A road selection judgment method based on prior information

Abstract

The present invention proposes a method for judging road selection based on prior information, comprising the following steps: step 1, initialization, obtaining the vehicle position and the vehicle heading angle according to the combined navigation, and calculating the horizontal protection level; step 2, determining whether to make a road selection judgment based on the prior information according to the vehicle state and the map data; step 3, making a road selection judgment according to the prior information. The present invention adopts a map matching algorithm with real-time computing capability, and at the same time ensures the accuracy of map matching through a sliding window of variable size, and reduces the amount of calculation. In addition, path planning and traffic regulations are included in the consideration of map matching, and the path planning coefficient and traffic regulations coefficient are set according to the shape and size of the intersection where the vehicle is located, and the statistical violation rate, and the known information is reasonably used to help the algorithm make path planning decisions.

Description

Technical Field

[0001] The present invention relates to a method for judging road selection, in particular to a method for judging road selection based on prior information. Background Art

[0002] With the development of social modernization, cities have become the most important places for human activities. As the main application in the urban environment, intelligent transportation has a huge application market. At the same time, in recent years, intelligent transportation in the urban environment is developing towards the new and intelligent directions, and more stringent requirements are also put forward for navigation and positioning services. The method of using the device coordinates obtained by the positioning sensor together with the road network database in this area to estimate the user's position on the road section is called map matching. GNSS (Global Navigation Satellite System) is used as the basic sensor for navigation and positioning. In an open environment, it can provide high-precision spatio-temporal information for the carrier. However, in the urban canyon environment, due to the blocking and reflection of GNSS signals by various obstacles, the signal quality for positioning and calculation often cannot meet the requirements, resulting in deviations in the results output by GNSS. Moreover, due to the complex and dense road structure in the urban environment, the algorithm may also match the user's position to the wrong road, making the positioning results output by the navigation software unable to meet the actual needs.

[0003] Some data processing algorithms are post-processing map matching algorithms, which determine the vehicle position by comparing the traveled trajectory with the road structure. Such algorithms usually use a hardware with a sampling rate or a processing method with a large amount of calculation for map matching. The post-processing map matching algorithm cannot meet the requirements of navigation and positioning during vehicle driving in the city.

[0004] The current multi-sensor map matching algorithms have the disadvantages of large amount of calculation and low calculation efficiency in data processing and map matching. On devices with low performance, it will cause the map matching result to lag behind the actual vehicle position during vehicle driving in the city.

[0005] The information sources of the current map matching algorithms are the sensors for vehicle positioning and the constraints that must be met such as road topology and elevation. However, traffic regulations also exist in the urban environment, and the information on navigation path planning in the navigation software is not used. Summary of the Invention

[0006] Object of the Invention: The technical problem to be solved by the present invention is to provide a method for judging road selection based on prior information in view of the deficiencies of the prior art.

[0007] To solve the above technical problem, the present invention discloses a method for judging road selection based on prior information, including the following steps:

[0008] Step 1, initialization: Obtain the vehicle position, the heading angle of the vehicle according to integrated navigation, and calculate the horizontal protection level. The specific method for obtaining the vehicle position includes:

[0009] The integrated navigation includes: an inertial measurement unit and a Global Navigation Satellite System (GNSS); perform mechanical alignment using the measurement results of the inertial measurement unit; calculate the least squares solution of the GNSS pseudorange; combine the above results and perform an extended Kalman filter to obtain the vehicle position.

[0010] The horizontal protection level is calculated according to the following method:

[0011] HPL = max(slope i )·P bias

[0012] where HPL represents the horizontal protection level, P bias represents the minimum measurement deviation value detected under the set false alarm rate and miss detection rate, and slope i is the characteristic slope of the i-th satellite, and the calculation method is as follows:

[0013]

[0014] where Λ 1j represents the element in the first row and the j-th column of the matrix Λ, Λ 2j represents the element in the first row and the j-th column of the matrix Λ, S jj represents the element in the j-th row and the j-th column of the matrix S, n represents the number of satellites. Among them, the matrix Λ = (A T A) -1 A T , the matrix S = [I - A(A T A) -1 A T , and A is the measurement matrix.

[0015] Step 2, determine whether to perform road selection judgment based on prior information according to the vehicle state and map data. The specific method is as follows:

[0016] Step 2-1, judge the speed of the vehicle. If the vehicle speed is zero, select the current road unchanged and end the judgment; otherwise, execute Step 2-2. Among them, the vehicle speed is calculated simultaneously when obtaining the vehicle position in Step 1.

[0017] Step 2-2: Combine the road structure information in the map data to determine whether there is a road that is connected to the road where the vehicle is located and allows entry within a circular area centered on the vehicle position obtained in Step 1 with the horizontal protection level as the radius. If there is such a road, mark the road as an alternative road and execute Step 3; otherwise, re-execute Step 1.

[0018] Step 3: Make a road selection judgment based on prior information, including the following steps:

[0019] Step 3-1: Calculate the road selection weight, which specifically includes:

[0020] Let the first epoch be the starting epoch. The method for calculating the road selection weights of all alternative road segments in the first epoch is as follows:

[0021] D ml =I m ·R m ·(A·d ,l +Δθ ml )

[0022] where D ml represents the road selection weight of the m-th alternative road, A, I m and R m are hyperparameters. Extend the alternative road along the tangent direction of the alternative road at the intersection. d ml is the distance between the vehicle position and the projection of the m-th alternative road and its extension line, in meters. Δθ ml is the difference between the vehicle's heading angle and the direction angle of the m-th alternative road, in degrees.

[0023] For the hyperparameters A, I m and R m , the calculation methods are as follows:

[0024] The hyperparameter A is the position and attitude weight parameter. When the spatial position dilution of precision (PDOP) value is less than 3, set the value of the hyperparameter A to 2; when the PDOP value is greater than 3, decrease the value of the hyperparameter A and set the value of the hyperparameter A to 2e 3-Y , where Y is the PDOP value;

[0025] The hyperparameter I m is the path planning weight parameter, and is set as follows:

[0026]

[0027] The hyperparameter R m is the traffic rule weight parameter, and is set as follows:

[0028]

[0029] Among them, the path planning and traffic rules are the prior information mentioned above.

[0030] Step 3-2: Normalize the road selection weights, which specifically includes:

[0031] The method for normalizing the road selection weights is as follows:

[0032]

[0033] Among them, P ml represents the normalized road selection weight of the m-th alternative road, and M is the set of alternative roads.

[0034] Step 3-3: Perform road selection, which specifically includes:

[0035] Apply a sliding window to determine the minimum value of the normalized road selection weights in k consecutive epochs. If the road corresponding to the minimum weight item in the sliding window does not change, select this road, complete the road selection and end; if the roads corresponding to the minimum weight items in the sliding window are inconsistent, make the following judgment: Determine whether the vehicle has entered a new road based on the vehicle position. If it has entered a new road, select the new road, complete the road selection and end. Otherwise, re-perform the road selection and update P il Re-perform the road selection until the road selection is completed and ended.

[0036] The size k of the sliding window mentioned above is calculated as follows:

[0037]

[0038] Among them, fs is the sampling rate of the Global Navigation Satellite System (GNSS), and q is the number of lanes of the road with the most connected lanes at the current intersection.

[0039] Beneficial effects:

[0040] Aiming at the problem that the post-processing map matching algorithm cannot meet the navigation positioning requirements, the map matching algorithm adopted by the present invention has real-time computing capabilities.

[0041] Aiming at the disadvantages of the existing multi-sensor map matching algorithm, such as large computational amount and low computational efficiency, which may lead to the map matching result lagging behind the actual vehicle position, the present invention proposes a novel method for calculating road selection weights, and at the same time uses a variable-size sliding window to ensure the accuracy of map matching. On the premise of ensuring accuracy, the computational amount is reduced.

[0042] Regarding the selection of information sources, the present invention proposes a path planning coefficient and a traffic regulation coefficient. Incorporate path planning and traffic regulation factors into the scope of consideration for map matching, and reasonably set the path planning coefficient and the traffic regulation coefficient based on the shape, size of the intersection where the vehicle is located, and the statistically recorded violation rate. Reasonably utilize the known information to assist the algorithm in making path planning decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The following further describes the present invention in detail with reference to the drawings and specific embodiments, and the above and / or other advantages of the present invention will become clearer.

[0044] Figure 1 It is a schematic diagram of the overall process of the method of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] The core content of the present invention is a road selection and judgment algorithm based on prior information, and its overall flowchart is as Figure 1 shown. A road selection and judgment method based on prior information, and the specific steps are as follows:

[0046] 1) Algorithm initialization:

[0047] Start. Use the mechanical arrangement of the IMU (Inertial Measurement Unit, IMU) and the least squares solution of the GNSS pseudorange calculated through pseudorange observations to perform extended Kalman filtering to obtain the vehicle position (reference: Comparison of direct and indirect filtering methods for UAV integrated navigation) and calculate the horizontal protection level. Also known as the protection level, it refers to the upper bound of the confidence interval of the positioning error that can satisfy the integrity risk. Simply put, the protection level is the envelope value of the positioning error calculated in real time by the navigation system. The horizontal protection level is the protection level in the horizontal direction.

[0048] Calculation of the horizontal protection level:

[0049] The horizontal protection level can be calculated according to the following formula:

[0050] HPL = max(slope i ) · P bias

[0051] In the formula, P bias represents the minimum measurement deviation value that can be detected under the given false alarm rate and miss detection rate /

[0052] slope i is the characteristic slope corresponding to the observation value of the i-th visible star and the ratio of the horizontal error to the test statistic in the case of no fault:

[0053]

[0054] where Λ = (A T A) -1 A T , S = [I - A(A T A) -1 A T , and A is the measurement matrix

[0055] 2) Algorithm start condition judgment:

[0056] First, judge the vehicle speed. If the vehicle speed is zero, i.e., v = 0, and the value of v is obtained from the result of integrated navigation, then select the current road where the vehicle is located, and the algorithm ends. Otherwise, proceed to the next step. Combining the road structure information of the map data, judge whether there is an accessible road connected to the current road of the vehicle within a circular area centered on the vehicle position obtained by the extended Kalman filter and with a calculated horizontal protection level as the radius. If there is such a road, mark this type of road as an alternative road and proceed to the next step. Otherwise, re-initialize the algorithm.

[0057] 3) Map matching process:

[0058] a) Road selection weight calculation:

[0059] Let the l-th epoch be the starting epoch of the algorithm, and use the road selection weight calculation formula proposed by the present invention to calculate the road selection weights of all alternative road segments in the l-th epoch. The formula is as follows:

[0060] D ml = I m ·R m ·(A·d ml + Δθ ml )

[0061] where A, I, R are hyperparameters, d ml is the distance in meters between the position estimate of the vehicle obtained by integrated navigation and the projection of the position estimate of integrated navigation on the m-th alternative road and its extension, and Δθ ml is the difference in degrees between the heading angle estimate output by integrated navigation and the direction angle of the m-th alternative road.

[0062] The following gives the calculation methods of the hyperparameters A, I, R and the sliding window size used later:

[0063] The hyperparameter A is the position and attitude weight parameter. When the Dilution of Precision (DOP, the Dilution of Precision is a quality indicator of the GNSS position, which takes into account the position of each satellite relative to other satellites in the constellation and their geometric position relative to the GNSS receiver. Here, PDOP is used, the full English name is Position Dilution of Precision, translated as "Spatial Position Dilution of Precision". The Spatial Position Dilution of Precision is a dimensionless number representing the relationship between the user's position error and the satellite position error.) value is less than 3, it can be considered that the GNSS signal quality is good. At this time, more reliance can be placed on the position information provided by the integrated navigation. Therefore, the value of A can be set to 2. When the DOP value is greater than 3, it can be considered that the GNSS signal quality is poor. At this time, when calculating the road selection weight, more consideration can be given to the attitude information obtained by the IMU. Therefore, the value of A can be reduced and the value of A can be set to 2e 3-Y , where Y is the DOP value.

[0064] I m is the path planning weight parameter. When the user uses the path planning function of the navigation program, it can be considered that the user has a stronger willingness to drive towards the road indicated by the path planning. Based on this, the road selection weight of this road can be increased. Therefore, the path selection parameter is set as follows:

[0065]

[0066] R m is the traffic rule weight parameter. Most drivers can be considered as traffic participants who abide by traffic rules. Therefore, even if the road is connected, subjectively, the driver will not drive into a road that violates traffic rules. Therefore, the traffic rule weight parameter is introduced to reduce the weight of the road that can only be entered by violating traffic rules. The traffic rule weight parameter is as follows:

[0067]

[0068] The size of the sliding window k determines the accuracy rate of road selection. The driving route of the driver at the intersection is affected by the size and structure of the intersection. For example, when a vehicle turns left and goes straight at an intersection, the trajectories of the two are not very different when they first enter the intersection. At this time, if a hasty judgment is made, it may lead to an error in the road selection after the intersection. Therefore, it is necessary to adjust the size of the sliding window to eliminate such errors. The calculation of the K value is as follows:

[0069]

[0070] where fs is the sampling rate of GNSS, and n is the number of lanes of the road with the most lanes connected to this intersection.

[0071] b) Normalization of road selection weights:

[0072] The normalization formula for the road selection weight value is as follows:

[0073]

[0074] M is the set of alternative roads.

[0075] c) Road selection process:

[0076] Apply a sliding window to the minimum value of the normalized road selection weight values for each of the consecutive k epochs, where k is the size of the sliding window. If the road corresponding to the minimum weight term within the window has not changed, then select this road and the algorithm ends. If the roads corresponding to the minimum weight terms within the window are inconsistent, then make the following judgment: Determine whether the vehicle has entered a new road through the integrated navigation and positioning solution. If it has entered a new road, then select the new road and the algorithm ends. Otherwise, continue with the road selection process and update P il Re - perform road selection until the algorithm ends.

[0077] In specific implementation, the present application provides a computer storage medium and a corresponding data processing unit. Among them, the computer storage medium can store a computer program, and when the computer program is executed by the data processing unit, it can run the inventive content of a road selection judgment method based on prior information provided by the present invention and some or all of the steps in each embodiment. The storage medium can be a magnetic disk, an optical disc, a read - only memory (ROM), or a random access memory (RAM), etc.

[0078] Those skilled in the art can clearly understand that the technical solutions in the embodiments of the present invention can be implemented by means of a computer program and its corresponding general - purpose hardware platform. Based on such an understanding, the technical solutions in the embodiments of the present invention, in essence, or the part that contributes to the prior art can be embodied in the form of a computer program, that is, a software product. This computer program software product can be stored in the storage medium and includes several instructions to enable a device (which can be a personal computer, a server, a single - chip microcomputer, a MUU, or a network device, etc.) including a data processing unit to execute the methods described in each embodiment or some parts of the embodiments of the present invention.

[0079] The present invention provides the idea and method of a road selection judgment method based on prior information. There are many methods and ways to specifically implement this technical solution. The above - mentioned is only the preferred implementation manner of the present invention. It should be noted that for those of ordinary skill in the art in this technical field, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention. Each component not clearly defined in this embodiment can be implemented by using the prior art.

Claims

1. A road selection judgment method based on prior information, characterized in that Including the following steps: Step 1, initialization: Obtain the vehicle position, the heading angle of the vehicle according to integrated navigation, and calculate the horizontal protection level; Step 2, according to the vehicle state and map data, decide whether to perform a road selection judgment based on prior information; Step 3, perform a road selection judgment according to prior information; Among them, the road selection judgment described in Step 3 includes the following steps: Step 3-1, calculate the road selection weight; Step 3-2, normalize the road selection weight; Step 3-3, perform road selection; The calculation of the road selection weight described in Step 3-1 specifically includes: Set the first epoch as the starting epoch, and the method for calculating the road selection weights of all alternative road segments in the first epoch is as follows: D ml = I m · R m · (A · d ml + Δθ ml ) Among them, D ml represents the road selection weight value of the m-th alternative road, A, I m and R m are hyperparameters. Extend the alternative road along the tangent direction of the alternative road at the intersection. d ml is the distance between the vehicle position and the projection on the m-th alternative road and its extension line, in meters. Δθ ml is the difference between the vehicle's heading angle and the direction angle of the m-th alternative road, in degrees; The hyperparameters A, I m and R m mentioned in Step 3-1 are calculated as follows: The hyperparameter A is the position and attitude weight parameter. When the spatial position accuracy factor PDOP value is less than 3, the value of the hyperparameter A is set to 2; when the spatial position accuracy factor PDOP value is greater than 3, the value of the hyperparameter A is decreased and the value of the hyperparameter A is set to 2e 3-Y , where Y is the spatial position accuracy factor PDOP value; Hyperparameter I m is the path planning weight parameter and is set as follows: Hyperparameter R m is the traffic rule weight parameter and is set as follows: Among them, path planning and traffic rules are the so-called prior information.

2. The method for judging road selection based on prior information according to claim 1, wherein, The specific method for obtaining the vehicle position described in Step 1 includes: The integrated navigation includes: an inertial measurement unit and a global navigation satellite system GNSS; perform mechanical alignment using the measurement results of the inertial measurement unit; calculate the least squares solution of the GNSS pseudorange; combine the above results and perform extended Kalman filtering to obtain the vehicle position.

3. The method for judging road selection based on prior information according to claim 2, wherein The horizontal protection level described in Step 1 is calculated according to the following method: HPL = max(slope i )·P bias Among them, HPL represents the horizontal protection level, and P bias represents the minimum measurement deviation value detected under the set false alarm rate and missed detection rate. slope i is the characteristic slope of the i-th satellite, and the calculation method is as follows: where, Λ 1j represents the element in the first row and the j-th column of matrix Λ, Λ 2j represents the element in the first row and the j-th column of matrix Λ, S jj represents the element in the j-th row and the j-th column of matrix S, n represents the number of satellites, where, matrix Λ = (A T A) -1 A T , matrix S = [I - A(A T A) -1 A T , and A is the measurement matrix.

4. The method for judging road selection based on prior information according to claim 3, characterized in that The specific method for deciding whether to perform a road selection judgment based on prior information described in Step 2 is as follows: Step 2-1, judge the speed of the vehicle. If the vehicle speed is zero, select the current road unchanged and the judgment ends. Otherwise, execute Step 2-2; among them, the vehicle speed is calculated while obtaining the vehicle position in Step 1; Step 2-2, combined with the road structure information in the map data, judge whether there is an allowed road connected to the road where the vehicle is located within a circular area centered on the vehicle position obtained in Step 1 with the horizontal protection level as the radius. If so, mark the road as an alternative road and execute Step 3. Otherwise, re-execute Step 1.

5. A method for judging road selection based on prior information according to claim 4, characterized in that, The normalization of the road selection weight described in Step 3-2 specifically includes: The normalization method of the road selection weight is: Among them, P ml represents the normalized road selection weight of the m-th alternative road, and M is the set of alternative roads.

6. The method for judging road selection based on prior information according to claim 5, wherein The road selection described in Step 3-3 specifically includes: Apply a sliding window to determine the minimum value of the normalized road selection weights in consecutive k epochs. If the road corresponding to the minimum weight term in the sliding window remains unchanged, select this road, complete the road selection, and end. If the roads corresponding to the minimum weight terms in the sliding window are inconsistent, make the following judgment: Determine whether the vehicle has entered a new road based on the vehicle's position. If it has entered a new road, select the new road, complete the road selection, and end. Otherwise, re - perform the road selection and update P ml Re - perform the road selection until the road selection is completed and ended.

7. The method for judging road selection based on prior information according to claim 6, wherein The calculation method of the size k of the sliding window described in Step 3-3 is as follows: Among them, fs is the sampling rate of the global navigation satellite system GNSS, and q is the number of lanes of the road with the most connected lanes at the current intersection.

Images