CN112097792A - An Ackerman Model Mobile Robot Odometer Calibration Method - Google Patents

Abstract

The invention discloses a method for calibrating the odometer of an Ackerman model mobile robot. Integrate the speed to obtain the moving distance of the robot; obtain the rotation angle of the mobile robot within a certain period of time through the IMU; obtain the distance and angle of the corner point relative to the mobile robot at different sampling times through the lidar tracking a single corner point in the environment. Calculate the estimated displacement and real displacement of the mobile robot according to the obtained relevant data, obtain the error coefficient, and complete the odometer calibration of the mobile robot. The invention utilizes the high-precision features of the IMU and the laser radar to complete the calibration of the robot odometer, improves the accuracy of the pose estimation of the robot during the movement process, and further improves the accuracy of the mobile robot during mapping, positioning and navigation, and is applied to the mobile robot. At the same time positioning and map construction technology field.

Description

An Ackerman Model Mobile Robot Odometer Calibration Method

technical field

The invention relates to the technology in the field of mobile robot odometer calibration, in particular to an Ackerman model mobile robot odometer calibration method.

Background technique

With the rapid development of computer technology, machine vision, artificial intelligence and other technologies, mobile robots have also received more in-depth research and increasingly widespread applications. In the field of national defense, drones and unmanned vehicles are used for reconnaissance, intelligence collection and tracking; in the field of logistics, AGV cars have become an important part of intelligent logistics systems. In the service field, various cleaning robots, welcome robots, catering robots, shopping guide robots and medical robots have also been launched one after another, and SLAM technology, one of the supporting technologies, is also developing continuously. In the SLAM technology, the pose estimation of the mobile robot is particularly important. However, due to the assembly error of the mobile robot, the gap between the teeth, the wheel slippage during the movement and other reasons, when the mobile robot uses the odometer information for pose estimation, it is often There is a relatively large error, so it is particularly important to calibrate the odometer.

Most of the existing odometer calibration methods are aimed at two-wheel differential mobile robots driven by two motors. The structure of the robot is relatively simple, and manual distance measurement is often used when measuring the relevant distance, which has low precision. When calibrating, the robot is often scheduled to run. The trajectory reduces the diversity of robot motion forms, which is often quite different from the actual motion situation, and different motion forms often produce different errors, which makes the calibration results not very universal and the efficiency is relatively low. . At the same time, the structure of the Ackerman model is relatively complex, and it is obviously different from the two-wheel differential mobile robot. Therefore, the calibration method for other models of mobile robots may not be suitable for the robot of this model.

In view of the defects of the prior art, the present invention proposes a method for calibrating the odometer of the Ackerman model robot. The present invention mainly uses the laser radar with high accuracy in ranging and the IMU inertial measurement unit with relatively accurate angle. Calibration of the odometer. The invention takes the laser radar and the IMU as the benchmark, calculates the displacement of the car within a certain time interval, and then uses the odometer information to estimate the displacement of the car, and compares the two to obtain an error coefficient, and then calculates the displacement estimated from the odometer information. Adjust to achieve the purpose of correctly estimating the car's pose.

SUMMARY OF THE INVENTION

In view of the above problems, the purpose of the present invention is to provide a method for calibrating the odometer of the Ackerman model mobile robot, by using the laser radar with higher accuracy in ranging and the LMU with more accurate angle in measuring the odometer to calibrate the odometer, Improve the calibration accuracy and calibration efficiency of the Ackerman model mobile robot odometer calibration.

To achieve the above object, the present invention adopts the following technical solutions:

(1) Obtain the real rotation angle of the mobile robot during the movement process through the IMU installed on the mobile robot;

(2) Calculate the movement speed of the mobile robot by calculating the number of pulses per unit time of the wheel encoder, and integrate the speed to obtain the movement distance of the mobile robot within a certain time interval;

(3) According to the moving distance of the mobile robot and the rotation angle of the mobile robot obtained by the IMU, use the track estimation algorithm to obtain the estimated displacement of the mobile robot;

(4) Track a single feature corner point in the environment through lidar, obtain its distance and angle relative to the mobile robot at the two sampling moments, combine the IMU rotation angle, and use geometric derivation to calculate the real displacement of the mobile robot;

(5) The error coefficient is obtained by comparing the estimated displacement of the mobile robot with the real displacement of the mobile robot, and the estimated displacement of the mobile robot is adjusted to realize the odometer calibration.

Preferably, in the step (3), the step of obtaining the estimated displacement of the mobile robot is as follows:

(3-1) Obtain the motor speed ω _m , then the moving linear velocity of the mobile robot body is v _c =r·ω _m ; where r is the wheel radius of the mobile robot;

(3-2) Integrate the speed of the mobile robot to obtain the moving distance s within a given time interval;

(3-3) Take the midpoint of the line connecting the rear wheels of the mobile robot as the reference point O, and the reference point moves from A(x, y) to B(x', y') around the O point;

其中s为小车给定时间内运动的距离，θ₁,θ₂分别为小车两次采样时的姿态角；(3-4) According to the track estimation algorithm, we get:

Among them, s is the distance that the car moves in a given time, and θ ₁ and θ ₂ are the attitude angles of the car when it is sampled twice;

(3-5) Calculate the displacement of the mobile robot according to the coordinates of the mobile robot obtained by the two samplings

Preferably, in the step (4), the step of obtaining the real displacement of the mobile robot is as follows:

(4-1) At the initial moment, the lidar observes the corner point, and outputs the relative positional relationship between the corner point and the radar coordinate system, the distance to the lidar is d ₁ , and the angle is α ₁ ;

(4-2) At the end point, the distance of the corner point relative to the lidar is d ₂ , and the angle is α ₂ ;

(4-3) In the time T, the rotation angle of the robot is Δθ, which is obtained by the IMU, and the angle between the two distance measurements of the lidar diagonal point is α ₃ , then α ₃ =α ₁ +α ₂ + can be calculated Δα;

(4-4) The robot displacement is:

其中d₃即为由激光雷达信息得到的小车真实位移。

where d ₃ is the real displacement of the car obtained from the lidar information.

Preferably, in the step (5), the step of obtaining the odometer error coefficient is as follows:

其中l_odom为由里程计信息的到的估计位移，l_laser为由激光雷达信息得到的真实位移；通过多次标定，取δ_l的平均值

(5-1) Calculate the error coefficient from multiple sets of data obtained by one calibration

where l _odom is the estimated displacement obtained from the odometer information, and l _laser is the actual displacement obtained from the lidar information; through multiple calibrations, the average value of δ _l is taken

取

的平均值，得到更具有普适性的

(5-2) Repeat the step (5-1), control the mobile robot to travel according to different trajectories, and obtain multiple sets of

Pick

, to obtain a more general

作为误差系数，加入到里程计信息估计

并控制小车运动，比较l与l_laser，验证该系数的正确性，其中l为标定后的估计位移。(5-3) with

As an error coefficient, added to the odometer information estimate

And control the movement of the trolley, compare l and l _laser to verify the correctness of the coefficient, where l is the estimated displacement after calibration.

Compared with the prior art, the present invention has the following obvious outstanding substantive features and remarkable technological progress:

1. The reference distance used for calibration in the present invention is obtained by geometrically deriving laser radar and IMU data, which has high precision and eliminates the interference of artificial ranging;

2. The present invention does not preset the movement trajectory of the robot, so that the movement forms of the robot are diversified, and the calibration results are more universal;

3. The present invention can sample and calculate multiple times during one movement of the robot to obtain multiple sets of error coefficients, which greatly improves the calibration efficiency.

Description of drawings

The accompanying drawings are only for illustrating the preferred embodiments and do not limit the present invention.

FIG. 1 is the motion environment required for the odometer calibration of the mobile robot according to the present invention.

FIG. 2 is a dynamic model diagram of the Ackerman model mobile robot of the odometer calibration method of the Ackerman model mobile robot according to the present invention.

FIG. 3 is a schematic diagram of inferring the displacement of the mobile robot by using the laser radar in the method for calibrating the odometer of the Ackerman model mobile robot according to the present invention.

FIG. 4 is a schematic diagram of the track reckoning of the mobile robot in the method for calibrating the odometer of the Ackerman model mobile robot according to the present invention.

FIG. 5 is a data processing flow chart of the method for calibrating the odometer of the Ackerman model mobile robot according to the present invention.

Detailed ways

In order to make the above-mentioned purposes, technical solutions and advantages of the present invention clearer and easier to understand, the present invention will be described in detail below with reference to the accompanying drawings and preferred embodiments. .

Example 1:

An Ackerman model mobile robot odometer calibration method, the operation steps are as follows:

(1) Obtain the real rotation angle of the mobile robot during the movement process through the IMU installed on the mobile robot;

(2) Calculate the movement speed of the mobile robot by calculating the number of pulses per unit time of the wheel encoder, and integrate the speed to obtain the movement distance of the mobile robot within a certain time interval;

(3) According to the moving distance of the mobile robot and the rotation angle of the mobile robot obtained by the IMU, use the track estimation algorithm to obtain the estimated displacement of the mobile robot;

(4) Track a single feature corner point in the environment through lidar, obtain its distance and angle relative to the mobile robot at the two sampling moments, combine the IMU rotation angle, and use geometric derivation to calculate the real displacement of the mobile robot;

(5) The error coefficient is obtained by comparing the estimated displacement of the mobile robot with the real displacement of the mobile robot, and the estimated displacement of the mobile robot is adjusted to realize the odometer calibration.

In this embodiment, the odometry calibration method of the Ackerman model mobile robot uses the laser radar with higher accuracy in ranging and the LMU with relatively accurate angle measurement to calibrate the odometer, so as to improve the odometer of the Ackerman model mobile robot. Calibration accuracy and calibration efficiency.

Embodiment 2:

This embodiment is basically the same as the first embodiment, and the special features are as follows:

In this embodiment, in the step (3), the steps of obtaining the estimated displacement of the mobile robot are as follows:

(3-1) Obtain the motor speed ω _m , then the moving linear velocity of the mobile robot body is v _c =r·ω _m ; where r is the wheel radius of the mobile robot;

(3-2) Integrate the speed of the mobile robot to obtain the moving distance s within a given time interval;

(3-3) Take the midpoint of the line connecting the rear wheels of the mobile robot as the reference point O, and the reference point moves from A(x, y) to B(x', y') around the O point;

其中s为小车给定时间内运动的距离，θ₁,θ₂分别为小车两次采样时的姿态角；(3-4) According to the track estimation algorithm, we get:

Among them, s is the distance that the car moves in a given time, and θ ₁ and θ ₂ are the attitude angles of the car when it is sampled twice;

(3-5) Calculate the displacement of the mobile robot according to the coordinates of the mobile robot obtained by the two samplings

In this embodiment, in the step (4), the step of obtaining the real displacement of the mobile robot is as follows:

(4-1) At the initial moment, the lidar observes the corner point, and outputs the relative positional relationship between the corner point and the radar coordinate system, the distance to the lidar is d ₁ , and the angle is α ₁ ;

(4-2) At the end point, the distance of the corner point relative to the lidar is d ₂ , and the angle is α ₂ ;

(4-3) In the time T, the rotation angle of the robot is Δα, which is obtained by the IMU, and the angle between the two distance measurements of the lidar diagonal point is α ₃ , then α ₃ =α ₁ +α ₂ + can be calculated Δα;

(4-4) The robot displacement is:

其中d₃即为由激光雷达信息得到的小车真实位移。

where d ₃ is the real displacement of the car obtained from the lidar information.

In the present embodiment, in the step (5), the step of obtaining the odometer error coefficient is as follows:

其中l_odom为由里程计信息的到的估计位移，l_laser为由激光雷达信息得到的真实位移；通过多次标定，取δ_l的平均值

(5-1) Calculate the error coefficient from multiple sets of data obtained by one calibration

where l _odom is the estimated displacement obtained from the odometer information, and l _laser is the actual displacement obtained from the lidar information; through multiple calibrations, the average value of δ _l is taken

取

的平均值，得到更具有普适性的

(5-2) Repeat the step (5-1), control the mobile robot to travel according to different trajectories, and obtain multiple sets of

Pick

, to obtain a more general

作为误差系数，加入到里程计信息估计

并控制小车运动，比较l与l_laser，验证该系数的正确性，其中l为标定后的估计位移。(5-3) with

As an error coefficient, added to the odometer information estimate

And control the movement of the trolley, compare l and l _laser to verify the correctness of the coefficient, where l is the estimated displacement after calibration.

The method of this embodiment utilizes the high-precision features of the IMU and the laser radar to complete the calibration of the robot odometer, improve the accuracy of the robot's pose estimation during the movement process, and further improve the accuracy of the mobile robot during mapping, positioning, and navigation.

Embodiment three:

This embodiment is basically the same as the above-mentioned embodiment, and the special features are as follows:

In this embodiment, FIG. 1 is a schematic diagram of a moving environment of a mobile robot. It is required that in this environment, there is only one corner point within the detectable range of the lidar for the lidar to track. Figure 2 is the dynamic model of the Ackerman model mobile robot. The lidar and IMU can be installed on the head and tail of the robot along the central axis of the robot. Figure 3 is a schematic diagram of the displacement of the mobile robot measured by the laser. According to the two samplings, the displacement of the mobile robot is estimated. Figure 4 is a schematic diagram of the dead reckoning of the mobile robot with the Ackerman model. Using the encoder and IMU information, the pose of the mobile robot at the two sampling moments is estimated. Figure 5 is the data processing flow of the Ackerman model mobile robot odometer calibration.

The odometer calibration method for the Ackerman model mobile robot in this embodiment includes steps S101-S108:

Step S101, controlling the mobile robot to move in the environment as shown in FIG. 1, and the movement form can be a straight line, a circle or a combination of the two;

Step S102, use the IMU to obtain the rotation angle Δθ of the mobile robot within a certain period of time;

Step S103, in conjunction with Fig. 2, use the encoder to obtain the motor speed ω _m of the mobile robot per unit time, then the linear velocity of the mobile robot motion is _vc =r·ω _m , where r is the radius of the wheel of the mobile robot, and _vc is integrated. The movement distance s that can be obtained from the odometer;

其中d₃即为由激光雷达信息得到的小车位移；Step S104, referring to Fig. 3, the coordinate system xOy and the coordinate system xO'y are respectively represented as the pose transformation of the lidar coordinate system within the time difference T; point A represents the only corner point in the environment, and at the initial moment, the lidar observed Corner point, and output the relative position relationship between the corner point and the radar coordinate system. The distance from the corner point to the lidar is d ₁ , and the angle is θ ₁ . At the end point, the distance between the corner point and the lidar is d ₂ , the angle is θ ₂ , the rotation angle of the robot is Δθ in time T, which is obtained by the IMU. The angle between the two distance measurements of the lidar diagonal point is θ ₃ , then θ ₃ =θ ₁ +θ ₂ +Δθ is calculated, and the robot displacement is:

where d ₃ is the car displacement obtained from the lidar information;

则在多次采样之后，小车相对于起始点的位移为：

其中θ₁，θ₂为移动机器人在A，B两点时的姿态角；Step S105, in conjunction with Fig. 4, take the midpoint O' of the rear wheel connection as the reference point, and the reference point moves from A(x, y) to B(x', y') around the O point, then obtains according to the track estimation algorithm:

Then after multiple sampling, the displacement of the car relative to the starting point is:

where θ ₁ , θ ₂ are the attitude angles of the mobile robot at points A and B;

Step S106, referring to Fig. 5, S _cor and θ _cor are the distance and angle of the corner point relative to the lidar, respectively, S _odom is the distance of the mobile robot estimated by the wheeled odometer within a certain period of time, and θ _IMU is the distance obtained by the IMU. According to the above data, the displacement l _laser calculated by the lidar and the displacement estimated by the odometer l _odom are obtained, and the error coefficient is defined as

则小车的真实位移

完成里程计标定；Step S107, through multiple calibrations, multiple groups of δ _l are obtained, and the average value of δ _l is taken

Then the true displacement of the car

Complete the odometer calibration;

Step S108, after the error coefficient is obtained, it is added to the mobile robot pose estimation to verify its correctness.

The method for calibrating the odometer of the Ackerman model mobile robot in this embodiment includes installing a wheel encoder, a laser radar, and an inertial measurement unit IMU on the mobile robot, respectively. The speed of the mobile robot is obtained by the wheel encoder installed on the motor, and the moving distance of the robot is obtained by integrating the speed; the rotation angle of the mobile robot in a certain period of time is obtained by the IMU; the single corner point in the environment is tracked by the lidar, and different The distance and angle of the lower corner point relative to the mobile robot at the sampling time; the estimated displacement and the actual displacement of the mobile robot are calculated according to the obtained relevant data, the error coefficient is obtained, and the odometer calibration of the mobile robot is completed. This embodiment uses the high-precision features of the IMU and lidar to complete the calibration of the robot's odometer, improve the accuracy of the robot's pose estimation during the movement process, and further improve the accuracy of the mobile robot during mapping, positioning, and navigation, and is suitable for applications. In the field of simultaneous localization and mapping of mobile robots (SLAM) technology.

The embodiments of the present invention have been described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned embodiments, and various changes can also be made according to the purpose of the invention and creation of the present invention. Changes, modifications, substitutions, combinations or simplifications should be equivalent substitution methods, as long as they meet the purpose of the present invention, as long as they do not deviate from the technical principles and inventive concepts of the present invention, all belong to the protection scope of the present invention.

Claims (4)

1. an Ackerman model mobile robot odometer calibration method, is characterized in that, operating steps are as follows: (1) Obtain the real rotation angle of the mobile robot during the movement process through the IMU installed on the mobile robot; (2) Calculate the movement speed of the mobile robot by calculating the number of pulses per unit time of the wheel encoder, and integrate the speed to obtain the movement distance of the mobile robot within a certain time interval; (3) According to the moving distance of the mobile robot and the rotation angle of the mobile robot obtained by the IMU, use the track estimation algorithm to obtain the estimated displacement of the mobile robot; (4) Track a single feature corner point in the environment through lidar, obtain its distance and angle relative to the mobile robot at the two sampling moments, combine the IMU rotation angle, and use geometric derivation to calculate the real displacement of the mobile robot; (5) The error coefficient is obtained by comparing the estimated displacement of the mobile robot with the real displacement of the mobile robot, and the estimated displacement of the mobile robot is adjusted to realize the odometer calibration. 2. Ackerman model mobile robot odometer calibration method according to claim 1, is characterized in that, in described step (3), the step of obtaining the estimated displacement of mobile robot is as follows: (3-1) Obtain the motor speed ω _m , then the moving linear velocity of the mobile robot body is v _c =r·ω _m ; where r is the wheel radius of the mobile robot; (3-2) Integrate the speed of the mobile robot to obtain the moving distance s within a given time interval; (3-3) Take the midpoint of the line connecting the rear wheels of the mobile robot as the reference point O, and the reference point moves from A(x, y) to B(x', y') around the O point; (3-4) According to the track estimation algorithm, we get:

Among them, s is the distance that the car moves in a given time, and θ ₁ and θ ₂ are the attitude angles of the car when it is sampled twice; (3-5) Calculate the displacement of the mobile robot according to the coordinates of the mobile robot obtained by the two samplings

3. Ackerman model mobile robot odometer calibration method according to claim 1, is characterized in that, in described step (4), the step of obtaining the true displacement of mobile robot is as follows: (4-1) At the initial moment, the lidar observes the corner point, and outputs the relative positional relationship between the corner point and the radar coordinate system, the distance to the lidar is d ₁ , and the angle is α ₁ ; (4-2) At the end point, the distance of the corner point relative to the lidar is d ₂ , and the angle is α ₂ ; (4-3) In the time T, the rotation angle of the robot is Δα, which is obtained by the IMU, and the angle between the two distance measurements of the lidar diagonal point is α ₃ , then α ₃ =α ₁ +α ₂ + can be calculated Δα; (4-4) The robot displacement is:

where d ₃ is the real displacement of the car obtained from the lidar information. 4. Ackerman model mobile robot odometer calibration method according to claim 1, is characterized in that, in described step (5), the step that obtains odometer error coefficient is as follows: (5-1) Calculate the error coefficient from multiple sets of data obtained by one calibration

where l _odom is the estimated displacement obtained from the odometer information, and l _laser is the actual displacement obtained from the lidar information; through multiple calibrations, the average value of δ _l is taken

(5-2) Repeat the step (5-1), control the mobile robot to travel according to different trajectories, and obtain multiple sets of

Pick

, to obtain a more general

(5-3) with

As an error coefficient, added to the odometer information estimate

And control the movement of the trolley, compare l and l _laser to verify the correctness of the coefficient, where l is the estimated displacement after calibration.

Images