation angles of a two-dimensional linear light source when it is transmitted in space at this moment. The light field structure is defined by multiple divergence angles, and the maximum size of the beam divergence angle is , the full angle of the beam angular spot divergence is .

[0019] Preferably, the simulation is accelerated by multiple processes, and the random variables in all processes are selected in accordance with the independent repeated experiment rules. The results are summarized after each process is completed, which accelerates the operation of the simulation program and facilitates accurate and fast analysis.

[0020] Compared with the prior art, the present application has the following beneficial effects: The present application can fully combine structured light fields with different shape characteristics and parameters, can automatically set the laser radar scanning track and mode for structured light fields with different shapes, and can generate unknown random motion targets in real application scenarios, accurately and quickly calculate the detection probability of the corresponding laser radar on the target, and has the characteristics of high precision, fast calculation, strong robustness and high scalability. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0022] Figure 1 A flow chart of a laser radar target detection performance evaluation method based on Monte Carlo simulation is provided in the embodiments of the present application.

[0023] Figure 2 A schematic diagram of a laser radar working specific scene is provided in the embodiments of the present application.

[0024] Figure 3 A convergence trend graph of detection probability with increasing Monte Carlo simulation times for three different light field structures is provided in the embodiments of the present application. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0026] It should be noted that the features in the following embodiments and implementation manners can be combined with each other without conflict.

[0027] As shown in the drawings, a laser radar target detection performance evaluation method based on Monte Carlo simulation includes the following steps: Figure 1 (1) Pre-configure the hardware system parameters and light field structure parameters of the laser radar, and establish a three-dimensional coordinate system based on the position of the laser radar.

[0028] The origin O of the three-dimensional coordinate system is the position of the laser radar, the XOY plane is the horizontal ground, and the z-axis direction is vertically upward. ​

[0029] (2) Based on the hardware system parameters of laser radar and the light field structure parameters, a spherical equidistant spiral scanning path is generated, and the unit vector group of beam pointing at each pulse moment in the three-dimensional coordinate system is obtained and the range of the allowed elevation angle of detection.

[0030] In the spherical equidistant spiral scanning path, the corresponding laser trajectory coordinates are as follows: ; ; ; In the formula, is the maximum detection distance, is the maximum size of the light field at , and is the parameter equation independent variable corresponding to the polar angle of the laser radar beam pointing vector.

[0031] The unit vector group of beam pointing at each pulse moment corresponds to a complete scanning trajectory from the minimum elevation angle to the maximum elevation angle. When used, it needs to be spliced in the order of first and last to simulate the continuous reciprocating scanning process of the laser radar in actual work.

[0032] The range of the allowed elevation angle of detection is: ; Among them, is the maximum detection height, is the maximum detection distance, is the maximum size of the light field at .

[0033] (3) Randomly generate the initial position and initial motion speed of the target based on the pre-configured maximum detection distance and the target motion parameters; the initial position of the target is located on the hemispherical surface with as the radius and at any position within the range of the allowed elevation angle of detection ; the initial motion speed of the target is randomly distributed between , and is the upper limit of the preset target motion speed.

[0034] (4) Detect the relationship between the generated initial position vector of the target and the initial motion speed vector . If the dot product , discard the sample and repeatedly step (3) and detect until the dot product , to ensure that the speed vector points to the inside of the detection hemisphere.​

[0035] (5) Calculation time Target position vector at time , the corresponding threshold judgment method and detection threshold are set according to different light field structures, and the corresponding Beam vector at time Compare and judge whether the laser radar can successfully detect the target at the current time.

[0036] Time is based on the discrete sampling points of the laser pulse repetition frequency , and the minimum sampling time interval is , so as to avoid unnecessary system time and space resource overhead caused by too short time division scale.

[0037] Calculation time Target position vector at time , the formula is as follows: ; Wherein, the target initial motion velocity vector determines the uniform linear motion in three-dimensional space.

[0038] According to different light field structure, the corresponding threshold judgment method and detection threshold are set, and the corresponding t Beam vector at time Compare and judge whether the laser radar can successfully detect the target at the current time, the specific formula is: Point light source: ; Two-dimensional linear light source: ; Hollow ring light source: ; Wherein, is the vector norm operator, are the minimum value and maximum value of the polar angle of the two-dimensional linear light source in space at time , respectively, are the polar angle and the elevation angle of the target position at time , respectively, are the minimum value and maximum value of the elevation angle of the two-dimensional linear light source in space at time , the light field structure is defined by multiple divergence angles, the maximum size of the beam divergence full angle is , and the beam angle spot divergence full angle is .

[0039] (6) Repeat step (5) until the detection is successful or the target flies out of the detection area without returning, then end the single Monte Carlo simulation and accumulate the corresponding capture results.

[0040] (7) Repeat steps (3) to (6) until the preset requirements are met. The target detection performance of the lidar is defined as the number of successful detections divided by the total number of simulations, that is, the detection probability.

[0041] The specific working scenario of the laser radar corresponding to the present invention is as follows Figure 2 As shown, the laser radar is located at the origin of the three-dimensional space coordinate system and scans with a certain structured light field (point light source, two-dimensional linear light source or hollow ring light source, the shape and parameters of the light field do not change during the scanning process) and its corresponding trajectory. Starting from any position on the sphere, a target moves in a straight line at a random speed (random between the maximum and minimum speeds). Monte Carlo methods can be used to evaluate the detection performance of lidar for such random targets with no prior information.

[0042] In this embodiment, the laser radar system parameters are set as follows: laser pulse repetition frequency 50 kHz, laser radar maximum speed 2 rad / s and 40 rad / s, the structured light field parameter settings are shown in Table 1.

[0043] Table 1 Set the upper and lower limits of the random target speed to (0, 50 m / s], and conduct a laser radar to detect such random targets. The probability of successful detection converges with the number of independent Monte Carlo. Figure 3 As shown, the horizontal axis is the number of independent Monte Carlo, and the vertical axis is the probability of successful detection. The left figure corresponds to a laser radar speed of 2 rad / s. The right figure corresponds to a lidar rotation speed of 40 rad / s. It can be seen that for all three light fields, the detection probability shows a stable convergence trend as the Monte Carlo order increases, and the convergence paths and respective convergence values ​​are inconsistent, indicating that this method is highly feasible and stable.

[0044] The embodiments described above provide a detailed description of the technical solutions and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for evaluating target detection performance of laser radar based on Monte Carlo simulation, characterized in that: The following steps are involved: (1) Pre-configure the hardware system parameters and light field structure parameters of the lidar and establish a three-dimensional coordinate system based on the lidar position; (2) Based on the hardware system parameters and light field structure parameters of the lidar, a spherical equidistant spiral scanning path is generated, and the beam pointing unit vector group at each pulse moment in the three-dimensional coordinate system is obtained. And detection allowed elevation angle scope; (3) Based on pre-configured maximum detection distance The target initial position and initial motion speed are randomly generated by the target motion parameters; the target initial position is located at A hemisphere with a radius of Any position within the range; the initial speed of the target is Randomly distributed, The preset upper limit of target movement speed; (4) Detect the generated target initial position vector With the initial motion velocity vector If the dot product of the two , then discard the sample and repeat step (3) and test until the dot product , ensuring that the velocity vector points into the detection hemisphere; (5) Calculation time Lower target position vector , according to different light field structures, set the corresponding threshold judgment method and detection threshold, and the corresponding Beam vector at time Compare and determine whether the lidar can successfully detect the target at the current moment; (6) Repeat step (5) until the detection is successful or the target flies out of the detection area without returning, then end the single Monte Carlo simulation and accumulate the corresponding capture results; (7) Repeat steps (3) to (6) until the preset requirements are met. The target detection performance of the lidar is defined as the number of successful detections divided by the total number of simulations, that is, the detection probability.

2. The method for evaluating laser radar target detection performance based on Monte Carlo simulation according to claim 1, wherein: In step (1), the origin O of the three-dimensional coordinate system is the location of the laser radar, the XOY plane is the horizontal ground, and the z-axis is vertically upward.

3. The method for evaluating laser radar target detection performance based on Monte Carlo simulation according to claim 1, wherein: In step (1), the light field structure includes a point light source, a two-dimensional linear light source, and a hollow ring light source.

4. The method for evaluating laser radar target detection performance based on Monte Carlo simulation according to claim 1, wherein: In step (2), the corresponding laser trajectory coordinates in the spherical equidistant spiral scanning path are as follows: ； ； ； Where, is the maximum detection distance, for The maximum size of the light field at is the independent variable of the parametric equation, corresponding to the polar angle of the laser radar beam pointing vector.

5. The method for evaluating laser radar target detection performance based on Monte Carlo simulation according to claim 1, wherein: In step (2), the beam pointing unit vector group at each pulse moment is Corresponding to a complete scanning trajectory from the minimum elevation angle to the maximum elevation angle.

6. The method for evaluating laser radar target detection performance based on Monte Carlo simulation according to claim 1, wherein: In step (2), the detection allowed elevation angle The range is: ； in, is the maximum detection height, is the maximum detection distance, for The maximum size of the light field at .

7. The method for evaluating laser radar target detection performance based on Monte Carlo simulation according to claim 1, wherein: In step (5), time Based on the laser pulse repetition frequency f The minimum sampling time interval is .

8. The method for evaluating laser radar target detection performance based on Monte Carlo simulation according to claim 1, wherein: In step (5), calculate the time Lower target position vector , the formula is as follows: ； Among them, the target initial motion velocity vector After determination, it moves in a straight line at a uniform speed in three-dimensional space.

9. The method for evaluating laser radar target detection performance based on Monte Carlo simulation according to claim 1, wherein: In step (5), the corresponding threshold judgment method and detection threshold are set according to different light field structures, and the corresponding Beam vector at time Compare and judge whether the laser radar can successfully detect the target at the current moment. The specific formula is: Point light: ； Two-dimensional linear light source: ； Hollow ring light source: ； in, is the vector modulo operator, They are The minimum and maximum polar angles of a two-dimensional linear light source when it is transmitted in space at this moment, They are The polar angle and elevation angle of the target position at the moment, They are The minimum and maximum elevation angles of a two-dimensional linear light source when it is transmitted in space at this moment. The light field structure is defined by multiple divergence angles, and the maximum size of the beam divergence angle is , the full angle of the beam angular spot divergence is .

Images