Lidar and imu-based instant positioning method

Abstract

The invention relates to an instant positioning method based on lidar and imu, and belongs to the field of intelligent robots and automatic driving. According to the invention, the data acquisition module obtains the lidar data and the imu data, and inputs the data into the instant positioning program; performing motion distortion correction on the point cloud in the lidar scanning data by utilizing imu data in a data optimization module; correcting the lidar data by using an adaptive data association module; the prediction state of the lidar data and the prediction state of the imu data are input into an extended Kalman filtering optimization model for optimization processing, and an optimized result is obtained; and performing pose estimation on the optimized result to obtain pose result information. According to the method, the maintainability and the flexibility of the program are improved; propagation of data errors in a program is reduced; and the workload of parameter adjustment of the program in different types of environments is reduced.

Description

Technical Field

[0001] The invention belongs to the field of intelligent robots and automatic driving, and in particular relates to an instant positioning method based on lidar and imu. Background Art

[0002] Real-time positioning method refers to the technology of using sensors to perceive the surrounding environment in real time and locate the position of the self in an unknown or partially known environment. Real-time positioning method is affected by factors such as environment and computing resources, which may easily lead to the positioning program not being able to run in real time and cumbersome tuning of different environmental parameters.

[0003] When the current real-time positioning method is run in different environments, the parameters need to be adjusted to better adapt to the environment and produce the best positioning effect. The real-time positioning method based on lidar and imu proposed in the present invention can reduce parameter adjustment and reduce the adaptation time cost when intelligent robots and autonomous driving technologies are applied to a variety of different environments. Summary of the invention

[0004] 1. Technical issues to be resolved

[0005] The technical problem to be solved by the present invention is how to provide a real-time positioning method based on lidar and imu to solve the problem that the positioning program cannot be run in real time and the tuning of different environmental parameters is cumbersome.

[0006] (II) Technical solution

[0007] In order to solve the above technical problems, the present invention proposes a real-time positioning method based on lidar and imu, which comprises the following steps:

[0008] Step S101, obtain lidar data and imu data through the data acquisition module, and input them into the real-time positioning program;

[0009] Step S102: using the imu data in the data optimization module to perform motion distortion correction on the point cloud in the lidar scanning data;

[0010] Step S103, using an adaptive data association module to correct lidar data;

[0011] Step S104: input the predicted state of lidar data and the predicted state of imu data into the extended Kalman filter optimization model for optimization processing to obtain an optimized result;

[0012] Step S105: perform pose estimation on the optimized result to obtain pose result information.

[0013] (III) Beneficial effects

[0014] The present invention proposes a real-time positioning method based on lidar and imu. The real-time positioning method designs an adaptive data association module, which takes into account the uncertainty of the predicted state, the map discretization error and the lidar distance noise. The processing of this module greatly reduces the workload of the program for parameter adjustment in different types of environments.

[0015] The present invention provides an instant positioning method based on lidar and imu, which can bring the following expected effects: 1) improve the maintainability and flexibility of the program; 2) reduce the propagation of data errors in the program; 3) reduce the workload of parameter adjustment of the program in different types of environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a flow chart of the instant positioning method based on lidar and imu of the present invention;

[0017] Figure 2 This is the framework diagram of the real-time positioning program based on lidar and imu;

[0018] Figure 3 This is a diagram showing the operating effect of the real-time positioning method based on lidar and imu. DETAILED DESCRIPTION

[0019] In order to make the purpose, content and advantages of the present invention more clear, the specific implementation methods of the present invention are further described in detail below in conjunction with the drawings and examples.

[0020] Currently, common real-time positioning methods mostly use different parameter configurations adjusted according to different environments, especially the adjustment of feature extraction, number of iterations and data association threshold. The present invention proposes an adaptive data association method, which comprehensively considers factors such as relative pose uncertainty, map discretization error and lidar noise, greatly reducing the workload of parameter adjustment in different types of environments.

[0021] The present invention relates to a real-time positioning method, the program can be stored on a computer-readable device, and provides data results to meet the needs of real-time positioning. The program relies on an adaptive data association method, which reduces the workload of parameter adjustment in different environments.

[0022] The present invention proposes an instant positioning method based on lidar and imu, and proposes an adaptive data association method by comprehensively considering factors such as relative pose uncertainty, map discretization error and lidar noise, which effectively reduces the workload of parameter adjustment in different types of environments and improves the robustness of the program.

[0023] The instant positioning method proposed in the present invention adopts the extended Kalman filter scheme of point-to-point registration and the adaptive data association module to provide a robust and real-time positioning technology for the instant positioning method. In terms of technical implementation, the instant positioning method provided by the present invention adopts modular architecture design, tightly coupled data processing, and adaptive threshold modules to provide robust, convenient, and fast-response instant positioning services for instant positioning software and products.

[0024] Figure 1 The instant positioning method based on lidar and imu provided by the present invention is demonstrated, and the method comprises the following steps:

[0025] Step S101, obtain lidar data and imu data through the data acquisition module, and input them into the real-time positioning program;

[0026] Step S102: In the data optimization module, the imu data is used to perform motion distortion correction on the point cloud in the lidar scanning data.

[0027] The specific method is as follows: in the interval between two adjacent frames of lidar scanning data, pre-integrate the imu data to obtain the distance and angle; perform motion distortion correction on the obtained distance and angle information and the two adjacent frames of lidar scanning data to optimize the point cloud position in each frame of lidar scanning data;

[0028] Step S103, using an adaptive data association module to correct lidar data;

[0029] The lidar data is corrected using the adaptive data association module. The specific method is as follows: In order to correct the association between point clouds and reduce the influence of outlier point clouds (point clouds with large deviations), a threshold τ is introduced to limit the maximum association distance between corresponding points. The threshold τ takes into account the uncertainty of the predicted state, the map discretization error, and the lidar distance noise.

[0030] The uncertainty of the predicted state refers to the distance between the initial estimated pose and the true pose. The error in the relative pose between two consecutive scans is calculated by integrating the noise from the IMU data over the time interval between the two lidar scan data.

[0031] Map discretization error refers to the use of a three-dimensional grid map to represent the point cloud on the map, while limiting the maximum number of point clouds in each grid to m.

[0032] Lidar range noise refers to the sensor noise of lidar distance measurement, which is expressed by the noise variance σ 2 Compute the threshold τ at time k.

[0033] Step S104: input the predicted state of lidar data and the predicted state of imu data into the extended Kalman filter optimization model for optimization processing to obtain an optimized result;

[0034] Specifically, the lidar data and the imu data are optimized using the data optimization module. The specific method is as follows: the lidar data and the imu data corrected by the adaptive data association module are respectively input into the extended Kalman filter optimization model for optimization processing to obtain the optimized lidar data and the imu data respectively;

[0035] Step S105: perform pose estimation on the optimized result to obtain pose result information.

[0036] Use the pose estimation module to obtain the pose results. The specific method is as follows: input the lidar data and imu data optimized by the extended Kalman filter optimization model into the state estimation model to obtain the state quantities of position and direction.

[0037] Embodiment 1:

[0038] The invention discloses a real-time positioning method based on lidar and imu, which specifically provides real-time positioning service for users.

[0039] Furthermore, pre-integration processing and motion de-distortion processing are performed on the imu and lidar data of the original sensor.

[0040] Furthermore, the lidar data processed with motion distortion correction is input into adaptive data association to optimize the data;

[0041] Furthermore, the corrected and predicted lidar data and the original imu data are input into the extended Kalman filter optimization module for optimization processing;

[0042] Furthermore, the optimized data is used for pose estimation to obtain accurate pose result information.

[0043] Furthermore, the instant positioning method adopts a modular architecture mode, providing users with flexible and easy-to-maintain program features.

[0044] Furthermore, the real-time positioning method adopts a tightly coupled data processing approach, which reduces the propagation of errors in the program.

[0045] The present invention also provides a computer service program, which implements the method when executed by a processor.

[0046] Innovation of the present invention

[0047] The instant positioning method adopts a technology based on modular architecture design, including: data acquisition module, data optimization module, adaptive data association module and posture estimation module, which reduces the complexity of the program and improves the maintainability and flexibility of the program;

[0048] This real-time positioning method adopts the technology of tightly coupled data processing, that is, lidar data and imu data are input into the pose estimation model together, and the final position, direction, speed and other state quantities are estimated with the purpose of minimizing the overall error of lidar data and imu data, which improves the accuracy and robustness of data processing in the program and reduces the propagation of data errors in the program.

[0049] This real-time positioning method designs an adaptive data association module, which takes into account the uncertainty of the predicted state, map discretization error and lidar distance noise. The processing of this module greatly reduces the workload of the program to adjust parameters in different types of environments.

[0050] The present invention provides an instant positioning method based on lidar and imu, which can bring the following expected effects: 1) improve the maintainability and flexibility of the program; 2) reduce the propagation of data errors in the program; 3) reduce the workload of parameter adjustment of the program in different types of environments.

[0051] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A real-time positioning method based on lidar and imu, characterized in that: The method comprises the following steps: Step S101, obtain lidar data and imu data through the data acquisition module, and input them into the real-time positioning program; Step S102: using the imu data in the data optimization module to perform motion distortion correction on the point cloud in the lidar scanning data; Step S103, using an adaptive data association module to correct lidar data; Step S104: input the predicted state of lidar data and the predicted state of imu data into the extended Kalman filter optimization model for optimization processing to obtain an optimized result; Step S105: perform pose estimation on the optimized result to obtain pose result information.

2. The real-time positioning method based on lidar and imu as claimed in claim 1, characterized in that: The step S102 includes: performing pre-integration processing on the imu data in the interval time period between two adjacent frames of lidar scanning data to obtain the distance and angle; performing motion distortion correction processing on the obtained distance and angle information and the two adjacent frames of lidar scanning data to optimize the point cloud position in each frame of lidar scanning data.

3. The real-time positioning method based on lidar and imu as claimed in claim 2, characterized in that: The step S103 includes: in order to correct the association between point clouds and reduce the influence of outlier point clouds, a threshold τ is introduced to limit the maximum association distance between corresponding points.

4. The real-time positioning method based on lidar and imu as claimed in claim 3, characterized in that: The threshold τ takes into account the uncertainty of the predicted state, the map discretization error, and the lidar range noise.

5. The real-time positioning method based on lidar and imu as claimed in claim 4, characterized in that: The uncertainty of the predicted state refers to the distance between the initial estimated pose and the true pose. The error in the relative pose between two consecutive scans is calculated by integrating the noise from the IMU data over the time interval between the two lidar scan data.

6. The real-time positioning method based on lidar and imu as claimed in claim 4, characterized in that: Map discretization error refers to the use of a three-dimensional grid map to represent the point cloud on the map, while limiting the maximum number of point clouds in each grid to m.

7. The real-time positioning method based on lidar and imu as claimed in claim 4, characterized in that: Lidar range noise refers to the sensor noise of lidar distance measurement, which is expressed by the noise variance σ 2 Compute the threshold τ at time k.

8. The real-time positioning method based on lidar and imu according to any one of claims 3 to 7, characterized in that: The step S104 includes: inputting the lidar data and imu data corrected by the adaptive data association module into the extended Kalman filter optimization model for optimization processing, and obtaining optimized lidar data and imu data respectively.

9. The real-time positioning method based on lidar and imu as claimed in claim 8, characterized in that: The step S104 uses a data optimization module to optimize lidar data and imu data.

10. The real-time positioning method based on lidar and imu as claimed in claim 8, characterized in that: The step S105 includes: inputting the lidar data and imu data optimized by the extended Kalman filter optimization model into the state estimation model to obtain the state quantities of position and direction.

Images