A fusion positioning method, vehicle, storage medium and program product

Abstract

The present application proposes a fusion positioning method, a vehicle, a storage medium and a program product. The method may include: obtaining a first positioning result determined for a target vehicle by a first positioning sensor, and obtaining a second positioning result determined for a target vehicle by a second positioning sensor, wherein the first positioning sensor and the second positioning sensor are of different types; extracting target positioning information in a target positioning direction based on the first positioning result, wherein the target positioning direction is determined based on the type of the first positioning sensor, and the target positioning direction is a partial positioning direction of all positioning directions corresponding to the first positioning result; fusing the target positioning information with the second positioning result to obtain a fusion positioning result of the target vehicle. In this way, it is possible to selectively fuse a more accurate portion of the first positioning result with the second positioning result, thereby improving the accuracy of the fusion positioning result.

Description

Technical Field

[0001] One or more embodiments of the present application relate to the field of positioning technology, and in particular, to a fusion positioning method, a vehicle, a storage medium, and a program product. Background Art

[0002] With the continuous development of autonomous driving technology, the requirements for positioning accuracy are getting higher and higher. How to obtain accurate positioning of target objects has become a problem that needs to be solved urgently.

[0003] At present, in order to improve positioning accuracy, positioning results of multiple types of positioning sensors such as visual positioning sensors, inertial navigation positioning sensors and navigation positioning sensors are usually obtained, and the positioning results of these multiple types of positioning sensors are fused to obtain a fused positioning result.

[0004] However, in practical applications, it is found that the fusion positioning results are not accurate enough. Summary of the invention

[0005] The present application provides a fusion positioning method, device, vehicle, storage medium and program product, which improve the accuracy of fusion positioning results.

[0006] According to a first aspect of one or more embodiments of the present application, a fusion positioning method is provided, the method comprising:

[0007] Obtaining a first positioning result determined for a target vehicle by a first positioning sensor, and obtaining a second positioning result determined for the target vehicle by a second positioning sensor, wherein the first positioning sensor and the second positioning sensor are of different types;

[0008] Extracting target positioning information in a target positioning direction based on the first positioning result, where the target positioning direction is determined based on the type of the first positioning sensor, and the target positioning direction is a part of all positioning directions corresponding to the first positioning result;

[0009] The target positioning information is fused with the second positioning result to obtain a fused positioning result of the target vehicle.

[0010] Optionally, all positioning directions corresponding to the first positioning result include directions respectively indicated by at least two coordinate axes constituting an N-dimensional coordinate system, where N is an integer not less than 2;

[0011] or,

[0012] The first positioning sensor is a visual positioning sensor, all positioning directions corresponding to the first positioning result include horizontal, vertical and elevation directions, and the target positioning direction is the horizontal direction;

[0013] The horizontal direction is the direction indicated by the X-axis constituting the three-dimensional coordinate system, the vertical direction is the direction indicated by the Y-axis constituting the three-dimensional coordinate system, and the elevation direction is the direction indicated by the Z-axis constituting the three-dimensional coordinate system.

[0014] Optionally, extracting target positioning information in the target positioning direction based on the first positioning result includes:

[0015] Predicting an overall error of the second positioning result in the target positioning direction based on the first positioning result, the overall error being used to represent the difference between the second positioning result and the actual positioning result of the target vehicle in the target positioning direction; determining the target positioning information based on the position error in the overall error; or

[0016] The target positioning information is determined based on the positioning result in the target positioning direction in the first positioning result.

[0017] Optionally, the obtaining of a first positioning result determined by a first positioning sensor for the target vehicle further includes:

[0018] Determine the dead position positioning result of the second positioning sensor at the current moment, and determine the dead position positioning result of the second positioning sensor at a historical moment;

[0019] Based on the change of the dead position positioning result at the current moment and the dead position positioning results at the historical moments, a first positioning result of the first positioning sensor is predicted.

[0020] Optionally, the dead position positioning result is obtained by:

[0021] determining a second attitude misalignment angle based on a partial misalignment angle in the first attitude misalignment angle of the second positioning sensor, wherein the partial misalignment angle is determined based on a type of the first positioning sensor;

[0022] Dead reckoning is performed based on the positioning data of the second positioning sensor and the second attitude misalignment angle to obtain the dead position positioning result.

[0023] Optionally, the first positioning sensor is a visual sensor, and the partial misalignment angle is a horizontal misalignment angle in the first posture misalignment angle.

[0024] Optionally, performing dead reckoning based on the positioning data of the second positioning sensor and the second attitude misalignment angle to obtain the dead position positioning result includes:

[0025] The second positioning sensor includes an inertial navigation positioning sensor, and dead reckoning is performed based on positioning data of the inertial navigation positioning sensor to obtain an initial dead position positioning result; based on the second attitude misalignment angle, the initial dead position positioning result is compensated to obtain a compensated dead position positioning result;

[0026] or,

[0027] The second positioning sensor includes an inertial navigation positioning sensor and a third positioning sensor. Dead reckoning is performed based on the positioning data of the inertial navigation positioning sensor to obtain an initial attitude. The initial attitude is compensated based on the second attitude misalignment angle to obtain the attitude in the dead position positioning result. The positioning data of the third positioning sensor is compensated based on the second attitude misalignment angle. Dead reckoning is performed based on the compensated positioning data to obtain the position in the dead position positioning result.

[0028] Optionally, the initial posture is the posture of the target vehicle in the carrier coordinate system; and the compensating the initial posture based on the second posture misalignment angle to obtain the posture in the dead position positioning result includes:

[0029] Based on the initial posture and the second posture misalignment angle, determining a first conversion relationship, wherein the first conversion relationship is used to represent a conversion relationship between a navigation coordinate system of the target vehicle and a carrier coordinate system;

[0030] The initial posture is transformed using the first transformation relationship to obtain the posture of the target vehicle in the navigation coordinate system.

[0031] Optionally, the first conversion relationship is represented as a posture conversion matrix, and the determining the first conversion relationship based on the initial posture and the second posture misalignment angle includes:

[0032] Based on the initial posture, a posture conversion matrix is ​​generated; based on the second posture misalignment angle, the posture conversion matrix is ​​compensated to obtain a compensated posture conversion matrix; or,

[0033] The initial posture is compensated based on the second posture misalignment angle to obtain a compensated posture; and a posture conversion matrix is ​​generated based on the compensated posture.

[0034] According to a second aspect of one or more embodiments of the present application, a fusion positioning device is provided, the device comprising:

[0035] an acquisition module, configured to acquire a first positioning result determined for a target vehicle by a first positioning sensor, and to acquire a second positioning result determined for the target vehicle by a second positioning sensor, wherein the first positioning sensor and the second positioning sensor are of different types;

[0036] an extraction module, configured to extract target positioning information in a target positioning direction based on the first positioning result, wherein the target positioning direction is determined based on a type of the first positioning sensor, and the target positioning direction is a part of all positioning directions corresponding to the first positioning result;

[0037] The fusion module is used to fuse the target positioning information with the second positioning result to obtain a fused positioning result of the target vehicle.

[0038] According to a third aspect of one or more embodiments of the present application, a vehicle is provided, comprising:

[0039] processor;

[0040] a memory for storing processor-executable instructions;

[0041] The processor implements the method described in the embodiment of the first aspect above by running the executable instructions.

[0042] According to a fourth aspect of one or more embodiments of the present application, a computer-readable storage medium is provided, on which computer instructions are stored. When the instructions are executed by a processor, the steps of the method described in the embodiment of the first aspect are implemented.

[0043] According to a fifth aspect of one or more embodiments of the present application, a computer-readable storage medium is provided, on which computer instructions are stored. When the instructions are executed by a processor, the steps of the method described in the embodiment of the first aspect are implemented.

[0044] It can be seen from the above technical solutions that in one or more embodiments of the present application, taking into account the different positioning accuracies of different types of positioning sensors in different positioning directions, when performing fusion positioning, the target positioning direction will be determined based on the type of the first positioning sensor, and only the target positioning information of the first positioning result in the target positioning direction will be fused with the second positioning result, while the positioning information of the first positioning result in other positioning directions will not be fused with the second positioning result. This can achieve the selective fusion of the more accurate information in the target positioning direction in the first positioning result with the second positioning result, which can not only improve the accuracy of the fused positioning result in the target positioning direction, but also reduce the negative impact of some less accurate information in the first positioning result on the accuracy of the fused positioning result, thereby improving the accuracy of the fused positioning result in each positioning direction.

[0045] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0047] Figure 1 It is a flowchart of a fusion positioning method provided by an exemplary embodiment.

[0048] Figure 2 It is a schematic diagram of a positioning direction provided by an exemplary embodiment.

[0049] Figure 3 It is a flowchart of another fusion positioning method provided by an exemplary embodiment.

[0050] Figure 4 It is a flowchart of another fusion positioning method provided by an exemplary embodiment.

[0051] Figure 5 It is a flowchart of another fusion positioning method provided by an exemplary embodiment.

[0052] Figure 6 It is a flowchart of a method for predicting a first positioning result provided by an exemplary embodiment.

[0053] Figure 7 It is a flowchart of a dead reckoning method provided by an exemplary embodiment.

[0054] Figure 8 It is a flowchart of a fusion positioning method provided by an exemplary embodiment.

[0055] Fig. 9 It is a structural block diagram of a vehicle provided by an exemplary embodiment.

[0056] Fig.10 It is a structural block diagram of a fusion positioning device provided by an exemplary embodiment.

[0057] Fig.11 It is a structural block diagram of another fusion positioning device provided by an exemplary embodiment. DETAILED DESCRIPTION

[0058] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0059] It should be noted that: in other embodiments, the steps of the corresponding method are not necessarily performed in the order shown and described in this application. In some other embodiments, the steps included in the method may be more or less than those described in this application. In addition, a single step described in this application may be decomposed into multiple steps for description in other embodiments; and multiple steps described in this application may also be combined into a single step for description in other embodiments.

[0060] Next, one or more embodiments of the present application are described in detail.

[0061] Figure 1 A flowchart of a fusion positioning method provided as an exemplary embodiment. The execution subject of the method may be a vehicle end, which may include but is not limited to any one or more vehicle-mounted modules such as a vehicle, a processing unit deployed on the vehicle, an intelligent driving unit or a positioning unit; the execution subject of the method may also be a service end that communicates with the vehicle, which may be a service platform, an independent terminal device, an Internet of Things end, an Internet of Vehicles end, a distributed system end, etc. The service platform may be an independent server or a server cluster, etc.; the execution subject of the method may also be completed by the vehicle end and the service end. For the description of the vehicle end and the service end, please refer to the previous content and will not be repeated here; the embodiments of this application do not limit the execution subject. Figure 1 As shown, the method may include the following steps:

[0062] S101. Obtain a first positioning result determined for a target vehicle by a first positioning sensor, and obtain a second positioning result determined for the target vehicle by a second positioning sensor, where the first positioning sensor and the second positioning sensor are of different types.

[0063] In an embodiment of the present application, different types of first positioning sensors and second positioning sensors are provided on the target vehicle, and a fused positioning result of the target vehicle is obtained by fusing the positioning result of the first positioning sensor and the positioning result of the second positioning sensor.

[0064] Among them, the first positioning sensor and the second positioning sensor can be any type of positioning sensor, for example, a visual positioning sensor, an inertial positioning sensor, a navigation positioning sensor, an odometer (Odom), a speedometer, etc., and can also be a new type of positioning sensor developed in the future. It only needs to ensure that the first positioning sensor and the second positioning sensor are of different types.

[0065] Among them, the visual positioning sensor is a visual sensor, which is an instrument that uses optical elements and imaging devices to obtain image information of the external environment. The visual positioning sensor in the embodiment of the present application may include but is not limited to any one or any combination of cameras, video recording equipment, lidar, monocular vision sensors, multi-eye vision sensors, etc. The embodiment of the present application does not limit the visual positioning sensor; that is, in the embodiment of the present application, the visual positioning sensor can analyze the semantic information (lane lines, etc.) of the vehicle's environment based on the information it collects to locate the vehicle. Therefore, the visual positioning sensor can also be called a semantic positioning sensor.

[0066] The first positioning sub-result is a positioning result determined by a visual positioning sensor. In one embodiment, the first positioning sub-result is data output by the visual positioning sensor. In one embodiment, the first positioning sub-result may include the position and / or attitude of the vehicle. For example, the first positioning sub-result includes the position and attitude of the vehicle, and the position may include but is not limited to the longitude and latitude of the vehicle (i.e., the longitude and latitude of the vehicle's location in the physical world) and the altitude (i.e., the vertical distance of the vehicle's location relative to the horizontal plane or sea level in the physical world), and the attitude may include but is not limited to the heading of the vehicle (i.e., the direction of travel of the vehicle). It should be noted that the embodiment of the present application is only an exemplary description of the content of the first positioning sub-result, and does not limit the content of the first positioning sub-result.

[0067] The inertial navigation positioning sensor is a positioning sensor based on IMU (Inertial Measurement Unit) for positioning. The inertial navigation positioning sensor can be an IMU, an INS (Inertial Navigation System), or a SINS (Strapdown Inertial Navigation System), etc. The embodiments of the present application do not limit the inertial navigation positioning sensor.

[0068] The navigation and positioning sensor is a GNSS (Global Navigation Satellite System) sensor. The GNSS sensor may be a GPS (Global Positioning System) sensor, a BDS (BeiDou Navigation Satellite System) sensor, a Galileo (Galileo Satellite Navigation System) sensor, etc. The embodiment of the present application does not limit the GNSS sensor.

[0069] It should be noted that the first positioning sensor is a type of positioning sensor, and the second positioning sensor can be a type of positioning sensor, or can include at least two types of positioning sensors. For example, the first positioning sensor is a visual positioning sensor, and the second positioning sensor includes an inertial positioning sensor and a navigation positioning sensor; for another example, the first positioning sensor is a visual positioning sensor, and the second positioning sensor includes an inertial positioning sensor and a wheel speed meter.

[0070] When the second positioning sensor includes at least two types of positioning sensors, the type of the first positioning sensor is different from the type of any positioning sensor in the second positioning sensor.

[0071] The first positioning result determined by the first positioning sensor for the target vehicle is a positioning result obtained by processing the data collected by the first positioning sensor during the positioning process. The first positioning result may be output by the first positioning sensor, that is, the first positioning sensor not only collects data but also processes the collected data; the first positioning result may also be output by other processing units, that is, the first positioning sensor only collects data, transmits the collected data to other processing units, and the other processing units process the collected data to obtain the positioning result. In one embodiment, the first positioning result may include the position and / or posture of the target vehicle. It should be noted that the embodiment of the present application is only an exemplary description of the content of the first positioning result, and does not limit the content of the first positioning result.

[0072] The second positioning result determined by the second positioning sensor for the target vehicle is a positioning result obtained by processing the data collected by the second positioning sensor during the positioning process. The second positioning result can be output by the second positioning sensor, that is, the second positioning sensor not only collects data but also processes the collected data; the second positioning result can also be output by other processing units, that is, the second positioning sensor only collects data, transmits the collected data to other processing units, and the other processing units process the collected data to obtain the positioning result. In one embodiment, the second positioning result may include the position and / or attitude of the target vehicle. For example, the position includes the longitude and latitude and altitude of the target vehicle, and the attitude includes the pitch angle, roll angle and heading angle of the target vehicle, then the second positioning result includes the longitude and latitude, altitude, pitch angle, roll angle and heading angle of the target vehicle. It should be noted that the embodiment of the present application is only an exemplary description of the content of the second positioning result, and does not limit the content of the second positioning result.

[0073] S102: Extract target positioning information in a target positioning direction based on the first positioning result, where the target positioning direction is determined based on a type of the first positioning sensor, and the target positioning direction is a part of all positioning directions corresponding to the first positioning result.

[0074] The first positioning result includes positioning sub-results of the target vehicle in multiple positioning directions. In one embodiment, all positioning directions corresponding to the first positioning result include directions respectively indicated by at least two coordinate axes constituting an N-dimensional coordinate system, where N is an integer not less than 2, for example, N is 2, 3, etc. For example, all positioning directions corresponding to the first positioning result include directions respectively indicated by the X-axis and the Y-axis constituting the two-dimensional coordinate system; for another example, all positioning directions corresponding to the first positioning include directions respectively indicated by the X-axis and the Y-axis constituting the three-dimensional coordinate system; for another example, all positioning directions corresponding to the first positioning include directions respectively indicated by the X-axis, the Y-axis, and the Z-axis constituting the three-dimensional coordinate system; wherein, the direction indicated by the X-axis constituting the three-dimensional coordinate system can also be referred to as the horizontal direction, the direction indicated by the Y-axis constituting the three-dimensional coordinate system can also be referred to as the vertical direction, and the direction indicated by the Z-axis constituting the three-dimensional coordinate system can also be referred to as the elevation direction.

[0075] In another embodiment, all positioning directions corresponding to the first positioning result may include the directions indicated by at least two coordinate axes constituting the N-dimensional coordinate system, and may also include the directions in which the target vehicle rotates around any one of the coordinate axes constituting the N-dimensional coordinate system, or the directions in which the target vehicle rotates around at least two coordinate axes constituting the N-dimensional coordinate system.

[0076] Exemplarily, all positioning directions corresponding to the first positioning result include at least one of a direction of rotation around the X-axis constituting the three-dimensional coordinate system, a direction of rotation around the Y-axis constituting the three-dimensional coordinate system, and a direction of rotation around the Z-axis constituting the three-dimensional coordinate system. The direction of rotation around the X-axis constituting the three-dimensional coordinate system can be called a pitch direction, the direction of rotation around the Y-axis constituting the three-dimensional coordinate system can be called a roll direction, and the direction of rotation around the Z-axis constituting the three-dimensional coordinate system can be called a heading. Figure 2 As shown, Figure 2 The lateral, longitudinal, elevation, roll, pitch and heading directions in the positioning directions are shown in FIG.

[0077] The positioning result in the first positioning result in the direction in which the target vehicle rotates around any one of the coordinate axes constituting the N-dimensional coordinate system is the angle of rotation of the target vehicle around the coordinate axis. The rotation angle can be represented based on the angle between the corresponding axis in the vehicle coordinate system and the corresponding plane in the N-dimensional coordinate system.

[0078] Exemplarily, the N-dimensional coordinate system is a geographic coordinate system. When the target vehicle rotates around the Y axis of the geographic coordinate system, the rotation angle is the angle between the X axis of the vehicle coordinate system and the XY plane of the geographic coordinate system, which can be called the roll angle, wherein the XY plane of the geographic coordinate system is a plane composed of the X axis and the Y axis of the geographic coordinate system; when the target vehicle rotates around the X axis of the geographic coordinate system, the rotation angle is the angle between the Y axis of the vehicle coordinate system and the XY plane of the geographic coordinate system, which can be called the pitch angle; when the target vehicle rotates around the Z axis of the geographic coordinate system, the rotation angle is the angle between the Z axis of the vehicle coordinate system and the YZ plane of the geographic coordinate system, which can be called the heading angle, wherein the YZ plane of the geographic coordinate system is a plane composed of the Y axis and the Z axis of the geographic coordinate system.

[0079] Among them, the vehicle coordinate system is a three-dimensional coordinate system with the center of mass of the target vehicle as the origin. The direction indicated by the X-axis of the vehicle coordinate system is the width direction of the target vehicle, such as the direction from the left side of the target vehicle to the right side of the target vehicle; the direction indicated by the Y-axis of the vehicle coordinate system is the length direction of the target vehicle, such as the direction from the rear of the target vehicle to the front of the target vehicle; the direction indicated by the Z-axis of the vehicle coordinate system is the height direction of the target vehicle, such as the direction from the bottom of the target vehicle to the roof of the target vehicle.

[0080] It should be noted that Figure 2 The X-axis, Y-axis, and Z-axis in are the X-axis, Y-axis, and Z-axis of the geographic coordinate system. Figure 2 The directions of the X-axis, Y-axis, and Z-axis in the current vehicle coordinate system of the target vehicle are the same as those in the geographic coordinate system. Figure 2The vehicle coordinate system is not shown in the figure. As the target vehicle moves, the directions of the X-axis, Y-axis and Z-axis in the vehicle coordinate system of the target vehicle may change. Based on the changes in the directions of the X-axis, Y-axis and Z-axis in the vehicle coordinate system of the target vehicle, the corresponding roll angle, pitch angle and heading angle can be calculated.

[0081] Taking the visual positioning sensor as an example, the visual positioning sensor is a sensor that performs positioning based on the semantic information of the vehicle's environment (lane lines, etc.). Therefore, the positioning result determined by the visual positioning sensor for the target vehicle has a higher positioning accuracy in the lateral direction of the three-dimensional coordinate system, but a lower positioning accuracy in the longitudinal and elevation directions of the three-dimensional coordinate system. Taking the inertial navigation positioning sensor as an example, the inertial navigation positioning sensor is a sensor that performs positioning through the internal gyroscope and accelerometer. Since the earth's gravity always applies a constant downward acceleration, it makes it difficult for the accelerometer to distinguish between acceleration and gravity caused by movement. Therefore, the positioning result determined by the inertial navigation positioning sensor for the target vehicle has a relatively higher positioning accuracy in the lateral and longitudinal directions, but a relatively lower positioning accuracy in the elevation direction.

[0082] It can be seen that different types of positioning sensors have different positioning accuracy in different positioning directions due to different working principles. The fusion positioning method provided in the embodiment of the present application is based on the type of positioning sensor, and selectively fuses the positioning information in the more accurate positioning direction in the positioning result, that is, in the method provided in the present application, the target positioning direction can be determined from all positioning directions of the first positioning sensor based on the type of the first positioning sensor and / or the type of the second positioning sensor (that is, the target positioning direction is the direction in which the positioning results of the first positioning sensor in all positioning directions are more accurate), and then the target positioning information in the target positioning direction is extracted based on the first positioning result to fuse the positioning information of the second positioning sensor; it can not only improve the accuracy of the fused positioning result in some positioning directions, but also reduce the negative impact of some inaccurate information in the first positioning result on the fused positioning result.

[0083] In the embodiment of the present application, since different types of positioning sensors have different positioning accuracies in different positioning directions, the target positioning direction is determined based on the type of the first positioning sensor. The target positioning direction is a positioning direction with a higher positioning accuracy of the first positioning sensor. In other words, the positioning accuracy of the first positioning sensor in the target positioning direction is higher than the positioning accuracy in other positioning directions except the target positioning direction among all positioning directions.

[0084] In one embodiment, a correspondence between the type of positioning sensor and the target positioning direction is preset. When performing fusion positioning, based on the type of the first positioning sensor, the target positioning direction corresponding to the first positioning sensor can be obtained from the correspondence, and the target positioning information in the target positioning direction can be extracted based on the first positioning result.

[0085] It should be noted that the target positioning direction may be one positioning direction or multiple positioning directions, and the embodiment of the present application does not limit the number of target positioning directions.

[0086] The target positioning information may be the positioning result of the target positioning direction in the first positioning result; illustratively, the first positioning result includes longitude, latitude and altitude, wherein longitude is the positioning result of the first positioning result in the horizontal direction, latitude is the positioning result of the first positioning result in the vertical direction, and altitude is the positioning result of the first positioning result in the elevation direction. If the target positioning direction is horizontal, the target positioning information may be the longitude in the first positioning result.

[0087] The target positioning information may also be information inferred in the target positioning direction based on the first positioning result; illustratively, the target positioning information may be the overall error of the second positioning result inferred in the target positioning direction based on the first positioning result, the position error in the overall error, etc. The embodiment of the present application does not limit the target positioning information.

[0088] S103: Fusing the target positioning information with the second positioning result to obtain a fused positioning result of the target vehicle.

[0089] The fused positioning result is used to indicate the geographic location of the target vehicle. In one embodiment, the fused positioning result may include the location of the target vehicle, and the fused positioning result may also include at least one other information such as the posture of the target vehicle, the speed of the target vehicle, and the direction of the target vehicle while including the location of the target vehicle. The embodiment of the present application does not limit the content of the fused positioning result.

[0090] Among them, when the extracted target positioning information is fused with the second positioning sub-result, at least one method such as filtering algorithm, optimization algorithm, weighted average, etc. can be used for fusion. The embodiment of the present application does not limit the fusion method.

[0091] The fusion positioning method provided in the embodiment of the present application takes into account the different positioning accuracies of different types of positioning sensors in different positioning directions. When performing fusion positioning, the target positioning direction is determined based on the type of the first positioning sensor, and only the target positioning information of the first positioning result in the target positioning direction is fused with the second positioning result, while the positioning information of the first positioning result in other positioning directions is not fused with the second positioning result. This can selectively fuse the more accurate information in the target positioning direction in the first positioning result with the second positioning result, which can not only improve the accuracy of the fusion positioning result in the target positioning direction, but also reduce the negative impact of some less accurate information in the first positioning result on the accuracy of the fusion positioning result, thereby improving the accuracy of the fusion positioning result in each positioning direction.

[0092] Next, through Figure 3 The embodiment shown, Figure 1 The illustrated embodiment is described in more detail. Figure 3 A flowchart of a fusion positioning method provided as an exemplary embodiment. The execution subject of the method may be a vehicle end, which may include but is not limited to any one or more vehicle-mounted modules such as a vehicle, a processing unit deployed on the vehicle, an intelligent driving unit or a positioning unit; the execution subject of the method may also be a service end that communicates with the vehicle, which may be a service platform, an independent terminal device, an Internet of Things end, an Internet of Vehicles end, a distributed system end, etc. The service platform may be an independent server or a server cluster, etc.; the execution subject of the method may also be completed by the vehicle end and the service end. For the description of the vehicle end and the service end, please refer to the previous content and will not be repeated here; the embodiments of this application do not limit the execution subject. Figure 3 As shown, the method may include the following steps:

[0093] S301: Obtain a first positioning result determined for a target vehicle by a first positioning sensor, and obtain a second positioning result determined for the target vehicle by a second positioning sensor, where the first positioning sensor and the second positioning sensor are of different types.

[0094] In one embodiment, the first positioning sensor and the second positioning sensor output positioning results at a certain frequency, and obtaining the first positioning result determined for the target vehicle by the first positioning sensor is to receive the first positioning result output by the first positioning sensor, and obtaining the second positioning result determined for the target vehicle by the second positioning sensor is to receive the second positioning result output by the second positioning sensor. Therefore, obtaining the first positioning result determined for the target vehicle by the first positioning sensor, and obtaining the second positioning result determined for the target vehicle by the second positioning sensor, include: receiving the first positioning result output by the first positioning sensor and the second positioning result output by the second positioning sensor.

[0095] In one embodiment, the positioning frequencies of the first positioning sensor and the second positioning sensor are different. At some times, only the second positioning result output by the second positioning sensor can be received, and at some times, both the second positioning result output by the second positioning sensor and the first positioning result output by the first positioning sensor can be received. The fusion positioning method provided in the embodiment of the present application is executed when both the first positioning result of the first positioning sensor and the second positioning result of the second positioning sensor are received.

[0096] S302: predicting an overall error of the second positioning result in the target positioning direction based on the first positioning result, where the overall error is used to represent a difference between the second positioning result and a true positioning result of the target vehicle in the target positioning direction.

[0097] In one embodiment, the overall error may include at least one of the position error, speed error, attitude error and orientation error, and the embodiment of the present application does not limit the content of the overall error. Among them, the position error is used to indicate the difference between the vehicle position in the second positioning result and the vehicle position in the real positioning result of the target vehicle; the speed error is used to indicate the difference between the vehicle speed in the second positioning result and the vehicle speed in the real positioning result of the target vehicle; the attitude error is used to indicate the difference between the vehicle attitude in the second positioning result and the vehicle attitude in the real positioning result of the target vehicle; the orientation error is used to indicate the difference between the vehicle orientation in the second positioning result and the vehicle orientation in the real positioning result of the target vehicle.

[0098] In an illustrated embodiment, when predicting the overall error of the second positioning result in the target positioning direction based on the first positioning result, the overall error of the second positioning result in the target positioning direction can be directly predicted. Among them, a filtering algorithm, a neural network model, and other methods can be used to directly predict the overall error of the second positioning result in the target positioning direction based on the first positioning result.

[0099] In another illustrated embodiment, when predicting the overall error of the second positioning result in the target positioning direction based on the first positioning result, the state error amount corresponding to the second positioning result can be predicted. Among them, the state error amount corresponding to the second positioning result includes the overall error of the second positioning result in all positioning directions. Therefore, by predicting the state error amount corresponding to the second positioning result, the overall error of the second positioning result in the target positioning direction can be obtained. It should be noted that the state error amount also includes the measurement error of the second positioning sensor, and the embodiment of the present application does not limit the content of the state error amount corresponding to the second positioning result.

[0100] Taking the second positioning sensor as an inertial navigation positioning sensor as an example, the state error amount corresponding to the second positioning result may include at least one of an attitude misalignment angle, a velocity error, a position error, a gyro zero bias, an added table zero bias and an installation deflection angle.

[0101] Among them, the attitude misalignment angle includes the rolling misalignment angle, the pitch misalignment angle and the heading misalignment angle. The rolling misalignment angle is used to indicate the difference between the vehicle attitude in the second positioning result and the vehicle attitude in the real positioning result of the target vehicle in the rolling direction. The pitch misalignment angle is used to indicate the difference between the vehicle attitude in the second positioning result and the vehicle attitude in the real positioning result of the target vehicle in the pitch direction. The heading misalignment angle is used to indicate the difference between the vehicle attitude in the second positioning result and the vehicle attitude in the real positioning result of the target vehicle in the heading direction.

[0102] The speed error includes a lateral speed sub-error, a longitudinal speed sub-error and an elevation speed sub-error. The lateral speed sub-error is used to indicate the lateral difference between the vehicle speed in the second positioning result and the actual positioning result of the target vehicle. The longitudinal speed sub-error is used to indicate the longitudinal difference between the vehicle speed in the second positioning result and the actual positioning result of the target vehicle. The elevation speed sub-error is used to indicate the elevation difference between the vehicle speed in the second positioning result and the actual positioning result of the target vehicle.

[0103] The position error includes a lateral position sub-error, a longitudinal position sub-error and an elevation direction position sub-error. The lateral position sub-error is used to indicate the lateral difference between the vehicle position in the second positioning result and the actual positioning result of the target vehicle. The longitudinal position sub-error is used to indicate the longitudinal difference between the vehicle position in the second positioning result and the actual positioning result of the target vehicle. The elevation direction position sub-error is used to indicate the elevation difference between the vehicle position in the second positioning result and the actual positioning result of the target vehicle.

[0104] Gyro bias is used to indicate the difference between the angular velocity measured by the gyroscope in the inertial navigation sensor and the actual angular velocity of the target vehicle. Table bias is used to indicate the difference between the acceleration measured by the accelerometer in the inertial navigation sensor and the actual acceleration of the target vehicle. Installation deflection is used to indicate the angular offset between the inertial navigation sensor and the target vehicle when the inertial navigation sensor is installed on the target vehicle.

[0105] Exemplarily, when the target positioning direction is lateral, the overall error of the second positioning result in the target positioning direction includes a position error and a velocity error, and the position error and velocity error included in the overall error in the target positioning direction are the position error and velocity error of the second positioning result in the target positioning direction, rather than the position error and velocity error of the second positioning result in all positioning directions. Exemplarily, the overall error of the second positioning result in the lateral direction includes a lateral position sub-error and a lateral velocity sub-error.

[0106] In an embodiment shown, a filter, a neural network model, or the like may be used to predict the state error amount corresponding to the second positioning result based on the first positioning result. The filtering algorithm may be a KF (Kalman Filter), an ESKF (Error State Kalman Filter), a PF (Particle Filter), a complementary filter, or the like.

[0107] Among them, ESKF is used to estimate the error between the current estimated state and the actual state. After the error is estimated, the error is superimposed on the previous estimated state to obtain a corrected state. In the fusion positioning scenario, ESKF estimates the error between the positioning result of the positioning sensor and the actual positioning result of the vehicle. After the error is estimated, the error is superimposed on the positioning result of the positioning sensor to obtain a fusion positioning result. Next, the embodiment of the present application takes the first positioning sensor as a visual positioning sensor, the second positioning sensor includes an inertial positioning sensor, and the prediction method used is ESKF as an example to exemplify the process of predicting the state error amount corresponding to the second positioning result based on the first positioning result:

[0108] ESKF has a state model and an observation model. The state model is used to describe how the system state evolves over time. The observation model describes how to calculate the predicted observation from the system state. It establishes a mathematical relationship between the system state and the sensor measurement. By comparing the actual observation and the predicted observation, the measurement residual is calculated, and then the state error is updated.

[0109] The state model is: , X is the state error, A is the state transfer matrix, G is the process noise driving matrix, W is the process noise, represents the derivative of X.

[0110] Exemplarily, the state error X is a 17-dimensional matrix, expressed as .in, It is the three-dimensional attitude misalignment angle, including roll misalignment angle, pitch misalignment angle and heading misalignment angle. It is a three-dimensional velocity error, including the lateral velocity sub-error, the longitudinal velocity sub-error and the elevation velocity sub-error. It is a three-dimensional position error, including lateral position sub-error, longitudinal position sub-error and elevation direction position sub-error.

[0111] It should be noted that the three-dimensional coordinate system can be a northeast celestial coordinate system or an upper right front coordinate system, etc. When the three-dimensional coordinate system is a northeast celestial coordinate system, the direction pointed by the X-axis, that is, the horizontal direction, can be called the east direction, the direction pointed by the Y-axis, that is, the vertical direction, can be called the north direction, and the direction pointed by the Z-axis, that is, the elevation direction, can be called the celestial direction; when the three-dimensional coordinate system is an upper right front coordinate system, the direction pointed by the X-axis, that is, the horizontal direction, can be called the right direction, the direction pointed by the Y-axis, that is, the vertical direction, can be called the front direction, and the direction pointed by the Z-axis, that is, the elevation direction, can be called the up direction.

[0112] It is the three-dimensional gyro bias, which is used to represent the difference between the angular velocity measured by the gyroscope in the inertial navigation positioning sensor and the actual angular velocity of the target vehicle; It is a three-dimensional table zero bias, which is used to represent the difference between the acceleration measured by the accelerometer in the inertial navigation positioning sensor and the actual acceleration of the target vehicle; is the pitch installation angle of the inertial navigation positioning sensor, It is the heading installation deflection angle of the inertial navigation positioning sensor.

[0113] Among them, the second positioning sensor can be a positioning sensor or multiple positioning sensors of different types. If the second positioning sensor also includes a wheel speed meter, the scale coefficient error k of the wheel speed meter can be added to the state error amount X to obtain an 18-dimensional state error amount.

[0114] A represents the state transfer matrix, which is used to describe the evolution of the system state over time when there is no external input. The specific values ​​in the state transfer matrix can be determined during the ESKF calculation process. G represents the process noise driving matrix, which is used to describe the impact of system model uncertainty on the system state. The specific values ​​in the process noise driving matrix can be determined during the ESKF calculation process. W represents the process noise, which is used to describe the content dynamics and external disturbances that are not captured by the system model. The value of the process noise can be determined during the ESKF calculation process.

[0115] The observation model is: Z=HX+V; where Z represents the observation value, which refers to the actual data obtained based on the positioning sensor. H represents the observation matrix, which is used to describe how to calculate the predicted observation value based on the system state. A represents the state transfer matrix. V represents the observation noise, which is used to represent the deviation between the measured value and the true value and can be configured based on the characteristics of the positioning sensor.

[0116] like Figure 4 As shown, after obtaining the first positioning result, the first positioning result can be observed and converted based on the second positioning result to obtain the observation quantity dp and the observation matrix H. The observation quantity dp and the observation matrix H are input into the ESKF for calculation. The ESKF outputs the state error quantity X, which is as follows:

[0117] (1) After obtaining the first positioning result, convert the first positioning result in the navigation coordinate system n to the carrier coordinate system b:

[0118] The first positioning result includes longitude, latitude and altitude. Through the UTM (Universal Transverse Mercator) coordinate system, the positions corresponding to the longitude, latitude and altitude in the first positioning result are converted into the easting, northing and celestial coordinates in the UTM coordinate system to obtain the position in the navigation coordinate system. At this time, the position in the first positioning result is expressed as Py_n= (Pe, Pn, Pu).

[0119] Based on the attitude of the target vehicle in the second positioning result of the inertial navigation positioning sensor, the conversion relationship between the carrier coordinate system and the navigation coordinate system can be determined, and the conversion relationship can be represented by an attitude conversion matrix. For example, the attitude of the target vehicle in the second positioning result is a roll angle a1, a pitch angle a2, and a heading angle a3, then the attitude conversion matrix can be expressed as Cbn=Rz(a3)Ry(a2)Rx(a1), where Rx is the rotation matrix around the x-axis, Ry is the rotation matrix around the y-axis, and Rz is the rotation matrix around the z-axis.

[0120]

[0121]

[0122]

[0123] Cbn=

[0124] Among them, sin is the sine function and cos is the cosine function.

[0125] The coordinate system of Py_n is transformed by the attitude conversion matrix Cbn to obtain Py_b=Cbn*Py_n. Among them, Py_b is the position vector of the first positioning result in the carrier coordinate system. Among them, the navigation coordinate system is the northeast celestial coordinate system, and the carrier coordinate system is the upper right front coordinate system. By converting the first positioning result from the navigation coordinate system to the carrier coordinate system, the position vector of the first positioning result is converted from the northeast celestial vector to the upper right front vector.

[0126] (2) Convert the second positioning result in the navigation coordinate system n to the carrier coordinate system b:

[0127] The position Ps_b is obtained by performing coordinate system transformation on the position in the second positioning result of the inertial navigation positioning sensor through the attitude transformation matrix Cbn.

[0128] (3) Based on the positioning results in the carrier coordinate system, construct the observation quantity and observation matrix:

[0129] The difference between the position of visual positioning and the position of inertial positioning in the carrier coordinate system is obtained as the observation quantity: dp = Ps_b- Py_b. Among them, dp is the observation quantity Z corresponding to the visual positioning sensor in the observation model. Since the observation quantity is constructed in the carrier coordinate system, the observation matrix H=(Cbn) can be constructed T , the observation matrix H is the transpose of the attitude transformation matrix Cbn.

[0130] (4) The observation quantity dp and the observation matrix H are input into ESKF for calculation. ESKF outputs the state error X, which is the state error corresponding to the second positioning result of the inertial navigation positioning sensor. Subsequently, the second positioning result of the inertial navigation positioning sensor can be corrected by the state error to obtain a fusion positioning result. In addition, the data collected by the inertial navigation positioning sensor can be corrected based on the gyro bias, table bias, installation deflection, etc. in the state error X to obtain a more accurate second positioning result.

[0131] It should be noted that Figure 4The process of calculating the state error amount corresponding to the second positioning result is only exemplified by taking the second positioning sensor as an inertial navigation positioning sensor as an example. In another embodiment, the second positioning sensor also includes a navigation positioning sensor and a wheel speed meter. The positioning results of the navigation positioning sensor and the wheel speed meter can also be input into the ESKF to calculate the state error amount X corresponding to the second positioning result of the inertial navigation positioning sensor together. The second positioning result of the inertial navigation positioning sensor is corrected by the state error amount X to obtain a fused positioning result, such as Figure 5 shown.

[0132] S303: Determine target positioning information based on the position error in the overall error.

[0133] According to the above step 302, the overall error may include at least one of the position error, speed error, attitude error and orientation error. In the embodiment of the present application, the position is the most important part of the positioning result, and the speed and orientation in the positioning result are mostly accurate, and the attitude has a relatively small impact on the subsequent processing based on the positioning result. Therefore, the target positioning information can be determined based only on the position error in the overall error.

[0134] It should be noted that, when the overall error is an overall error in the horizontal direction, the position error in the overall error is the position error in the horizontal direction, which can also be called the horizontal position sub-error; when the overall error is an overall error in the vertical direction, the position error in the overall error is the position error in the vertical direction, which can also be called the longitudinal position sub-error; when the overall error is an overall error in the elevation direction, the position error in the overall error is the position error in the elevation direction, which can also be called the elevation direction position sub-error.

[0135] In one embodiment, the above step 303 may include: determining the position error in the overall error as the target positioning information; or performing coordinate system conversion on the position error in the overall error, and determining the converted position error as the target positioning information.

[0136] For example, the position error in the overall error is in the navigation coordinate system n. The position error in the target positioning direction can be converted to (1) Convert to carrier coordinate system b: =(Cbn) T .in, (1) is the position error in the target positioning direction in the carrier coordinate system. (1) Determined as target positioning information, (Cbn) T is the transpose of the posture transformation matrix Cbn.

[0137] In another embodiment, the overall error of the second positioning result in the target positioning direction is obtained by predicting the state error amount corresponding to the second positioning result. In this case, the target positioning information can be directly determined based on the position error of the second positioning result in the target positioning direction in the state error amount. Wherein, determining the target positioning information based on the position error of the second positioning result in the target positioning direction in the state error amount includes: determining the position error in the overall error as the target positioning information; or performing a coordinate system conversion on the position error in the overall error, and determining the converted position error as the target positioning information.

[0138] For example, the position error in the overall error It is in the navigation coordinate system n. The position error in the target positioning direction can be first calculated based on the attitude transformation matrix. Convert to carrier coordinate system b: =(Cbn) T . is the position error in the carrier coordinate system. The three components of the position error are in the order of right, top, and front. Now only the right position error (i.e., horizontal) is fed back, and the forward (i.e., vertical) and upward (elevation) errors are set to zero, that is, the second and third items are set to zero. (2) = (3) = 0. Then, the second and third terms are set to zero through the attitude transformation matrix. Converted to the navigation coordinate system, we get =Cbn .Should That is the target positioning information.

[0139] It should be noted that the embodiment of the present application only takes the determination of target positioning information based on the position error in the overall error as an example to exemplify the process of obtaining target positioning information. In another embodiment, the target positioning information can also be determined directly based on the overall error of the second positioning result in the target positioning direction.

[0140] Another point that needs to be explained is that the embodiment of the present application only takes the example of determining the target positioning information by predicting the error of the second positioning result as an example to exemplify the process of obtaining the target positioning information. In another embodiment, extracting the target positioning information in the target positioning direction based on the first positioning result includes: determining the target positioning information based on the positioning result in the target positioning direction in the first positioning result. Among them, determining the target positioning information based on the positioning result in the target positioning direction in the first positioning result includes: directly determining the positioning result in the target positioning direction in the first positioning result as the target positioning information; or, performing a coordinate system conversion on the positioning result in the target positioning direction in the first positioning result, and determining the converted positioning result as the target positioning information.

[0141] For example, the target positioning direction is horizontal, and the first positioning result includes longitude, latitude and altitude, wherein the longitude is the positioning result of the first positioning result in the horizontal direction, and the target positioning information is determined based on the longitude in the first positioning result.

[0142] S304: Fuse the target positioning information with the second positioning result to obtain a fused positioning result of the target vehicle.

[0143] In one embodiment, the target positioning information is a position error in an overall error of the second positioning result in the target positioning direction. The target positioning information is fused with the second positioning result to obtain a fused positioning result of the target vehicle, including: subtracting the target positioning information from the positioning result in the target positioning direction in the second positioning result to obtain a fused positioning result of the target vehicle.

[0144] In one embodiment, the target positioning information is a positioning result in the target positioning direction in the first positioning result. The target positioning information is fused with the second positioning result to obtain a fused positioning result of the target vehicle, including: using at least one optimization algorithm such as a filtering algorithm, an optimization algorithm, or a weighted average method to fuse the target positioning information with the second positioning result to obtain a fused positioning result of the target vehicle.

[0145] Exemplarily, taking the first positioning sensor as a visual positioning sensor as an example, the fusion positioning method provided in the embodiment of the present application is exemplarily described:

[0146] The first positioning sensor is a visual positioning sensor. All positioning directions corresponding to the first positioning result determined by the visual positioning sensor for the target vehicle include lateral, longitudinal and elevation directions. Of course, it can also include at least one of the rolling direction, pitching direction, heading, etc. Based on the working principle of the visual positioning sensor, it can be known that the lateral positioning result of the visual positioning sensor is very accurate. Therefore, the target positioning direction is lateral. Among them, the lateral direction is the direction indicated by the X-axis constituting the three-dimensional coordinate system, the longitudinal direction is the direction indicated by the Y-axis constituting the three-dimensional coordinate system, and the elevation direction is the direction indicated by the Z-axis constituting the three-dimensional coordinate system.

[0147] The fusion positioning method includes: obtaining a first positioning result determined for a target vehicle by a visual positioning sensor, and obtaining a second positioning result determined for the target vehicle by a second positioning sensor, wherein the visual positioning sensor and the second positioning sensor are of different types; extracting lateral target positioning information based on the first positioning result; and fusing the target positioning information with the second positioning result to obtain a fusion positioning result of the target vehicle.

[0148] Among them, extracting the target positioning information in the horizontal direction based on the first positioning result includes: predicting the horizontal position error of the second positioning result based on the first positioning result, and determining the target positioning information based on the position error; or determining the target positioning information based on the horizontal positioning result in the first positioning result.

[0149] The state error amount corresponding to the second positioning result includes a lateral position error of the second positioning result. Therefore, the lateral position error of the second positioning result can be predicted by predicting the state error amount corresponding to the second positioning result.

[0150] It should be noted that, in one embodiment, the second positioning sensor includes a navigation positioning sensor and an inertial navigation positioning sensor, and the positioning frequency of the navigation positioning sensor is different from that of the inertial navigation positioning sensor. Among them, the positioning frequency of the inertial navigation positioning sensor is higher than the positioning frequency of the visual positioning sensor. Therefore, at some moments, only the positioning results of the inertial navigation positioning sensor and the first positioning sensor can be obtained. At this time, the solution provided in the embodiment of the present application can be adopted. At some moments, the positioning results of the inertial navigation positioning sensor, the first positioning sensor and the navigation positioning sensor can be obtained. Since the positioning accuracy of the navigation positioning sensor in the longitudinal and elevation directions is high, and the filtering algorithm will filter the input information, therefore, based on ESKF, these three positioning results are processed to obtain a relatively accurate fusion positioning result, thereby eliminating the need to perform the step of extracting the target positioning information in the embodiment of the present application.

[0151] Optionally, the fusion positioning method also includes: if the inertial navigation positioning result of the inertial navigation positioning sensor and the navigation positioning result of the navigation positioning sensor are obtained at the current moment, and the first positioning result of the first positioning sensor is not obtained, then the inertial navigation positioning result and the navigation positioning result are directly fused to obtain the fused positioning result of the target vehicle; if the inertial navigation positioning result of the inertial navigation positioning sensor and the first positioning result of the first positioning sensor are obtained at the current moment, and the navigation positioning result of the navigation positioning sensor is not obtained, then the target positioning information in the target direction is extracted based on the first positioning result, and the target positioning information is fused with the second positioning result to obtain the fused positioning result of the target vehicle.

[0152] The fusion positioning method provided in the embodiment of the present application takes into account the different positioning accuracies of different types of positioning sensors in different positioning directions. When performing fusion positioning, the target positioning direction is determined based on the type of the first positioning sensor, and only the target positioning information of the first positioning result in the target positioning direction is fused with the second positioning result, while the positioning information of the first positioning result in other positioning directions is not fused with the second positioning result. This achieves the selective fusion of the more accurate information in the target positioning direction in the first positioning result with the second positioning result, which can not only improve the accuracy of the fused positioning result in the target positioning direction, but also reduce the negative impact of the less accurate part of the information in the first positioning result on reducing the accuracy of the fused positioning result, thereby improving the accuracy of the fused positioning result in each positioning direction.

[0153] It should be noted that the embodiment of the present application further provides a new dead reckoning algorithm, through which the first positioning result of the first positioning sensor is predicted. Figure 6 A flowchart of a method for predicting a first positioning result provided as an exemplary embodiment. The execution subject of the method may be a vehicle end, which may include but is not limited to any one or more vehicle-mounted modules such as a vehicle, a processing unit deployed on the vehicle, an intelligent driving unit or a positioning unit; the execution subject of the method may also be a service end that communicates with the vehicle, which may be a service platform, an independent terminal device, an Internet of Things end, an Internet of Vehicles end, a distributed system end, etc. The service platform may be an independent server or a server cluster, etc.; the execution subject of the method may also be completed by the vehicle end and the service end. For the description of the vehicle end and the service end, please refer to the previous content and will not be repeated here; the embodiment of the present application does not limit the execution subject. Figure 6 As shown, the method may include the following steps:

[0154] S601: Determine the dead position positioning result of the second positioning sensor at the current moment, and determine the dead position positioning result of the second positioning sensor at the historical moment.

[0155] The historical moment may be a previous moment, which may be the moment when the second positioning sensor last performed positioning, or the moment when the first positioning sensor last performed positioning. In some embodiments, the first positioning sensor and the second positioning sensor have different positioning frequencies, so the moment when the second positioning sensor last performed positioning and the moment when the first positioning sensor last performed positioning may be different moments.

[0156] In one embodiment, each time the second positioning sensor performs positioning, a current dead position positioning result is generated. Therefore, in the above step 601, the current dead position positioning result is calculated in real time, and the historical dead position positioning result is calculated at the historical moment, and only needs to be obtained from the historical data.

[0157] It should be noted that the method provided in the embodiment of the present application predicts the first positioning result based on the change of the dead position positioning result. In this way, it is sufficient that the change amount of the dead position positioning result is accurate, and the requirement for the overall accuracy of the dead position positioning result is not high, which can avoid the influence of the cumulative error in the process of determining the dead position positioning result.

[0158] In one embodiment, since the embodiment of the present application predicts the first positioning result based on the change of the dead position positioning result, based on the type of the first positioning sensor, only the change amount of the dead position positioning result in certain positioning directions can be accurately determined, while the change amount of the dead position positioning result in other positioning directions can be ignored.

[0159] Taking the case where the first positioning sensor is a visual positioning sensor, the visual positioning sensor can determine the lateral position of the target vehicle very quickly and accurately based on environmental semantic information such as lane lines. Therefore, even if there is a large deviation in the heading angle measured by the second positioning sensor, resulting in a large deviation in the lateral position of the determined dead position positioning result, it will not have any impact on the positioning of the first positioning sensor. Therefore, when determining the dead position positioning result of the second positioning sensor, it is only necessary to calibrate part of the posture measured by the second positioning sensor.

[0160] In some embodiments, the dead position positioning result is obtained by: determining the second attitude misalignment angle based on the partial misalignment angle of the first attitude misalignment angle of the second positioning sensor, wherein the partial misalignment angle is determined based on the type of the first positioning sensor; performing dead reckoning based on the positioning data of the second positioning sensor and the second attitude misalignment angle to obtain the dead position positioning result. The positioning data of the second positioning sensor refers to the data used for positioning collected by the second positioning sensor during the positioning process. Taking the second positioning sensor as an inertial navigation positioning sensor as an example, the positioning data is the acceleration collected by the accelerometer and the angular velocity collected by the gyroscope in the inertial navigation positioning sensor.

[0161] The attitude misalignment angle is used to indicate the difference between the attitude of the target vehicle determined by the second positioning sensor and the actual attitude of the target vehicle. The first attitude misalignment angle is the total attitude misalignment angle of the second positioning sensor, and the first attitude misalignment angle includes the horizontal misalignment angle and the azimuth misalignment angle. The horizontal misalignment angle includes the roll misalignment angle and the pitch misalignment angle, and the azimuth misalignment angle is the heading misalignment angle. Among them, the first attitude misalignment angle can be obtained based on the method provided in the above step 302, that is, the first attitude misalignment angle can be the attitude misalignment angle of the second positioning sensor determined in the last fusion positioning process.

[0162] Wherein, when determining the second attitude misalignment angle based on the partial misalignment angle in the first attitude misalignment angle of the second positioning sensor, the partial misalignment angle in the first attitude misalignment angle can be directly used as the second attitude misalignment angle. For example, the first attitude misalignment angle includes a horizontal misalignment angle and an azimuth misalignment angle, and the azimuth misalignment angle is set to zero to obtain the second attitude misalignment angle, that is, the horizontal misalignment angle in the first attitude misalignment angle is used as the second attitude misalignment angle.

[0163] Among them, the partial misalignment angle is determined based on the type of the first positioning sensor, and the partial misalignment angle is used to enhance the positioning accuracy of the dead position positioning result in the first positioning direction, and the first positioning direction is a positioning direction different from the target positioning direction. Since the partial misalignment angle of the second positioning sensor is taken into account in the dead reckoning process, the positioning accuracy of the dead position positioning result in the first positioning direction is higher, so when the first positioning result is predicted based on the change amount of the dead position positioning result, the positioning accuracy of the first positioning result in the first positioning direction can be improved, and the target positioning direction is the positioning direction with higher positioning accuracy of the first positioning sensor itself, and there is no need to further improve it based on the dead position positioning result. It can be seen that predicting the first positioning result based on the change amount of the dead position positioning result not only makes the first positioning result have a higher positioning accuracy in the target positioning direction, but also improves the positioning accuracy of the first positioning result in the first positioning direction.

[0164] Exemplarily, the first positioning sensor is a visual sensor, and the partial misalignment angle is a horizontal misalignment angle in the first posture misalignment angle.

[0165] In one embodiment, when dead reckoning is performed based on the positioning data of the second positioning sensor and the second attitude misalignment angle to obtain a dead position positioning result, the dead position positioning result can be obtained first, and then the dead position positioning result can be corrected by the second attitude misalignment angle; or the positioning data can be corrected first by the second attitude misalignment angle, and then the dead position positioning result can be determined based on the corrected positioning data.

[0166] Optionally, performing dead reckoning based on the positioning data of the second positioning sensor and the second attitude misalignment angle to obtain a dead positioning result includes: performing dead reckoning based on the positioning data of the second positioning sensor to obtain an initial dead positioning result; and compensating the initial dead positioning result based on the second attitude misalignment angle to obtain a compensated dead positioning result.

[0167] Optionally, performing dead reckoning based on the positioning data of the second positioning sensor and the second attitude misalignment angle to obtain a dead positioning result includes: compensating the positioning data of the second positioning sensor based on the second attitude misalignment angle; and performing dead reckoning based on the compensated positioning data to obtain the dead positioning result.

[0168] In one embodiment, the second positioning sensor includes an inertial navigation positioning sensor. Since the dead-positioning result of the inertial navigation positioning sensor is in the carrier coordinate system, and the first positioning result of the first positioning sensor is in the navigation coordinate system. Therefore, it is necessary to convert the dead-positioning result of the inertial navigation positioning sensor into the navigation coordinate system, and the conversion relationship is determined based on the attitude in the dead-positioning result of the inertial navigation positioning sensor. Therefore, for the inertial navigation positioning sensor, the dead-positioning result can be obtained first, and then the dead-positioning result can be compensated by the second attitude misalignment angle. Dead-positioning is performed based on the positioning data of the second positioning sensor and the second attitude misalignment angle to obtain the dead-positioning result, including: performing dead-positioning based on the positioning data of the inertial navigation positioning sensor to obtain an initial dead-positioning result; based on the second attitude misalignment angle, the initial dead-positioning result is compensated to obtain a compensated dead-positioning result.

[0169] In one embodiment, the second positioning sensor includes an inertial navigation positioning sensor and a third positioning sensor, and the dead-positioning result includes the attitude and position of the target vehicle; wherein the inertial navigation positioning sensor is used to determine the attitude included in the dead-positioning result, and the wheel speed meter is used to determine the position included in the dead-positioning result. Since the attitude determined by the inertial navigation positioning sensor is in the carrier coordinate system, and the first positioning result of the first positioning sensor is in the navigation coordinate system. Therefore, it is necessary to convert the attitude determined by the inertial navigation positioning sensor to the navigation coordinate system, and the conversion relationship is determined based on the attitude determined by the inertial navigation positioning sensor. Therefore, for the inertial navigation positioning sensor, the dead-positioning result can be obtained first, and then the dead-positioning result can be compensated by the second attitude misalignment angle; for the third positioning sensor, the positioning data can be compensated by the second attitude misalignment angle first, and then the dead-positioning result can be determined based on the compensated positioning data.

[0170] Optionally, dead reckoning is performed based on the positioning data of the second positioning sensor and the second attitude misalignment angle to obtain a dead positioning result, including: performing dead reckoning based on the positioning data of the inertial navigation positioning sensor to obtain an initial attitude; compensating the initial attitude based on the second attitude misalignment angle to obtain the attitude in the dead positioning result; compensating the positioning data of the third positioning sensor based on the second attitude misalignment angle; and performing dead reckoning based on the compensated positioning data to obtain the position in the dead positioning result.

[0171] The third positioning sensor is a positioning sensor of a different type from the first positioning sensor and the inertial navigation positioning sensor. Exemplarily, the third positioning sensor may be a wheel speed meter, and the embodiment of the present application does not limit the third positioning sensor.

[0172] Among them, the initial posture is the posture of the target vehicle in the carrier coordinate system, and the initial posture needs to be converted to the navigation coordinate system. The posture determined by the inertial navigation positioning sensor determines the conversion relationship between the carrier coordinate system and the navigation coordinate system. Therefore, the conversion relationship between the carrier coordinate system and the navigation coordinate system can be determined based on the posture misalignment angle and the initial posture.

[0173] Optionally, based on the second attitude misalignment angle, the initial attitude is compensated to obtain the attitude in the dead position positioning result, including: based on the initial attitude and the second attitude misalignment angle, determining a first transformation relationship, the first transformation relationship is used to represent the transformation relationship between the navigation coordinate system of the target vehicle and the carrier coordinate system; using the first transformation relationship to transform the initial attitude to obtain the attitude of the target vehicle in the navigation coordinate system.

[0174] In one embodiment, the first conversion relationship is represented by a posture conversion matrix. Since the posture conversion matrix and the posture can be converted to each other, the initial posture is first compensated based on the posture misalignment angle, and then the posture conversion matrix is ​​generated based on the compensated posture; or the posture conversion matrix is ​​first generated based on the initial posture, and then the posture conversion matrix is ​​compensated based on the posture misalignment angle, the results of the two are the same.

[0175] Therefore, based on the initial posture and the second posture misalignment angle, the first conversion relationship is determined, including: based on the initial posture, generating a posture conversion matrix; based on the second posture misalignment angle, compensating the posture conversion matrix to obtain a compensated posture conversion matrix; or, based on the second posture misalignment angle, compensating the initial posture to obtain a compensated posture; based on the compensated posture, generating a posture conversion matrix.

[0176] When the first conversion relationship is a posture conversion matrix, the initial posture is converted using the first conversion relationship to obtain the posture of the target vehicle in the navigation coordinate system, including: multiplying the initial posture by the posture conversion matrix to obtain the posture of the target vehicle in the navigation coordinate system.

[0177] Next, take the second positioning sensor including the inertial navigation positioning sensor as an example. Figure 7 The following example illustrates the process of determining the attitude in the dead position positioning result:

[0178] Among them, the inertial navigation positioning sensor is used to determine the attitude in the dead position positioning result, and the wheel speed meter is used to determine the position in the dead position positioning result. At the last moment, the state error corresponding to the positioning result of the inertial navigation positioning sensor at the last moment is generated by the method shown in the above step 302, and the state error includes gyro zero bias and table zero bias. The positioning data of the inertial navigation positioning sensor is compensated based on the gyro zero bias and table zero bias, and the compensated positioning data is integrated to obtain the initial attitude Att_dr. Based on the initial attitude, the attitude conversion matrix Cbn_dr is generated. The attitude misalignment angle in the above state error is obtained to obtain the first attitude misalignment angle, the azimuth misalignment angle in the first attitude misalignment angle is set to zero, the horizontal misalignment angle is retained, and the second attitude misalignment angle is obtained. Based on the second attitude misalignment angle, the attitude conversion matrix is ​​compensated to obtain the attitude conversion matrix Cbn_dr_re, and the initial attitude Att_dr is multiplied by the attitude conversion matrix Cbn_dr_re to obtain the attitude Att_dr_re of the target vehicle in the navigation coordinate system.

[0179] In addition, when the inertial navigation positioning sensor is installed on the target vehicle (carrier), the orientation of the inertial navigation positioning sensor needs to be parallel to the center axis of the target vehicle and remain horizontal. However, in the actual installation process, it is difficult to ensure that the orientation of the inertial navigation positioning sensor is completely parallel to or completely horizontal with the center axis of the target vehicle. It is often the case that the inertial navigation positioning sensor is slightly tilted or installed crookedly. In this case, the inertial navigation positioning sensor's own coordinate system does not completely overlap with the carrier coordinate system. Since the coordinate system corresponding to the wheel speed meter is the carrier coordinate system, and the inertial navigation positioning sensor's own coordinate system does not completely overlap with the carrier coordinate system, when converting the wheel speed meter's positioning data based on the first conversion relationship, the installation angle of the inertial navigation positioning sensor also needs to be considered.

[0180] Based on the second attitude misalignment angle, the positioning data of the wheel speed meter is compensated, and the dead reckoning is performed based on the compensated positioning data to obtain the position in the dead reckoning positioning result, including: acquiring a first conversion relationship; generating a second conversion relationship based on the installation deflection angle of the inertial navigation positioning sensor, the second conversion relationship being used to represent the direction of the inertial navigation positioning sensor relative to the navigation coordinate system; based on the first conversion relationship and the second conversion relationship, the positioning data of the wheel speed meter is converted to obtain the converted positioning data; and the dead reckoning is performed based on the converted positioning data to obtain the position of the target vehicle in the navigation coordinate system.

[0181] Optionally, the first conversion relationship is expressed as a posture conversion matrix, and the second conversion relationship is expressed as a direction cosine matrix.

[0182] Next, take the second positioning sensor including the inertial navigation positioning sensor and the wheel speed meter as an example. Figure 6 The following example illustrates the process of determining the position in the dead position positioning result:

[0183] Get the wheel speed v output by the tachometer, and compensate the wheel speed based on the scale coefficient error of the tachometer (which can be obtained from the state error corresponding to the positioning result of the inertial navigation positioning sensor), and get V_odom=[0,(1+ )*v,0] T , V_odom is the wheel speed after compensation. Wherein, V_odom is a scalar. Based on the attitude conversion matrix Cbn_dr_re and the direction cosine matrix Cbv, V_odom can be decomposed into the velocity vector V_calib in the three directions of northeast, northeast and sky, such as: V_calib=Cbn_dr_re*Cbv*V_odo-m. Integrating V_calib can get the distance vector, which is superimposed on the position in the last determined dead position positioning result to get the position in this dead position positioning result.

[0184] Exemplarily, the installation deflection angle includes a pitch installation deflection angle , Roll installation angle and heading installation deflection , the direction cosine matrix is ​​expressed as: .

[0185] S602: Predict a first positioning result of a first positioning sensor based on a change in a current dead position positioning result and a change in a historical dead position positioning result.

[0186] In an embodiment shown, the positioning frequency of the first positioning sensor is different from the positioning frequency of the second positioning sensor. Therefore, when the dead position positioning result is obtained, there is not necessarily a positioning requirement of the first positioning sensor. In the embodiment of the present application, when the first positioning sensor has a positioning requirement, the above steps S601 to S602 can be performed.

[0187] In one embodiment, the first positioning result of the current positioning of the first positioning sensor can be predicted based on the change of the positioning result of the last positioning of the first positioning sensor and the positioning time of the current positioning. When the first positioning sensor was last positioned, the first positioning result of the first positioning sensor and the second positioning result of the second positioning sensor were fused to obtain a fused positioning result. Therefore, the fused positioning result can be regarded as the positioning result corresponding to the last positioning time of the first positioning sensor. Optionally, based on the change of the current positioning result and the historical positioning result, the first positioning result of the first positioning sensor is predicted, including: summing the difference between the current positioning result and the previous positioning result with the fused positioning result corresponding to the previous moment to obtain the reference positioning result of the first positioning sensor. Among them, the reference positioning result of the first positioning sensor can be used as the initial positioning result of the first positioning sensor in the process of calculating the first positioning result based on the collected positioning data.

[0188] Among them, the last moment is the positioning moment of the first positioning sensor's last positioning. In this way, dead reckoning and positioning fusion form a closed loop, such as Figure 8 shown.

[0189] Optionally, when the first positioning sensor collects positioning data, a corresponding signal can be transmitted to trigger the execution of the step of "determining the dead position positioning result of the second positioning sensor at the current moment, and determining the dead position positioning result of the second positioning sensor at a historical moment; based on the change of the dead position positioning result at the current moment and the dead position positioning result at a historical moment, predicting the first positioning result of the first positioning sensor".

[0190] Optionally, when the acquisition time of the last fused positioning result reaches a preset time, the step of "determining the dead position positioning result of the second positioning sensor at the current moment, and determining the dead position positioning result of the second positioning sensor at a historical moment; predicting the first positioning result of the first positioning sensor based on the change of the dead position positioning result at the current moment and the dead position positioning result at a historical moment" can be performed. The preset time is determined based on the positioning frequency of the first positioning sensor, and when the acquisition time of the last fused positioning result reaches the preset time, the first positioning sensor performs positioning again.

[0191] In one embodiment, the dead reckoning algorithm can calculate the attitude, position and speed. However, in the embodiment of the present application, only the attitude and position are required when determining the reference positioning result. Therefore, in the embodiment of the present application, when performing dead reckoning, the speed may not be calculated; or, when determining the reference positioning result, the speed in the dead reckoning result is not considered. The reference positioning result is six-dimensional data, including a three-dimensional attitude and a three-dimensional position.

[0192] The dead reckoning method provided in the embodiment of the present application predicts the first positioning result based on the change of the dead position positioning result. In this way, only the change amount of the dead position positioning result needs to be accurate, and the requirement for the overall accuracy of the dead position positioning result is not high, which can avoid the influence of the accumulated error in the process of determining the dead position positioning result.

[0193] Moreover, the partial misalignment angle is determined based on the type of the first positioning sensor and is used to enhance the positioning accuracy of the dead position positioning result in the first positioning direction, and the first positioning direction is the direction in which the positioning accuracy of the first positioning sensor is relatively poor. Considering the partial misalignment angle of the second positioning sensor in the dead reckoning process can improve the positioning accuracy of the dead position positioning result in the first positioning direction, so that when predicting the first positioning result based on the change in the dead position positioning result, the positioning accuracy of the first positioning result in the first positioning direction can be improved, and the target positioning direction is the positioning direction in which the first positioning sensor itself has a relatively high positioning accuracy, and there is no need to further improve it based on the dead position positioning result. It can be seen from this that when predicting the first positioning result based on the change in the dead position positioning result, only considering the partial misalignment angle of the second positioning sensor not only makes the first positioning result have a higher positioning accuracy in the target positioning direction, but also improves the positioning accuracy of the first positioning result in the first positioning direction.

[0194] It should be noted that the user authority information, user information (including but not limited to user login account, user login password, location, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this manual are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0195] Corresponding to the above method embodiment, the present application also provides an embodiment of a fusion positioning device.

[0196] Fig. 9 is a schematic diagram of a vehicle structure according to an exemplary embodiment of the present application. Fig. 9 At the hardware level, the vehicle includes a processor 902, an internal bus 904, a network interface 906, a memory 908, and a non-volatile memory 910, and may also include hardware required for other services. The processor 902 reads the corresponding computer program from the non-volatile memory 910 into the memory 908 and then runs it. Of course, in addition to the software implementation, this application does not exclude other implementations, such as logic devices or a combination of software and hardware, etc., that is, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0197] Fig.10 is a block diagram of a fusion positioning device according to an exemplary embodiment of the present application. Fig.10 The device includes an acquisition module 1001, an extraction module 1002, and a fusion module 1003, wherein:

[0198] An acquisition module 1001 is used to acquire a first positioning result determined for a target vehicle by a first positioning sensor, and to acquire a second positioning result determined for the target vehicle by a second positioning sensor, wherein the first positioning sensor and the second positioning sensor are of different types;

[0199] An extraction module 1002 is used to extract target positioning information in a target positioning direction based on the first positioning result, where the target positioning direction is determined based on the type of the first positioning sensor, and the target positioning direction is a part of all positioning directions corresponding to the first positioning result;

[0200] The fusion module 1003 is used to fuse the target positioning information with the second positioning result to obtain a fused positioning result of the target vehicle.

[0201] In one embodiment, all positioning directions corresponding to the first positioning result include directions indicated by at least two coordinate axes constituting an N-dimensional coordinate system, where N is an integer not less than 2;

[0202] or,

[0203] The first positioning sensor is a visual positioning sensor, all positioning directions corresponding to the first positioning result include horizontal, vertical and elevation directions, and the target positioning direction is the horizontal direction;

[0204] The horizontal direction is the direction indicated by the X-axis constituting the three-dimensional coordinate system, the vertical direction is the direction indicated by the Y-axis constituting the three-dimensional coordinate system, and the elevation direction is the direction indicated by the Z-axis constituting the three-dimensional coordinate system.

[0205] In one embodiment, the extraction module 1002 is used to predict the overall error of the second positioning result in the target positioning direction based on the first positioning result, and the overall error is used to represent the difference between the second positioning result and the actual positioning result of the target vehicle in the target positioning direction; determine the target positioning information based on the position error in the overall error; or

[0206] The extraction module 1002 is used to determine the target positioning information based on the positioning result in the target positioning direction in the first positioning result.

[0207] like Fig.11 As shown, in one embodiment, the device further includes:

[0208] The determination module 1004 is used to determine the dead position positioning result of the second positioning sensor at the current moment, and determine the dead position positioning result of the second positioning sensor at the historical moment;

[0209] The prediction module 1005 is used to predict the first positioning result of the first positioning sensor based on the change of the current dead position positioning result and the dead position positioning results at historical moments.

[0210] In one embodiment, the determination module 1004 is used to determine a second attitude misalignment angle based on a partial misalignment angle in the first attitude misalignment angle of the second positioning sensor, where the partial misalignment angle is determined based on the type of the first positioning sensor; and perform dead reckoning based on the positioning data of the second positioning sensor and the second attitude misalignment angle to obtain the dead position positioning result.

[0211] In one embodiment, the first positioning sensor is a visual sensor, and the partial misalignment angle is a horizontal misalignment angle in the first posture misalignment angle.

[0212] In one embodiment, the second positioning sensor includes an inertial navigation positioning sensor, and the determination module 1004 is used to perform dead reckoning based on the positioning data of the inertial navigation positioning sensor to obtain an initial dead position positioning result; based on the second attitude misalignment angle, the initial dead position positioning result is compensated to obtain a compensated dead position positioning result;

[0213] or,

[0214] The second positioning sensor includes an inertial navigation positioning sensor and a third positioning sensor, and the determination module 1004 is used to perform dead reckoning based on the positioning data of the inertial navigation positioning sensor to obtain an initial attitude; based on the second attitude misalignment angle, compensate the initial attitude to obtain the attitude in the dead position positioning result; based on the second attitude misalignment angle, compensate the positioning data of the third positioning sensor; and perform dead reckoning based on the compensated positioning data to obtain a position in the dead position positioning result.

[0215] In one embodiment, the initial posture is the posture of the target vehicle in the carrier coordinate system; the determination module 1004 is used to determine a first transformation relationship based on the initial posture and the second posture misalignment angle, and the first transformation relationship is used to represent the transformation relationship between the navigation coordinate system of the target vehicle and the carrier coordinate system; the initial posture is transformed using the first transformation relationship to obtain the posture of the target vehicle in the navigation coordinate system.

[0216] In one embodiment, the first conversion relationship is represented as a posture conversion matrix; the determination module 1004 is used to generate a posture conversion matrix based on the initial posture; based on the second posture misalignment angle, the posture conversion matrix is ​​compensated to obtain a compensated posture conversion matrix; or, based on the second posture misalignment angle, the initial posture is compensated to obtain a compensated posture; based on the compensated posture, a posture conversion matrix is ​​generated.

[0217] The implementation process of the functions and effects of each module in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, which will not be repeated here.

[0218] The devices or modules described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a vehicle, which may be a car, a bus, a truck, or the like.

[0219] In a typical configuration, the vehicle includes one or more processors, including a central processing unit (CPU) and a graphics processing unit (GPU), input / output interfaces, network interfaces, and memory. The CPU is used for computational simulation, and the GPU is used for outputting high-quality 3D images.

[0220] Memory may include non-permanent storage in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0221] Computer-readable media include permanent and non-permanent, removable and non-removable media that can be used to store information by any method or technology. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, disk storage, quantum memory, graphene-based storage media or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0222] Corresponding to the embodiments of the aforementioned fusion positioning method, the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps of any embodiment of the aforementioned fusion positioning method are implemented.

[0223] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A fusion positioning method, characterized in that: The method comprises: Obtaining a first positioning result determined for a target vehicle by a first positioning sensor, and obtaining a second positioning result determined for the target vehicle by a second positioning sensor, wherein the first positioning sensor and the second positioning sensor are of different types, and all positioning directions corresponding to the first positioning result include directions respectively indicated by at least two coordinate axes constituting an N-dimensional coordinate system, where N is an integer not less than 2; Extracting target positioning information in a target positioning direction based on the first positioning result, where the target positioning direction is determined based on the type of the first positioning sensor, and the target positioning direction is part of all positioning directions corresponding to the first positioning result, and a positioning accuracy of the first positioning sensor in the target positioning direction is higher than a positioning accuracy in at least one other positioning direction except the target positioning direction among all positioning directions; Correcting the second positioning result based on the target positioning information to obtain a fused positioning result of the target vehicle; The extracting target positioning information in the target positioning direction based on the first positioning result includes: The state error amount of the second positioning result is predicted based on the first positioning result by a filtering algorithm, and the position error in the target positioning direction in the state error amount is determined as the target positioning information.

2. The method according to claim 1, characterized in that The first positioning sensor is a visual positioning sensor, all positioning directions corresponding to the first positioning result include horizontal, vertical and elevation directions, and the target positioning direction is the horizontal direction; The horizontal direction is the direction indicated by the X-axis constituting the three-dimensional coordinate system, the vertical direction is the direction indicated by the Y-axis constituting the three-dimensional coordinate system, and the elevation direction is the direction indicated by the Z-axis constituting the three-dimensional coordinate system.

3. The method according to any one of claims 1 to 2, characterized in that: The obtaining of a first positioning result determined by the first positioning sensor for the target vehicle further includes: Determine the dead position positioning result of the second positioning sensor at the current moment, and determine the dead position positioning result of the second positioning sensor at a historical moment; Based on the change of the dead position positioning result at the current moment and the dead position positioning results at the historical moments, a first positioning result of the first positioning sensor is predicted.

4. The method according to claim 3, characterized in that The dead position positioning result is obtained by: determining a second attitude misalignment angle based on a partial misalignment angle in the first attitude misalignment angle of the second positioning sensor, wherein the partial misalignment angle is determined based on a type of the first positioning sensor; Dead reckoning is performed based on the positioning data of the second positioning sensor and the second attitude misalignment angle to obtain the dead position positioning result.

5. The method according to claim 4, characterized in that The first positioning sensor is a visual sensor, and the partial misalignment angle is a horizontal misalignment angle in the first posture misalignment angle.

6. The method according to claim 4, characterized in that The performing dead reckoning based on the positioning data of the second positioning sensor and the second attitude misalignment angle to obtain the dead position positioning result includes: The second positioning sensor includes an inertial navigation positioning sensor, and dead reckoning is performed based on positioning data of the inertial navigation positioning sensor to obtain an initial dead position positioning result; based on the second attitude misalignment angle, the initial dead position positioning result is compensated to obtain a compensated dead position positioning result; or, The second positioning sensor includes an inertial navigation positioning sensor and a third positioning sensor. Dead reckoning is performed based on the positioning data of the inertial navigation positioning sensor to obtain an initial attitude. The initial attitude is compensated based on the second attitude misalignment angle to obtain the attitude in the dead position positioning result. The positioning data of the third positioning sensor is compensated based on the second attitude misalignment angle. Dead reckoning is performed based on the compensated positioning data to obtain the position in the dead position positioning result.

7. The method according to claim 6, characterized in that The initial posture is the posture of the target vehicle in the carrier coordinate system; the initial posture is compensated based on the second posture misalignment angle to obtain the posture in the dead position positioning result, including: Based on the initial posture and the second posture misalignment angle, determining a first conversion relationship, wherein the first conversion relationship is used to represent a conversion relationship between a navigation coordinate system of the target vehicle and a carrier coordinate system; The initial posture is transformed using the first transformation relationship to obtain the posture of the target vehicle in the navigation coordinate system.

8. The method according to claim 7, characterized in that The first conversion relationship is represented as a posture conversion matrix, and the first conversion relationship is determined based on the initial posture and the second posture misalignment angle, including: Based on the initial posture, a posture conversion matrix is ​​generated; based on the second posture misalignment angle, the posture conversion matrix is ​​compensated to obtain a compensated posture conversion matrix; or, The initial posture is compensated based on the second posture misalignment angle to obtain a compensated posture; and a posture conversion matrix is ​​generated based on the compensated posture.

9. A vehicle, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor implements the method according to any one of claims 1 to 8 by running the executable instructions.

10. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the instructions are executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

11. A computer program product having a computer program / instructions stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images