Multi-hypothesis map matching positioning method
By employing a multi-hypothesis map matching method, the probability of position assumptions is calculated using the perpendicular foot curve of the GNSS error ellipse and vehicle speed. This solves the problem of inaccurate map matching caused by GNSS positioning errors and achieves high-precision positioning in complex road scenarios.
Patent Information
- Application Number
- CN202511727511.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-24
- Publication Date
- 2026-02-10
AI Technical Summary
Existing technologies fail to effectively account for GNSS positioning errors in different directions, resulting in low map matching accuracy, especially prone to mismatches in complex road scenarios.
A multi-hypothesis map matching positioning method is adopted. By acquiring vehicle motion information and electronic maps, the position covariance matrix of GNSS measurement points is calculated. Map matching is performed using the foot curve of the error ellipse, and vehicle running speed is fused to calculate the probability of multiple position hypotheses. The position hypotheses are periodically iterated and maintained.
It improves the accuracy of map matching, can quickly identify and eliminate incorrect location assumptions, and obtain accurate vehicle positioning results.
Smart Images
Figure CN121498722A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of vehicle positioning technology, and more specifically to a multi-hypothesis map matching positioning method. Background Technology
[0002] Vehicle positioning provides crucial information support for the perception, planning, and control of intelligent vehicles. Map matching, as a common auxiliary positioning method, utilizes electronic maps and positioning information from sensors to determine the vehicle's position on the road network, which can improve positioning accuracy to a certain extent. Satellite positioning is one of the main ways to obtain the absolute position of a vehicle, and combining it with other sensors (wheel speed sensors, inertial sensors, etc.) enhances overall positioning performance.
[0003] In certain complex road scenarios, such as exiting a ramp, map matching based on the current location information is prone to errors due to the uncertainty of location and the limitations of general map matching algorithms, thus seriously affecting subsequent driving decisions. Specifically: (1) GNSS-based map matching does not take into account the measurement error of the positioning location in different directions, resulting in low matching accuracy; (2) By matching the current measurement location with the electronic map, vehicles are easily matched to the wrong road segment in dense and intersecting road segments. Summary of the Invention
[0004] The technical problem this invention aims to solve is that existing technologies do not consider GNSS positioning errors in different directions, resulting in low accuracy of the obtained map matching points; the accuracy of map matching is also lower at densely packed roads and intersections.
[0005] To address this, the present invention provides a multi-hypothesis map matching and positioning method that considers the error distribution characteristics of GNSS measurement points in different directions, adopts a probability-based map matching approach, allows for the existence of multiple map matching hypotheses, and improves map matching performance.
[0006] The technical solution adopted by this invention to solve its technical problem is: A multi-hypothesis map matching and localization method includes the following steps: S1, acquire vehicle movement information and read electronic maps; S2, performs map matching on the locations of GNSS measurement points; S3 integrates vehicle speed with map matching results and calculates multiple location hypotheses and their corresponding probabilities; S4 periodically iterates and maintains the probabilities of multiple location hypotheses. When the probability of a hypothesis drops to a set threshold, the corresponding location hypothesis is eliminated.
[0007] To avoid obtaining erroneous map matching results, this scheme allows for the existence of multiple location assumptions and calculates their respective probabilities, thereby improving the accuracy of map matching for vehicles in complex road scenarios.
[0008] Furthermore, in S1, the vehicle motion information includes latitude, longitude, elevation, timestamp, etc. collected by the GNSS receiver, and vehicle speed, timestamp, etc. collected by the wheel speed sensor.
[0009] Furthermore, in S1, the electronic map contains information such as the precise location of the vehicle's travel route and the coordinates of key points.
[0010] Furthermore, step S2 specifically includes the following steps: S201 calculates the position covariance matrix of GNSS measurement points: Based on the spatial distribution of satellites and pseudorange measurement errors, the covariance matrix is calculated to obtain the standard deviation of GNSS measurement points in the latitude direction. Standard deviation of GNSS measurement points in the longitude direction and the measurement covariance of GNSS measurement points in latitude and longitude. ; S202 uses the perpendicular bisector curve to obtain GNSS map matching points: the two-dimensional error distribution based on GNSS measurement points can be represented by an error ellipse, through... , An error ellipse is established, and the perpendicular curve corresponding to the error ellipse is used for map matching to obtain the map matching point of the current GNSS measurement point on the road segment in the electronic map.
[0011] By adopting the above technical solution and applying the perpendicular curve of the error ellipse to map matching, the measurement error in the direction of the map matching point can be accurately obtained, thereby more accurately calculating the probability of a vehicle being located on that road segment and improving the accuracy of map matching.
[0012] Furthermore, the process of obtaining GNSS map matching points using the perpendicular curve is as follows: by The major semi-axis of the error ellipse for GNSS measurement points is used as the reference. The length of the minor semi-axis of the error ellipse is taken as the length of the error ellipse. A tangent line is drawn to the ellipse from any point on the error ellipse. A perpendicular line is drawn from the origin (GNSS measurement point) to the tangent line. The trajectory of the foot of the perpendicular line forms the foot curve of the error ellipse. The perpendicular curve is scaled up or down proportionally until it is tangent to the electronic map of the road segment in the neighborhood of the GNSS measurement point. The tangency point is the map matching point of the current GNSS measurement point on that road segment. The positioning standard deviation of the map matching point direction can be calculated using the following formula:
[0013] in, The angle between the line connecting the GNSS measurement point and the map matching point and the minor semi-axis of the ellipse; the standard deviation of the map matching point's orientation. .
[0014] Meanwhile, the coordinates of the map matching point can be calculated and determined based on the coordinates of the map measurement points on both sides of the map matching point.
[0015] Furthermore, step S3 specifically includes the following steps: S301 merges GNSS map matching points with vehicle speed to obtain the merged vehicle position; S302 calculates the probability that the vehicle is located on all possible road segments based on the fused vehicle position, GNSS measurement points, vehicle speed, etc., where: Whether a vehicle is likely to be located on a certain road segment is determined by the distance between the GNSS measurement point and its map matching point on the electronic map of that road segment. If the distance is less than a set threshold, the vehicle is likely to be located on that road segment, indicating that the location assumption is valid. The probability of the vehicle being located on that road segment is calculated. S303 compares the conditional probabilities of all the location assumptions and takes the road segment corresponding to the location assumption with the highest probability as the road segment where the vehicle is located.
[0016] When the traffic network is complex, a large number of location assumptions exist at the same time. However, as vehicles continue to move, the probability of each erroneous location assumption will decrease rapidly, so as to achieve the technical effect of eliminating erroneous location assumptions.
[0017] The beneficial effects of this invention are: 1. By using the perpendicular curve of the GNSS error ellipse for map matching, the measurement error in the direction of the map matching point can be obtained, and the probability of a vehicle being located on the corresponding assumed road segment can be calculated more accurately, thereby improving the accuracy of map matching. 2. By periodically iterating and maintaining the probabilities of multiple location hypotheses, incorrect location hypotheses can be quickly identified and eliminated, thereby obtaining accurate vehicle positioning results. Attached Figure Description
[0018] The present invention will be further described below with reference to the accompanying drawings and embodiments.
[0019] Figure 1 This is a flowchart of the positioning method of the present invention; Figure 2 This is a schematic diagram of map matching based on the perpendicular bisector curve of the present invention; Figure 3 This is a schematic diagram of the probability distribution of the distance between GNSS measurement points and road sections in this invention. Detailed Implementation
[0020] The invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention and therefore showing only the components relevant to the invention. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values of the components and steps set forth in these embodiments do not limit the scope of this disclosure.
[0021] It should also be understood that, for ease of description, the dimensions of the various parts shown in the accompanying drawings are not drawn to actual scale. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit this disclosure or its application or use.
[0022] Techniques, methods, and apparatus known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and apparatus should be considered part of the specification. In all examples shown and discussed herein, any specific values should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.
[0023] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, features defined with "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.
[0024] It should be noted that similar labels and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be discussed further in subsequent figures.
[0025] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.
[0026] Vertical projection is a common map matching algorithm that can quickly find map matching points, but it cannot obtain the positioning error of the GNSS measurement point in the direction of that matching point. This application provides a multi-hypothesis map matching positioning method that considers direction-dependent GNSS measurement errors in map matching and uses the perpendicular curve of the GNSS measurement error ellipse for map matching. The perpendicular curve can be used to represent the positioning measurement error in any direction. Using the perpendicular curve for map matching allows for both obtaining the map matching point by zooming in or out of the perpendicular curve and obtaining the error in the direction of that matching point, thereby more accurately calculating the probability of a vehicle being located on that road segment and improving the accuracy of map matching.
[0027] Reference Figure 1 A multi-hypothesis map matching and localization method includes the following steps: S1, acquire vehicle motion information sensed by onboard sensors. Vehicle motion information includes latitude, longitude, elevation, speed, timestamp, etc. collected by the onboard GNSS receiver, as well as vehicle running speed, timestamp, etc. collected by the wheel speed sensor. The system reads the electronic map stored in the vehicle controller, which contains the precise location of the vehicle's route and the coordinates of key points.
[0028] S2, performs map matching on the locations of GNSS measurement points; S201 Calculates and obtains the position covariance matrix of the GNSS measurement points: Optionally, let the covariance matrix of the GNSS measurement points be... for:
[0029] in, This represents the standard deviation of GNSS measurement points along the latitude direction. This represents the standard deviation of GNSS measurement points along the longitude direction. This represents the measurement covariance of GNSS measurement points in latitude and longitude.
[0030] The covariance matrix is calculated based on the spatial distribution of satellites and pseudorange measurement errors. When performing pseudorange point positioning, the GNSS receiver satisfies the least squares matrix expression commonly used in GNSS positioning calculations (existing technology):
[0031] GNSS measurement point coordinates are represented in the northeast-northeast coordinate system, with east as the orientation. x The axis is north. y Axis, celestial direction is z Axis, error vector , c At the speed of light, For receiver clock bias, For the first n The pseudorange measurement error of the satellite, A This is the pseudorange observation matrix, reflecting the linear relationship between pseudorange observations and the parameters to be determined. It is assumed that the pseudorange errors of each satellite are equal, independent, and satisfy zero mean and variance. If the error vector follows a Gaussian distribution, then the covariance matrix (based on current technology, reflecting the correlation between different parameter estimation errors and the variance of individual parameter errors) is expressed as:
[0032] Among them, the main diagonal elements ( , (etc.) correspond to the receivers respectively. x , y , z The variance of orientation and position errors and the variance of clock errors; non-main diagonal elements ( , (etc.) is the covariance between different errors, reflecting the degree of correlation between two errors.
[0033] Under the northeast celestial coordinate system
[0034] in, Let be the elevation angle of the nth satellite. The satellite azimuth angle can be obtained from the position measurement information output by the GNSS receiver. The position covariance matrix of the GNSS measurement point can be obtained by taking the first two rows and two columns of the covariance matrix of the error vector.
[0035] It should be noted that the position covariance matrix of GNSS measurement points describes the two-dimensional error distribution of GNSS.
[0036] S202 uses the foot of the perpendicular curve to obtain GNSS map matching points: the two-dimensional error distribution of GNSS measurement points can be represented by an error ellipse.
[0037] This application uses the perpendicular curve corresponding to the error ellipse for map matching and determines the standard deviation of the direction of the map matching point. , specifically: The semi-major axis of the error ellipse of the GNSS measurement point is The length of the minor semi-axis is If a tangent line is drawn from any point on the error ellipse, and a perpendicular line is drawn from the origin (GNSS measurement point) to the tangent line, the trajectory of the foot of the perpendicular forms the foot curve of the error ellipse. like Figure 2 As shown, the solid curve represents the GNSS measurement point. The error ellipse perpendicular foot curve, after being enlarged, is compared with the road segment. Tangent to , That is, the current GNSS measurement point is at Map matching points; The standard deviation of the orientation can be calculated using the following formula:
[0038] in, GNSS measurement points Matching points on the map The angle between the line connecting the two points and the minor axis of the ellipse; Standard deviation of orientation ; The coordinates can be based on Matching points on the map coordinates of measurement points on both sides of the map , Calculation determined.
[0039] S3 integrates vehicle speed with map matching results and calculates the probability of multiple location hypotheses; The S301 integrates the GNSS map location with the vehicle speed obtained from wheel speed sensors to obtain the fused vehicle position: Optionally, a CTRA (Constant Turn Rate and Acceleration) motion model is adopted, and an EKF (Extended Kalman Filter) is used to fuse and estimate the vehicle state. The system state of the EKF is... These represent the vehicle's lateral position, longitudinal position, velocity direction, velocity, and acceleration, respectively. The measurement information used in the fusion process includes the lateral position, longitudinal position, and wheel speed of the GNSS map matching points.
[0040] It should be noted that using the CTRA-based EKF algorithm to obtain vehicle location is a relatively mature technical method, and this application will not elaborate on it in the embodiments.
[0041] Based on the merged lateral and longitudinal positions of the vehicle, the nearest point is found on the electronic map of each road segment within the neighborhood of the GNSS measurement point, and this point is used to determine the vehicle's position on that road segment. , The location is determined by calculating the positions of map measurement points on both sides of the map matching point.
[0042] S302 calculates the probability that a vehicle is located on all possible road segments based on the merged vehicle location, GNSS measurement points, vehicle speed, etc.: Whether a vehicle is likely to be located on a certain road segment is determined by the distance between the GNSS measurement point and its map matching point on the map of that road segment. If the distance is less than a set threshold, the vehicle is likely to be located on that road segment (the location assumption is valid), and the probability of the vehicle being located on that road segment is calculated.
[0043] The following two scenarios illustrate the probability calculation of location assumptions: Scenario 1: If the vehicle is not on a certain road segment for the first time, then at the given GNSS measurement point coordinates... Vehicle speed measurement Under these conditions, the vehicle is currently in the road segment and vehicles in The position above is The joint conditional probability is:
[0044] Based on the location hypothesis With position It can directly determine whether the vehicle is located Above, that is The value is 0 or 1, which indicates whether the road segment matches the location assumption.
[0045] It should be noted that the location assumption Indicates that the vehicle is located on the road section superior; Indicates that the vehicle is located on the road section The specific location on it.
[0046] Section Not referring to a specific road segment; road segment Indicates a specific road segment.
[0047] If the vehicle is located Above, assuming the vehicle's position The coordinates of the map endpoints on both sides are respectively and The positions are respectively and , Represented as:
[0048] in, , , , , .
[0049] Assuming at a given location Under these conditions, the distance from the GNSS measurement point to the corresponding road segment is assumed to follow a Gaussian distribution, such as... Figure 3 The probability density curve in the figure is shown as follows:
[0050] in, This represents the distance between the GNSS measurement point and the map matching point.
[0051] Location assumption Prior probability Based on the posterior probability of the previous period Replace. Thus, the calculation yields all the hypotheses. Afterwards, for each After normalization, it is then included in the calculation of the joint conditional probability.
[0052] Scenario 2: If the vehicle is first located on a certain road segment Above, the vehicle's initial position on this road segment is determined by GNSS map matching points, and the corresponding position conditional probability is:
[0053] The right side of the above formula , The specific calculation process has been given in Scenario 1. Indicates positional assumption and road section The correspondence is either 0 or 1.
[0054] S303. Compare the conditional probabilities of all the hypotheses and take the road segment corresponding to the location with the highest probability as the road segment where the vehicle is located.
[0055] S4 involves periodically iterating and maintaining the probabilities of multiple location hypotheses. When the probability of a hypothesis drops to a set threshold, the corresponding location hypothesis is eliminated. Specifically: When the traffic network is complex, a large number of location assumptions are generated. As vehicles move, GNSS measurements gradually move away from most of these assumptions, and the probability of the corresponding location assumptions decreases rapidly. When the probability of an assumption drops to a set threshold, the corresponding assumption is eliminated to improve computational efficiency.
[0056] This concludes the detailed description of a multi-hypothesis map matching and localization method according to this disclosure. To avoid obscuring the concept of this disclosure, some details known in the art have not been described. Those skilled in the art will fully understand how to implement the technical solutions disclosed herein based on the above description.
[0057] Based on the above-described preferred embodiments of the present invention, and through the foregoing description, those skilled in the art can make various changes and modifications without departing from the inventive concept. The technical scope of this invention is not limited to the contents of the specification, but must be determined by the scope of the claims.
Claims
1. A multi-hypothesis map matching and localization method, characterized in that, Includes the following steps: S1, acquire vehicle movement information and read electronic maps; S2, use the foot curve of the GNSS measurement error ellipse to perform map matching on the location of GNSS measurement points; S3, merge the vehicle speed from the vehicle motion information with the map matching results to obtain the merged vehicle position, and calculate the probability of multiple position hypotheses; S4 periodically iterates and maintains the probabilities of multiple location hypotheses. When the probability of a hypothesis drops to a set threshold, the corresponding location hypothesis is eliminated.
2. The multi-hypothesis map matching and positioning method according to claim 1, characterized in that, S2 specifically includes the following steps: S201 calculates the position covariance matrix of GNSS measurement points: Based on the spatial distribution of satellites and pseudorange measurement errors, the covariance matrix is calculated to obtain the standard deviation of GNSS measurement points in the latitude direction. Standard deviation of GNSS measurement points in the longitude direction and the measurement covariance of GNSS measurement points in latitude and longitude. ; S202 uses perpendicularity curves to obtain GNSS map matching points: based on , Error ellipses for GNSS measurement points are established as the minor and major semi-axes, respectively. The perpendicular foot curves corresponding to the error ellipses are used for map matching to obtain the map matching points of the current GNSS measurement points on the road segments in the electronic map.
3. The multi-hypothesis map matching and positioning method according to claim 2, characterized in that, The process of obtaining GNSS map matching points using the aforementioned perpendicular curve is as follows: Draw a tangent line to the ellipse from any point on the error ellipse, and draw a perpendicular line from the GNSS measurement point (which is the origin of the coordinate system) to the tangent line. The trajectory of the foot of the perpendicular forms the foot curve of the error ellipse. The perpendicular curve is scaled up or down proportionally until it is tangent to the electronic map of the road segment within the neighborhood of the GNSS measurement point. The tangency point is the map matching point of the current GNSS measurement point on that road segment; the positioning standard deviation of the map matching point direction. It can be calculated using the following formula: in, The angle between the line connecting the GNSS measurement point and the map matching point and the minor semi-axis of the ellipse.
4. The multi-hypothesis map matching and positioning method according to claim 3, characterized in that, The coordinates of the map matching point can be calculated and determined based on the coordinates of the map measurement points on both sides of the map matching point.
5. The multi-hypothesis map matching and positioning method according to claim 4, characterized in that, S3 specifically includes the following steps: S301 merges GNSS map matching points with vehicle speed to obtain the merged vehicle position; S302 calculates the probability that the vehicle is located on all possible road segments based on the fused vehicle position, GNSS measurement points, and vehicle speed, where: Whether a vehicle is likely to be located on a certain road segment is determined by the distance between the GNSS measurement point and its map matching point on the electronic map of that road segment. If the distance is less than a set threshold, the vehicle is likely to be located on that road segment, the location assumption is valid, and the probability of the vehicle being located on that road segment is calculated. S303 compares the conditional probabilities of all the location assumptions and takes the road segment corresponding to the location assumption with the highest probability as the road segment where the vehicle is located.
6. The multi-hypothesis map matching and positioning method according to claim 5, characterized in that, The distance from the GNSS measurement point to the location is assumed to follow a Gaussian distribution.
7. The multi-hypothesis map matching and positioning method according to claim 1, characterized in that, In S1, the vehicle motion information includes latitude, longitude, elevation, and timestamp collected by the GNSS receiver.
8. The multi-hypothesis map matching and positioning method according to claim 1, characterized in that, The electronic map contains the precise location of the vehicle's route and the coordinates of key points.
9. The multi-hypothesis map matching and positioning method according to claim 8, characterized in that, The electronic map is pre-stored in the vehicle controller.
10. The multi-hypothesis map matching and positioning method according to claim 7, characterized in that, The vehicle's operating speed is collected by wheel speed sensors.