Laser slam positioning method and system fusing gnss observation information

Abstract

The application discloses a kind of laser SLAM positioning method and system of fusion GNSS observation information.The method comprises: obtaining the laser radar data of target space inner carrier, inertial sensor data and GNSS data;Inertial sensor data is constructed based on inertial sensor data constraint factor;Laser radar data is constructed based on laser radar constraint factor;Doppler data constraint factor, code pseudorange constraint factor and clock error constraint factor are constructed based on GNSS data;The constraint factor is optimized based on sliding window optimization method, and the positioning information of target space inner carrier is obtained.The application combines GNSS / IMU and laser SLAM by the above method, utilizes the high-precision positioning advantage of laser SLAM in short time, and GNSS provides global coordinate, and there is no error accumulation characteristics.The application fuses the observation data of GNSS, inertial measurement sensor and laser radar three kinds of sensors, and carries out real-time indoor and outdoor seamless high-precision positioning.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of multi-sensor fusion positioning, in particular to a laser SLAM positioning method and system fusing GNSS observation information. BACKGROUND

[0002] With the maturity of Beidou positioning technology, China's outdoor positioning and navigation field has reached a new height. In the outdoor field, multi-satellite positioning system composed of Beidou can be relied on for joint positioning to achieve high-precision positioning results. However, based on GNSS signals alone, positioning may fail in indoor and semi-occluded areas, which cannot meet the needs of real-time positioning.

[0003] Simultaneous Localization And Mapping (SLAM) technology as a technology that relies only on its own sensors for autonomous positioning has achieved rapid development in the field of robots. This method mainly relies on vision and laser sensors, and visual SLAM mainly based on vision and laser SLAM mainly based on laser radar have emerged. Compared with visual SLAM, laser SLAM is more robust and does not rely on environmental texture information. However, the real-time positioning obtained by SLAM is relative position information, which belongs to a local coordinate system and has cumulative errors. SUMMARY

[0004] To solve the above problems, the present application provides a laser SLAM positioning method and system fusing GNSS observation information, which combines GNSS / IMU with laser SLAM to achieve seamless indoor and outdoor positioning and greatly improve positioning accuracy and robustness.

[0005] To achieve the above purpose, the present application provides the following solutions:

[0006] A laser SLAM positioning method fusing GNSS observation information, comprising:

[0007] acquiring laser radar data, inertial sensor data and GNSS data of a carrier in a target space;

[0008] constructing an inertial sensor data constraint factor based on the inertial sensor data;

[0009] constructing a laser radar constraint factor based on the laser radar data;

[0010] constructing a Doppler data constraint factor, a code pseudorange constraint factor and a clock difference constraint factor based on the GNSS data;

[0011] The inertial sensor data constraint factor, the laser radar constraint factor, the Doppler data constraint factor, the code pseudorange constraint factor, and the clock difference constraint factor are jointly optimized based on a sliding window optimization method to obtain positioning information of the target space carrier.

[0012] Optionally, the inertial sensor data constraint factor is constructed based on the inertial sensor data, and specifically includes:

[0013] The inertial sensor data between the key frame laser radar data collected by the laser radar is pre-integrated by using a pre-integration processing formula to obtain an inertial sensor data constraint factor; the key frame laser radar data is laser radar data in which the motion distance and the rotation angle of the adjacent two frames of laser radar data are greater than the corresponding set threshold; the inertial sensor data constraint factor includes position, velocity, and rotation constraints between the key frames of the laser radar.

[0014] Optionally, the laser radar constraint factor is constructed based on the laser radar data, and specifically includes:

[0015] The laser radar data is de-distorted by using the inertial sensor to obtain de-distorted laser radar data.

[0016] A three-dimensional point cloud map is generated by using the de-distorted laser radar data and the key frame laser radar data.

[0017] The laser radar constraint factor is obtained by using the three-dimensional point cloud map for matching based on a nearest point iteration algorithm.

[0018] Optionally, the laser radar data is de-distorted by using the inertial sensor, and specifically includes:

[0019] The position and the attitude of the laser radar when collecting each frame of the laser radar data are determined by using a timestamp and a motion trajectory of the inertial sensor to obtain coordinate values of each frame of the laser radar data.

[0020] The coordinate values of each frame of the laser radar data are converted to the laser radar coordinate system at the starting time of each frame of the laser radar data to obtain de-distorted laser radar data.

[0021] Optionally, the inertial sensor data constraint factor, the laser radar constraint factor, the Doppler data constraint factor, the code pseudorange constraint factor, and the clock difference constraint factor are jointly optimized based on a sliding window optimization method, and specifically includes:

[0022] A sliding window optimization equation is constructed based on the inertial sensor data constraint factor, the laser radar constraint factor, the Doppler data constraint factor, the code pseudorange constraint factor, and the clock difference constraint factor.

[0023] Solving the sliding window optimization equation by using the L-M algorithm to obtain the positioning information of the carrier in the target space.

[0024] The application further provides a laser SLAM positioning system fusing GNSS observation information, comprising:

[0025] A data acquisition module is configured to acquire lidar data, inertial sensor data and GNSS data of a carrier in a target space.

[0026] An inertial sensor data constraint factor construction module is configured to construct an inertial sensor data constraint factor based on the inertial sensor data.

[0027] A lidar constraint factor construction module is configured to construct a lidar constraint factor based on the lidar data.

[0028] A constraint factor, code pseudorange constraint factor and clock error constraint factor construction module is configured to construct a Doppler data constraint factor, a code pseudorange constraint factor and a clock error constraint factor based on the GNSS data.

[0029] An optimization module is configured to jointly optimize the inertial sensor data constraint factor, the lidar constraint factor, the Doppler data constraint factor, the code pseudorange constraint factor and the clock error constraint factor based on a sliding window optimization method to obtain the positioning information of the carrier in the target space.

[0030] Optionally, the inertial sensor data constraint factor construction module specifically comprises:

[0031] A pre-integration processing unit is configured to perform pre-integration processing on the inertial sensor data between key frame lidar data collected by the lidar by using a pre-integration processing formula to obtain an inertial sensor data constraint factor; the key frame lidar data is lidar data whose motion distance and rotation angle are both greater than a corresponding set threshold; and the inertial sensor data constraint factor comprises position, velocity and rotation constraints between key frame lidar data.

[0032] Optionally, the lidar constraint factor construction module specifically comprises:

[0033] A distortion removal unit is configured to perform distortion removal on the lidar data by using the inertial sensor to obtain distortion-removed lidar data.

[0034] A three-dimensional point cloud map generation unit is configured to generate a three-dimensional point cloud map by using the distortion-removed lidar data and key frame lidar data.

[0035] A matching unit is configured to perform matching based on a nearest point iteration algorithm using the three-dimensional point cloud map to obtain a laser radar constraint factor.

[0036] Optionally, the distortion removing unit specifically comprises:

[0037] A coordinate value determination subunit is configured to determine the position and posture of the laser radar when collecting each frame of the laser radar data by using the time stamp and the motion trajectory of the inertial sensor to obtain the coordinate value of each frame of the laser radar data.

[0038] A coordinate value conversion subunit is configured to convert the coordinate value of each frame of the laser radar data to the laser radar coordinate system at the starting moment of each frame of the laser radar data to obtain the distortion-removed laser radar data.

[0039] Optionally, the optimization module specifically comprises:

[0040] A sliding window optimization equation construction unit is configured to construct a sliding window optimization equation based on the inertial sensor data constraint factor, the laser radar constraint factor, the Doppler data constraint factor, the code pseudorange constraint factor and the clock difference constraint factor.

[0041] A solving subunit is configured to solve the sliding window optimization equation by using an L-M algorithm to obtain the positioning information of the carrier in the target space.

[0042] According to the embodiments of the present application, the following technical effects are provided:

[0043] The laser SLAM positioning method fusing GNSS observation information provided by the present application comprises the following steps: acquiring laser radar data, inertial sensor data and GNSS data of a carrier in a target space; constructing an inertial sensor data constraint factor based on the inertial sensor data; constructing a laser radar constraint factor based on the laser radar data; constructing a Doppler data constraint factor, a code pseudorange constraint factor and a clock difference constraint factor based on the GNSS data; and jointly optimizing the above constraint factors based on a sliding window optimization method to obtain the positioning information of the carrier in the target space. The present application combines GNSS / IMU and laser SLAM by the above method, utilizes the high-precision positioning advantage of laser SLAM in a short time, and provides global coordinates by GNSS without error accumulation. The present application fuses the observation data of GNSS, inertial measurement sensor and laser radar, and realizes real-time high-precision positioning indoors and outdoors seamlessly. BRIEF DESCRIPTION OF DRAWINGS

[0044] In order to make the technical solutions in the embodiments of the present application or the prior art clearer, the accompanying drawings needed in the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description only represent some of the embodiments of the present application, and other drawings can be obtained by those of ordinary skill in the art without any creative effort based on these drawings.

[0045] Figure 1 A flowchart of the laser SLAM positioning method fusing GNSS observation information provided by the present application. DETAILED DESCRIPTION

[0046] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments only represent some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without any creative effort fall within the protection scope of the present application.

[0047] The purpose of the present application is to provide a laser SLAM positioning method fusing GNSS observation information, which combines GNSS / IMU with laser SLAM, can realize seamless positioning indoors and outdoors and greatly improve positioning accuracy and robustness.

[0048] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0049] Embodiment one

[0050] As shown in the accompanying drawings, Figure 1 The present application provides a laser SLAM positioning method fusing GNSS observation information, comprising the following steps:

[0051] Step 101: acquiring laser radar data, inertial sensor data and GNSS data of a carrier in a target space.

[0052] Step 102: constructing an inertial sensor data constraint factor based on the inertial sensor data.

[0053] Step 103: constructing a laser radar constraint factor based on the laser radar data.

[0054] Step 104: constructing a Doppler data constraint factor, a code pseudorange constraint factor and a clock difference constraint factor based on the GNSS data.

[0055] Step 105: jointly optimizing the inertial sensor data constraint factor, the lidar constraint factor, the Doppler data constraint factor, the code pseudo-range constraint factor and the clock bias constraint factor based on the sliding window optimization method to obtain the positioning information of the carrier in the target space.

[0056] In the step 101, the lidar data is collected by the laser radar (3D Lidar) on the carrier; the inertial sensor data is collected by the inertial sensor (IMU); and the GNSS data is the pseudo-range and Doppler signal received by the GNSS receiver.

[0057] In the step 102, the step 102 specifically includes:

[0058] The inertial sensor data between the key frame lidar data collected by the lidar is pre-integrated by using the formula of pre-integration processing to obtain the inertial sensor data constraint factor; the key frame lidar data is the laser radar data whose motion distance and rotation angle between two adjacent frames are greater than the corresponding set threshold; and the inertial sensor data constraint factor includes the position, velocity and rotation constraint between the key frame lidar.

[0059] In the embodiment, in order to reduce the calculation amount and computer storage space, only the laser key frame data is stored; sometimes the 3D Lidar does not move or the motion distance is very small, which will cause data redundancy and unnecessary calculation. Therefore, in the following pose optimization and local mapping process, only the laser Lidar key frame data is processed.

[0060] The laser Lidar key frame is determined according to the distance information and the rotation angle size, and the motion exceeding 5m or the rotation exceeding 10° will be considered as the key frame.

[0061] The scanning frequency of the laser Lidar is between 10-30HZ, and the frequency of the IMU receiving data is between 100-500. In order to prevent repeated integration and save calculation amount, the IMU acceleration data and angular velocity data between the radar key frame data need to be integrated, as shown in formula (1), to obtain the position, velocity and rotation constraint between the laser Lidar key frame, that is, the IMU constraint factor (inertial sensor data constraint factor). The formula of pre-integration processing is as follows:

[0062]

[0063] represents the acceleration and angular velocity measurement value measured by the IMU, i and i+1 represent two time points of the IMU measurement data between the kth key frame and the k+1th key frame, and δt represents the time interval between the two IMU data, respectively represent the acceleration zero offset and the angular velocity zero offset. respectively represent the position, velocity and quaternion representation of the pose in the coordinate system at time i based on key frame b k position, velocity and quaternion representation of the pose in the coordinate system at time i+1 based on key frame b respectively represent the position, velocity and quaternion representation of the pose in the coordinate system at time i based on key frame b k position, velocity and quaternion representation of the pose in the coordinate system at time i+1 based on key frame b is a three-dimensional zero vector, is a unit quaternion, denotes the conversion from quaternion form to rotation matrix. The measurement model constructed by the pre-integration term, i.e. the calculation formula of the inertial sensor data constraint factor, is shown in equation (2), where g w denotes the gravity vector g w in the world coordinate system. T .

[0064]

[0065] wherein, denote the position, velocity and angular pre-integration increment between key frame b k and key frame b k+1 ; is a unit quaternion; g w denotes the gravity vector g w in the world coordinate system. T , g denotes the gravitational acceleration; denotes the rotation matrix from the world coordinate system to key frame b k ; denotes the position of key frame b k in the world coordinate system; denotes the position of key frame b k+1 in the world coordinate system; Δt k denotes the time difference between key frame b k and key frame b k+1 ; denotes the velocity of key frame b k in the world coordinate system; denotes the inverse matrix of , denotes the quaternion of the rotation from the world coordinate system to key frame b k ; denotes the quaternion of the rotation from key frame b k+1 to the world coordinate system; denotes the inverse matrix of ; denotes the acceleration bias of the k-th frame; denotes the angular velocity bias of the k-th frame; acceleration bias of the k+1th frame; angular velocity bias of the k+1th frame; inertial sensor data constraint factor is the data after pre-integration processing.

[0066] The step 103 specifically comprises:

[0067] Step 1031: using the inertial sensor to de-warp the lidar data to obtain de-warping lidar data.

[0068] 1) Adopting the time stamp and the motion trajectory of the inertial sensor to determine the position and posture of the lidar when collecting each frame of lidar data, and obtaining the coordinate value of each frame of lidar data.

[0069] 2) According to the time interval of each frame of lidar data and the motion posture of the process, performing coordinate interpolation to calculate the carrier posture corresponding to the time of each laser point cloud, so as to convert the coordinate value of each frame of lidar data to the lidar coordinate system at the starting moment of the frame of lidar data through the external parameter, and obtain the de-warping lidar data. Steps 1) and 2) realize the de-warping of the lidar data by using the inertial sensor.

[0070] Step 1032: generating a three-dimensional point cloud map by using the de-warping lidar data and the key frame lidar data.

[0071] Step 1033: using the three-dimensional point cloud map to perform matching based on the iterative closest point algorithm to obtain the lidar constraint factor.

[0072] The steps 1032-1033 specifically comprise:

[0073] Using the de-warping lidar data to extract line features and surface features, and performing pose transformation (R, T) based on distance minimization (lidar constraint factor) according to the line features and surface features between adjacent key frames. Using the pose matching of adjacent key frames as the initial value, and then matching with the submap to obtain a better pose result, that is, the lidar constraint factor. The set of line features and surface features of continuous several key frames constitutes a local submap (submap), which is used for scan-submap matching with the features extracted from the current frame point cloud.

[0074] The step 104 specifically comprises:

[0075] 1) Construction of Doppler data constraint factor

[0076] The Doppler shift is measured from the difference between the received carrier signal and the design signal, which reflects the relative motion of the receiver-satellite along the signal propagation path. Due to the characteristics of GNSS signal structure, the accuracy of Doppler measurement is usually one order of magnitude higher than that of code pseudorange. The Doppler shift is modeled as:

[0077]

[0078] The above formula and respectively represent the coordinates of the receiver and the satellite in the ECI coordinate frame, and λ represents the wavelength of the carrier signal. is the unit vector from the receiver to the satellite in the ECI frame, represents the drift rate of the satellite clock error, which can be obtained from the satellite navigation information, and c represents the speed of light, represents the Doppler measurement noise.

[0079] Based on the modeling equation, t k the system state at time t j The residual error related to the Doppler measurement in satellite s can be expressed as where n s represents the broadcast satellite spatial accuracy index, n dp is the receiver measurement noise index, and θ el represents the satellite elevation angle in the receiver's view.

[0080] 2) Construction of code pseudorange constraint factor

[0081] The code pseudorange measurement modeling formula can be expressed as:

[0082]

[0083] In the formula, and are the ECI coordinates of satellite s and receiver r, respectively; ζ s is a 4x1 index vector, with the corresponding satellite constellation entity being 1 and the other three entities being 0; Δt s is the satellite clock error, which can be obtained from the broadcast navigation message; and represent the tropospheric and ionospheric delays, respectively; represents the delay caused by multipath error; represents the measurement noise, which is assumed to follow a zero-mean Gaussian distribution The modeling is as follows:

[0084]

[0085] n s denotes the broadcast satellite spatial precision index; n pr denotes the receiver reported code pseudorange measurement noise index; θ el denotes the satellite elevation angle on the receiver view.

[0086] Due to the time difference between the satellite code pseudorange signal transmission and arrival time, represents the satellite position at the signal arrival time in the ECEF coordinate frame, represents the satellite position at the signal transmission time in the ECEF coordinate frame, and the two have the following transformation relationship:

[0087]

[0088] R z (θ) denotes the rotation around the z-axis of the ECI by an angle θ, ω E denotes the Earth rotation angular velocity, t f is the GNSS signal propagation time. Finally, at the time t k , the system state and the code pseudorange residual measured on the satellite s j can be expressed as:

[0089] where r k denotes the GNSS receiver at the time t k .

[0090] 3) Clock bias constraint factor

[0091] The receiver clock bias at t k and t k-1 has the following relationship:

[0092]

[0093] where I 4×1 denotes a 4x1 matrix with elements of 1, and the residual form in the discrete case is expressed as follows:

[0094]

[0095] In the formula, is the time difference between the measurement time k and k-1.

[0096] The receiver clock drift rate is modeled as a random walk process, and its residual form is as follows:

[0097]

[0098] The step 105 specifically comprises: constructing a sliding window optimization equation based on the inertial sensor data constraint factor, the laser radar constraint factor, the Doppler data constraint factor, the code pseudo-range constraint factor and the clock difference constraint factor; and solving the sliding window optimization equation by using an L-M algorithm to obtain the positioning information of the carrier in the target space.

[0099] The system state to be estimated comprises: a position of the carrier relative to a local world coordinate system attitude velocity acceleration bias b a , magnetometer bias b w . The local world coordinate system and the heading angle ψ of the East-North-Sky (ENU) coordinate system, the receiver clock difference δt, and the receiver clock difference drift rate

[0100] The method for optimizing and solving the constraint factors by using the sliding window-based graph optimization method in the application can summarize the states χ in the window as follows:

[0101] χ = [x 0, x 1,... x n ] n , ψ]

[0102]

[0103] δt = [δt 0, δt 1,... δt n ] G , δt n-1, δt n ] R , δt n-1, δt n ] E , δt n-1, δt n ] C

[0104] wherein n is the window size. The four components in δt correspond to: the clock bias of the receiver relative to GPS, GLONASS, GALILEO and Beidou.

[0105] ​When the system transmits a new observation of a moment, a new constraint factor is constructed, and more system state variables are generated. However, as time goes on, the number of system state variables increases, and the continuous increase in the calculation burden will lead to the decline of the performance of the entire combination algorithm. In order to effectively reduce the calculation amount of the system, the application adopts a sliding window-based factor graph optimization. In the optimization process, only the state variables most relevant to the latest state of the system are stored in the window, and the previous historical state is marginalized to construct a prior constraint term for the state in the window to participate in the optimization of the subsequent sliding window, thereby maximizing the preservation of the original observation information. In the application, the window size of the sliding window is set to 10, that is, only the latest 10 frames of system state are retained in the window. When the number of state variables in the window reaches the window size, if a new frame of observation arrives, the "oldest frame" in the window needs to be marginalized, and the constraint term generated after marginalization is added to the prior constraint term. For the marginalization of the historical state, the application adopts the Schur complement method to construct the marginalization term. In addition, the L-M algorithm is used to optimize and solve the nonlinear equation set constructed by the constraint factor, to obtain real-time high-precision position, attitude, velocity, time and other information.

[0106] Embodiment two

[0107] In order to perform the method corresponding to the above-mentioned embodiment one, to realize the corresponding functions and technical effects, the following provides a laser SLAM positioning system fusing GNSS observation information, comprising:

[0108] A data acquisition module is configured to acquire lidar data, inertial sensor data and GNSS data of a carrier in a target space.

[0109] An inertial sensor data constraint factor construction module is configured to construct an inertial sensor data constraint factor based on the inertial sensor data.

[0110] A lidar constraint factor construction module is configured to construct a lidar constraint factor based on the lidar data.

[0111] A constraint factor, code pseudorange constraint factor and clock error constraint factor construction module is configured to construct a Doppler data constraint factor, a code pseudorange constraint factor and a clock error constraint factor based on the GNSS data.

[0112] An optimization module is configured to jointly optimize the inertial sensor data constraint factor, the lidar constraint factor, the Doppler data constraint factor, the code pseudorange constraint factor and the clock error constraint factor based on a sliding window optimization method, to obtain positioning information of the carrier in the target space.

[0113] The inertial sensor data constraint factor construction module specifically comprises:

[0114] The pre-integration processing unit is configured to perform pre-integration processing on the inertial sensor data between the key frame lidar data collected by the lidar by using a formula of pre-integration processing to obtain an inertial sensor data constraint factor, wherein the key frame lidar data is lidar data whose motion distance and rotation angle between two adjacent frames are greater than corresponding set thresholds, and the inertial sensor data constraint factor includes position, velocity and rotation constraints between the key frame lidar data.

[0115] The laser radar constraint factor construction module specifically includes:

[0116] The de-distortion unit is configured to perform de-distortion on the lidar data by using the inertial sensor to obtain de-distorted lidar data.

[0117] The three-dimensional point cloud map generation unit is configured to generate a three-dimensional point cloud map by using the de-distorted lidar data and the key frame lidar data.

[0118] The matching unit is configured to perform matching based on a nearest point iteration algorithm by using the three-dimensional point cloud map to obtain the laser radar constraint factor.

[0119] The de-distortion unit specifically includes:

[0120] The coordinate value determination sub-unit is configured to determine the position and attitude of the laser radar when collecting each frame of lidar data by using a time stamp and a motion trajectory of the inertial sensor to obtain a coordinate value of each frame of lidar data.

[0121] The coordinate value conversion sub-unit is configured to convert the coordinate value of each frame of lidar data to the coordinate system of the laser radar at the starting moment of each frame of lidar data to obtain the de-distorted lidar data.

[0122] The optimization module specifically includes:

[0123] The sliding window optimization equation construction unit is configured to construct a sliding window optimization equation based on the inertial sensor data constraint factor, the laser radar constraint factor, the Doppler data constraint factor, the code pseudorange constraint factor and the clock bias constraint factor.

[0124] The solving sub-unit is configured to solve the sliding window optimization equation by using an L-M algorithm to obtain the positioning information of the carrier in the target space.

[0125] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts of each embodiment can be referred to each other. For the system disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the related parts can be referred to the method part.

[0126] The principles and implementation manners of the present application are described by using specific examples in the present application, and the above examples are only used to help understand the method of the present application and its core idea; meanwhile, for the general technical personnel in the art, the specific implementation manners and application range will be changed according to the idea of the present application. In conclusion, the content of the present specification should not be understood as the limitation of the present application.

Claims

1. A method of laser SLAM localization fusing GNSS observation information, characterized in that, The method comprises the following steps: acquiring laser radar data, inertial sensor data and GNSS data of a carrier in a target space; constructing an inertial sensor data constraint factor based on the inertial sensor data; constructing a laser radar constraint factor based on the laser radar data; constructing a Doppler data constraint factor, a code pseudorange constraint factor and a clock difference constraint factor based on the GNSS data; jointly optimizing the inertial sensor data constraint factor, the laser radar constraint factor, the Doppler data constraint factor, the code pseudorange constraint factor and the clock difference constraint factor based on a sliding window optimization method to obtain positioning information of the carrier in the target space. 2.The fusion GNSS observation information and laser SLAM positioning method according to claim 1, wherein, The method of constructing an inertial sensor data constraint factor based on the inertial sensor data comprises the following steps: performing pre-integration processing on the inertial sensor data between key frame laser radar data collected by the laser radar by using a pre-integration processing formula to obtain an inertial sensor data constraint factor; the key frame laser radar data are laser radar data whose motion distance and rotation angle are greater than corresponding set thresholds; the inertial sensor data constraint factor comprises position, speed and rotation constraints between key frame laser radar data. 3.The fusion GNSS observation information and laser SLAM positioning method according to claim 1, wherein, The method of constructing a laser radar constraint factor based on the laser radar data comprises the following steps: de-distorting the laser radar data by using the inertial sensor to obtain de-distorted laser radar data; generating a three-dimensional point cloud map by using the de-distorted laser radar data and key frame laser radar data; performing matching based on a nearest point iteration algorithm by using the three-dimensional point cloud map to obtain a laser radar constraint factor. 4.The fusion GNSS observation information and laser SLAM positioning method according to claim 3, wherein, The method of de-distorting the laser radar data by using the inertial sensor comprises the following steps: determining the position and posture of the laser radar when collecting each frame of the laser radar data by using a timestamp and the motion trajectory of the inertial sensor to obtain coordinate values of each frame of the laser radar data; converting the coordinate values of each frame of the laser radar data to the coordinate system of the laser radar at the starting moment of each frame of the laser radar data to obtain de-distorted laser radar data. 5.The fusion GNSS observation information laser SLAM positioning method according to claim 1, wherein, The method of jointly optimizing the inertial sensor data constraint factor, the laser radar constraint factor, the Doppler data constraint factor, the code pseudorange constraint factor and the clock difference constraint factor based on a sliding window optimization method comprises the following steps: constructing a sliding window optimization equation based on the inertial sensor data constraint factor, the laser radar constraint factor, the Doppler data constraint factor, the code pseudorange constraint factor and the clock difference constraint factor; solving the sliding window optimization equation by using an L-M algorithm to obtain the positioning information of the carrier in the target space.

6. A laser SLAM positioning system fusing GNSS observation information, characterized in that, The method comprises the following steps: a data acquisition module is configured to acquire laser radar data, inertial sensor data and GNSS data of a carrier in a target space; an inertial sensor data constraint factor construction module is configured to construct an inertial sensor data constraint factor based on the inertial sensor data; a laser radar constraint factor construction module is configured to construct a laser radar constraint factor based on the laser radar data; A constraint factor constructing module, configured to construct Doppler data constraint factors, code pseudo-range constraint factors and clock difference constraint factors based on the GNSS data; An optimization module, configured to jointly optimize the inertial sensor data constraint factors, the lidar constraint factors, the Doppler data constraint factors, the code pseudo-range constraint factors and the clock difference constraint factors based on a sliding window optimization method, to obtain the positioning information of the carrier in the target space.

7. The laser SLAM positioning system fusing GNSS observation information according to claim 6, characterized in that, The inertial sensor data constraint factor constructing module specifically comprises: A pre-integration processing unit, configured to perform pre-integration processing on the inertial sensor data between the key frame lidar data collected by the lidar by using a pre-integration processing formula, to obtain inertial sensor data constraint factors; the key frame lidar data are lidar data whose motion distance and rotation angle are both greater than a corresponding set threshold; the inertial sensor data constraint factors comprise position, velocity and rotation constraints between key frames of the lidar.

8. The laser SLAM positioning system fusing GNSS observation information according to claim 6, characterized in that, The lidar constraint factor constructing module specifically comprises: A de-distortion unit, configured to perform de-distortion on the lidar data by using the inertial sensor, to obtain de-distorted lidar data; A three-dimensional point cloud map generating unit, configured to generate a three-dimensional point cloud map by using the de-distorted lidar data and the key frame lidar data; A matching unit, configured to perform matching based on a nearest point iteration algorithm by using the three-dimensional point cloud map, to obtain lidar constraint factors.

9. The laser SLAM positioning system fusing GNSS observation information according to claim 8, characterized in that, The de-distortion unit specifically comprises: A coordinate value determining subunit, configured to determine the position and posture of the lidar when each frame of the lidar data is collected by using a time stamp and the motion trajectory of the inertial sensor, to obtain the coordinate value of each frame of the lidar data; A coordinate value converting subunit, configured to convert the coordinate value of each frame of the lidar data to the coordinate system of the lidar at the starting moment of each frame of the lidar data, to obtain de-distorted lidar data.

10. The laser SLAM positioning system fusing GNSS observation information according to claim 6, characterized in that, The optimization module specifically comprises: A sliding window optimization equation constructing unit, configured to construct a sliding window optimization equation based on the inertial sensor data constraint factors, the lidar constraint factors, the Doppler data constraint factors, the code pseudo-range constraint factors and the clock difference constraint factors; A solving subunit, configured to solve the sliding window optimization equation by using an L-M algorithm, to obtain the positioning information of the carrier in the target space.

Images