Yaw determination method and device, computer equipment and storage medium

Abstract

The invention discloses a yaw determination method and device, computer equipment and a storage medium, and the method comprises the steps: obtaining the original track point information, actual yaw time, re-planned path information and re-planned path time of a target vehicle when the target vehicle is yawed; determining an actual driving path of the target vehicle based on the original track point information; determining a yaw track based on the actual yaw time, the time for re-planning the path and the actual driving path; determining whether the target vehicle is in an error-prone scene or not based on the yaw trajectory and a planned path before the target vehicle yaws; and if the target vehicle is in the error-prone scene, determining whether yaw misjudgment occurs or not. The method not only can more accurately restore the driving path of a driver, but also can be compatible with complex scenes such as original track point flat drift, road network equidistant deviation and parallel roads, so as to determine whether the yaw judgment of the navigation engine is wrong or not, thereby improving the accuracy and integrity of the yaw judgment of the navigation engine.

Description

Technical Field

[0001] The present invention relates to the field of navigation technology, and in particular to a method, device and storage medium for determining a yaw motion, a computer device and a storage medium. Background Art

[0002] With the development of mobile Internet technology and navigation technology, people are increasingly relying on map navigation applications for daily travel, and people's requirements for navigation accuracy and sensitivity are also getting higher and higher. Among them, in navigation engine applications, the experience of yaw judgment is crucial. Users expect yaw judgment to be accurate and sensitive. If the yaw judgment is wrong, it will lead to the wrong provision of new paths and broadcasts, causing unnecessary detours and safety hazards. If the yaw judgment is too slow, it will cause the induced information to be delayed or lost, and there will be errors with the real road information, which will cause a bad experience. Summary of the invention

[0003] Based on this, it is necessary to provide a yaw determination method, device, computer equipment and storage medium to address the above technical issues, so as to solve at least one problem existing in the above prior art.

[0004] In a first aspect, an embodiment of the present application provides a yaw determination method, comprising:

[0005] When the target vehicle deviates, the original track point information, actual deviation time, re-planned path information and re-planned path time of the target vehicle are obtained;

[0006] Based on the original trajectory point information, determining the actual driving path of the target vehicle;

[0007] Determining a yaw trajectory based on the actual yaw time, the re-route time, and the actual travel path;

[0008] Determining whether the target vehicle is in an error-prone scenario based on the yaw trajectory and the planned path of the target vehicle before the yaw;

[0009] If the target vehicle is in a non-error-prone scenario, it is determined whether a yaw misjudgment occurs.

[0010] In one embodiment, determining whether the target vehicle is in an error-prone scenario based on the yaw trajectory and the planned path of the target vehicle before the yaw includes:

[0011] Determining a first trajectory curve and a yaw path corresponding to the yaw trajectory;

[0012] Based on the yaw path, determining a second trajectory curve;

[0013] Based on the sizes of the first trajectory curve and the second trajectory curve, intercepting a corresponding third trajectory curve in the planned path before the target vehicle deviates;

[0014] Comparing the first trajectory curve, the second trajectory curve and the third trajectory curve in pairs;

[0015] Based on the comparison result, it is determined whether the target vehicle is in an error-prone scenario.

[0016] In one embodiment, determining whether the target vehicle is in an error-prone scenario based on the comparison result includes:

[0017] If the second trajectory curve does not overlap with the third trajectory curve, the similarity between any two of the first trajectory curve, the second trajectory curve and the third trajectory curve is greater than a preset similarity threshold, and the similarity between the first trajectory curve and the second trajectory curve is greater than the similarity between the first trajectory curve and the third trajectory curve, it is determined that the target vehicle is in an error-prone scenario.

[0018] In one embodiment, if the target vehicle is in an error-prone scenario, determining whether a deviation misjudgment occurs includes:

[0019] If the target vehicle is in an error-prone scenario, determining the comprehensive probability between different roads in the actual driving path;

[0020] Determining the best matching path other than the actual driving path from the candidate paths based on the product of the comprehensive probability and the preset penalty weight;

[0021] Comparing the optimal matching path with the planned path of the target vehicle before the deviation;

[0022] Based on the comparison results, it is determined whether a yaw misjudgment occurs.

[0023] In one embodiment, after determining whether the target vehicle is in an error-prone scenario, the method further includes:

[0024] If the target vehicle is in a non-error-prone scenario, it is determined whether a yaw misjudgment occurs.

[0025] In one embodiment, determining whether a deviation misjudgment occurs includes:

[0026] determining a yaw path corresponding to the yaw trajectory;

[0027] comparing the yaw path with a planned path of the target vehicle before the yaw;

[0028] Based on the comparison results, determine whether a yaw misjudgment occurs.

[0029] In one embodiment, determining the yaw trajectory based on the actual yaw time, the time of re-planning the path, and the actual driving path includes:

[0030] Determining a target yaw trajectory point based on the actual yaw time;

[0031] Determining a starting trajectory point and an ending trajectory point in the actual driving path based on the target off-course trajectory point;

[0032] The yaw trajectory is obtained based on the start trajectory point and the end trajectory point.

[0033] In one embodiment, after determining whether a deviation misjudgment occurs, the method further includes:

[0034] When no yaw misjudgment occurs, determining a yaw misjudgment rate of the navigation engine;

[0035] Determine the yaw response time for this navigation determination result;

[0036] The yaw determination effect of the navigation engine is evaluated based on the yaw error rate and the yaw response time.

[0037] In a second aspect, a yaw determination device is provided, comprising:

[0038] A deviation information acquisition unit, used to acquire original track point information, actual deviation time, re-planned path information and re-planned path time of the target vehicle when the target vehicle deviates;

[0039] An actual driving path determining unit, used to determine the actual driving path of the target vehicle based on the original trajectory point information;

[0040] a yaw trajectory determination unit, configured to determine a yaw trajectory based on the actual yaw time, the time for re-planning the path, and the actual driving path;

[0041] An error-prone scene judgment unit, used to determine whether the target vehicle is in an error-prone scene based on the yaw trajectory and the planned path of the target vehicle before the yaw;

[0042] The yaw determination unit is used to determine whether a yaw misjudgment occurs if the target vehicle is in an error-prone scenario.

[0043] In a third aspect, a readable storage medium is provided, on which computer-readable instructions are stored, wherein the computer-readable instructions, when executed by a processor, implement the yaw determination method as described above.

[0044] The above-mentioned yaw determination method, device, computer equipment and storage medium, the method of which is implemented, includes: when the target vehicle yaws, obtaining the original trajectory point information, actual yaw time, re-planned path information and re-planned path time of the target vehicle; based on the original trajectory point information, determining the actual driving path of the target vehicle; based on the actual yaw time, the re-planned path time and the actual driving path, determining the yaw trajectory; based on the yaw trajectory and the planned path before the target vehicle yaws, determining whether the target vehicle is in an error-prone scenario; if the target vehicle is in an error-prone scenario, determining whether a yaw misjudgment occurs. In an embodiment of the present application, when the navigation engine detects that the vehicle has deviated, the vehicle's trajectory points, deviation time, re-planned route and re-planned route time can be obtained to determine the vehicle's actual driving path and deviated trajectory, and based on the vehicle's current scenario, determine whether a deviated misjudgment problem has occurred. This can not only more accurately restore the driver's driving path, but also be compatible with complex scenarios such as original trajectory point drift, road network equidistant deviation, parallel roads, etc., thereby determining whether the navigation engine's current deviated judgment is incorrect, thereby improving the accuracy and completeness of the navigation engine's deviated judgment. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0046] Figure 1 is a schematic diagram of a flow chart of a yaw determination method in one embodiment of the present invention;

[0047] Figure 2 is a schematic diagram of an application environment of a yaw determination method in an embodiment of the present invention;

[0048] Figure 3 It is a schematic diagram of an application environment of a road network matching method in an error-prone scenario in one embodiment of the present invention;

[0049] Figure 4 is a structural schematic diagram of a yaw determination device in one embodiment of the present invention;

[0050] Figure 5 is a schematic diagram of a computer device in one embodiment of the present invention. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] In one embodiment, if Figure 1 As shown, a method for determining yaw is provided, comprising the following steps:

[0053] In step S110, when the target vehicle deviates, the original track point information, the actual deviation time, the re-planned path information and the re-planned path time of the target vehicle are obtained;

[0054] Alternatively, see Figure 2 , a navigation application can be pre-installed in a front-end device, such as a mobile phone or a car computer. When the vehicle is driving or when navigation begins, the navigation application will collect the location information of the target vehicle in real time, report it to the cloud, and store it in the database through the cloud. When the navigation engine detects that the path of the target vehicle does not match the original planned path and there is deviation, a deviation detection request can be initiated. When the cloud obtains the deviation detection request, it will record the actual deviation timestamp and send the re-planned path information, such as path ID, path shape point string, etc. to the front-end device. At the same time, the cloud will store the path information (path id, path information (including LinkID and driving direction LinkDir), path shape point string) offline to the specified database. When the front-end navigation application receives the re-planned path, it will also record the time of the re-planned path, that is, the display timestamp of the re-planned path displayed by the front-end, and record and report it in real time. In this way, the original track point information, actual deviation time, re-planned path information and re-planned path time of the target vehicle can be obtained.

[0055] The location information may include the timestamp, speed, direction, longitude and latitude, positioning accuracy, etc. of the original track point. It is understandable that the track point information may be collected through a positioning system, such as a Global Positioning System (GPS) / Global Navigation Satellite System (GNSS).

[0056] In step S120, based on the original trajectory point information, the actual driving path of the target vehicle is determined;

[0057] Optionally, the pre-stored complete original trajectory points can be sorted according to preset rules, such as time sequence or position sequence, etc. Then the complete trajectory point information of the target vehicle can be extracted from the database, and a preset path planning algorithm, such as a hidden Markov model (HMM), a geometric matching algorithm, a naive Bayesian map matching algorithm, etc., is used to match the trajectory points to the corresponding road network to obtain a matching path, and the matching path is used as the actual driving path, that is, the optimal path M. total The above-mentioned actual driving path matching process can be understood as a path matching method in an offline matching scenario.

[0058] Among them, M total is the actual driving path of the target vehicle, which can be composed of multiple road network links, which are sorted in chronological order or in the order of driving positions. Then M total :Link1→Link2→...→Link n .

[0059] Taking the HMM algorithm as an example, it has the characteristics of high noise and high signal sparsity tolerance. In road network matching, the original trajectory points can be regarded as observation sequences, and the road states in the road network can be regarded as hidden states. By defining the state transition probability (the probability of a vehicle transferring from one road to another), the observation probability (the probability of generating a certain observation trajectory point on a certain road) and the initial state probability, the most likely hidden state sequence is found using methods such as the Viterbi algorithm, so that the trajectory points can be accurately mapped to the road network. The vehicle trajectory can be accurately mapped to the road network data, and the actual driving path of the target vehicle can be restored.

[0060] Optionally, a backtracking algorithm based on connectivity constraints can be used to select the best possible connected path to obtain the actual driving path. Each link (road segment) has a starting point, an end point, and direction information. For any trajectory point, in the forward process, the connectivity relationship between the trajectory point and the previous trajectory point can be checked according to the path planning algorithm and the road network topology. If it is connected, that is, there is a legal path from the trajectory point to the previous trajectory point, it will be counted in the connected list N. The objective function in the backtracking can be as follows:

[0061] Given the model θ and the observation sequence (trajectory point string) Z = (Z1, Z2, ..., Z n ), and the connected list set is N, find the optimal hidden state sequence X in list N = (X1, X2, ..., X n ), i.e., the most likely candidate road segment.

[0062]

[0063] That is, when no candidate matching point can be found in the connected list N, the candidate matching point is selected according to the principle of maximum probability. Since the forward process of the existing HMM model road network matching algorithm sets its disconnected transition probability to a minimum probability value, if the trajectory drifts continuously, the comprehensive probability of disconnection will still be maximized, resulting in matching errors. Therefore, the backtracking algorithm based on connectivity constraints avoids trajectory drift and matches to disconnected paths, thereby reducing matching errors.

[0064] It should be noted that the specific implementation process of the optimal actual driving path obtained by the backtracking algorithm based on connectivity constraints is as follows: based on the original trajectory point information, the trajectory point feature information, such as speed, direction and other related feature information, is extracted. Based on these feature information, the road segment screening work is started, which includes setting the candidate range, that is, defining the possible road segment range in the road network according to the trajectory point features; then selecting candidate road segments from these set ranges, and determining the number of candidate road segments, etc. Then the observation probability can be calculated, and the possibility of the trajectory point appearing on each candidate road segment can be evaluated based on the previously extracted trajectory point features and the selected candidate road segments. Then the state transition probability is calculated, which is based on the road network data and the sequence of trajectory points to determine the probability of transferring from one road segment to another. After obtaining the observation probability and state transition probability, the comprehensive probability is further calculated, and the above two probabilities are comprehensively considered to evaluate the possibility of the entire trajectory matching. Subsequently, the Viterbi algorithm can be used to solve the comprehensive probability, and the most likely matching path can be found through this algorithm. The road segment matching is performed according to the solution result of the Viterbi algorithm to determine the final matching road segment. Finally, a backtracking operation is performed to check and optimize the matching results, and finally an accurate matching result is output, that is, the actual driving trajectory of the target vehicle in the road network. The above-mentioned backtracking algorithm based on connectivity constraints can more accurately restore the driver's driving trajectory, with a matching accuracy rate of more than 98%.

[0065] In step S130, a yaw trajectory is determined based on the actual yaw time, the time for re-planning the path, and the actual driving path;

[0066] Optionally, the target vehicle's driving trajectory may be segmented according to the time of the re-planned route, that is, the timestamp T of the re-planned route after the deviation may be obtained. B Find the timestamp T corresponding to the yaw request initiated by the front end A , based on the timestamp T A Find the corresponding target vehicle's trajectory point P, intercept 100m forward from P as the starting trajectory point S1, and intercept the ending trajectory point less than or equal to the timestamp T B The last trajectory point S2 is intercepted and used as the standard trajectory [S1->S2] during this yaw period. The standard trajectory [S1->S2] can be used as the yaw trajectory during this yaw period.

[0067] It can be understood that the position of the trajectory point P corresponds to the timestamp T B The time node represented can reflect the specific geographical location of the vehicle when it deviates.

[0068] In step S140, based on the yaw trajectory and the planned path of the target vehicle before the yaw, it is determined whether the target vehicle is in an error-prone scenario;

[0069] It should be noted that error-prone scenarios refer to scenarios where the original trajectory point happens to drift to other paths that can be connected, there is an equidistant deviation between the road network data and the trajectory point of the target vehicle, or parallel roads (including main and auxiliary roads / viaducts, etc.), which makes offline matching prone to matching errors. Figure 3 As shown in the figure, there is a partial unmatched area between the original trajectory point A of the target vehicle and the road network mapping point B matched by the offline matching algorithm, that is, Figure 3 In the middle of the matched route, there is a mapping point B on the auxiliary road, but the target vehicle is actually driving on the main road. The track point drift is closer to the auxiliary road, resulting in a matching error, and the driving path obtained after matching is wrong. However, in this scenario, the original route planning is for the target vehicle to drive along the main road, and the target vehicle is actually driving on the main road, but the navigation engine determines that the target vehicle has deviated based on the offline matching results. This is an incorrect judgment behavior and is a false deviation. However, if Figure 2 The matching result shown will still believe that the track here is traveling on the auxiliary road and has deviated behavior, so the navigation engine's yaw judgment will be considered correct. Since one false yaw is not recognized once, it is not conducive to the subsequent evaluation of the navigation engine's yaw judgment effect, which can easily lead to an increase in the navigation engine's misjudgment rate and a decrease in the accuracy of yaw judgment.

[0070] Based on the above reasons, it is possible to determine whether the target vehicle is in an error-prone scenario based on the obtained yaw trajectory and the planned path before the target vehicle yaws. Optionally, a first trajectory curve (curve_1) can be obtained based on the yaw trajectory [S1->S2], a yaw path can be obtained according to the yaw trajectory, and a corresponding second trajectory curve (curve_2) can be obtained based on the matching longitude and latitude points and link point strings of the yaw path. At the same time, the latitude and longitude of the current route point string in the planned path before the target vehicle yaws are intercepted by using the map size of the first trajectory curve and the second trajectory curve to generate a third trajectory curve (curve_3).

[0071] Then, the similarity between curve_1, curve_2, and curve_3 is calculated through a similarity algorithm. If curve_3 and curve_2 do not overlap, but are similar and satisfy Similar(curve_1, curve_3)>Similar(curve_1, curve_2), it is determined to be an error-prone scenario. The main reasons for error-prone scenarios are that the original trajectory point of the target vehicle drifts to other paths that can be connected, there is an equidistant deviation between the road network data and the driver's trajectory point, or the scene is in a parallel road (such as main and auxiliary roads, viaducts, etc.), which can easily cause matching errors.

[0072] It can be understood that based on the timestamps corresponding to the starting trajectory point and the ending trajectory point of the yaw trajectory, a corresponding path can be intercepted in the actual driving path, and this path is the yaw path. The yaw path refers to the path traveled by the target vehicle when the yaw occurs.

[0073] In step S150, if the target vehicle is in an error-prone scenario, it is determined whether a deviation misjudgment occurs.

[0074] It is understandable that in different scenarios, different methods can be used to determine whether the navigation engine has erroneously deviated. For example, if the target vehicle is in a non-error-prone scenario, the path M matching the yaw trajectory and the planned route R of the target vehicle before the yaw can be obtained according to the yaw trajectory [S1->S2] obtained above for judgment. That is, the paths M and R can be compared. If the comparison results are different, it can be considered that the target vehicle actually deviated, that is, the navigation engine's judgment is correct; if the comparison results are the same, it can be considered that the target vehicle did not deviate, so the navigation engine has a yaw judgment error.

[0075] The above-mentioned yaw determination method includes: when the target vehicle yaws, obtaining the original track point information, actual yaw time, re-planned path information and re-planned path time of the target vehicle; determining the actual driving path of the target vehicle based on the original track point information; determining the yaw trajectory based on the actual yaw time, the re-planned path time and the actual driving path; determining whether the target vehicle is in an error-prone scene based on the yaw trajectory and the planned path before the target vehicle yaws; if the target vehicle is in a non-error-prone scene, determining whether a yaw misjudgment occurs. In the embodiment of the present application, when the navigation engine detects that the vehicle yaws, the vehicle's track points, yaw time, re-planned path and re-planned path time can be obtained to determine the vehicle's actual driving path and yaw trajectory, and based on the vehicle's current scene, determine whether a yaw misjudgment problem occurs, which can not only restore the driver's driving path more accurately, but also be compatible with complex scenes such as original track point flat drift, road network equidistant deviation, parallel roads, etc., so as to determine whether the navigation engine's current yaw judgment is wrong, so as to improve the accuracy and completeness of the navigation engine's yaw judgment.

[0076] In an embodiment of the present application, determining the yaw trajectory based on the actual yaw time, the time of re-planning the path, and the actual driving path includes:

[0077] Determining a target yaw trajectory point based on the actual yaw time;

[0078] Determining a starting trajectory point and an ending trajectory point in the actual driving path based on the target off-course trajectory point;

[0079] The yaw trajectory is obtained based on the start trajectory point and the end trajectory point.

[0080] Optionally, the optimal path M can be obtained through the first round of offline matching based on the original trajectory points obtained. total , where M total The driver's driving path M total :Link1:→Link2→...→Link n , which is composed of multiple road networks Link. Then, the target vehicle's driving trajectory can be segmented according to the time of the re-planned route, that is, the timestamp T of the re-planned route after the deviation can be obtained. B Find the timestamp T corresponding to the yaw request initiated by the front end A , based on the timestamp T A Find the corresponding target vehicle's trajectory point P, intercept 100m forward from P as the starting trajectory point S1, and intercept the ending trajectory point less than or equal to the timestamp T BThe last trajectory point S2 is intercepted and used as the standard trajectory [S1->S2] during this yaw period. The standard trajectory [S1->S2] can be used as the yaw trajectory during this yaw period.

[0081] The position of the trajectory point P corresponds to the timestamp T B The time node represented can reflect the specific geographical location of the vehicle when it deviates.

[0082] In one embodiment of the present application, determining whether the target vehicle is in an error-prone scenario based on the yaw trajectory and the planned path before the target vehicle yaws includes:

[0083] Determining a first trajectory curve and a yaw path corresponding to the yaw trajectory;

[0084] Based on the yaw path, determining a second trajectory curve;

[0085] Based on the sizes of the first trajectory curve and the second trajectory curve, intercepting a corresponding third trajectory curve in the planned path before the target vehicle deviates;

[0086] Comparing the first trajectory curve, the second trajectory curve and the third trajectory curve in pairs;

[0087] Based on the comparison result, it is determined whether the target vehicle is in an error-prone scenario.

[0088] It should be noted that error-prone scenarios refer to scenarios where the original trajectory points happen to drift to other paths that can be connected, there are equidistant deviations between the road network data and the trajectory points of the target vehicle, or there are parallel roads (including main and auxiliary roads / viaducts, etc.), which makes offline matching prone to matching errors.

[0089] Optionally, if the target vehicle is in an error-prone scenario, when the yaw trajectory of the target vehicle is obtained, the yaw trajectory may include multiple trajectory points, each trajectory point may correspond to a specific longitude and latitude value, and the longitude and latitude values ​​are arranged in the order of time or position of the target vehicle passing, so that a series of longitude and latitude points can be obtained. The longitude and latitude points are connected in sequence to obtain a yaw curve that can represent the actual driving trajectory of the target vehicle during the yaw period, that is, the first trajectory curve curve_1.

[0090] Based on the timestamps corresponding to the starting and ending points of the yaw trajectory, the corresponding path is intercepted in the actual driving path. This path is the yaw path, which refers to the path that the target vehicle travels when it deviates. Based on the yaw path, the node set contained in each section of the yaw path can be obtained, that is, the link point string. Each link corresponds to a longitude and latitude number, and a matching series of longitude and latitude points can be obtained. The longitude and latitude points are connected in sequence to obtain the second trajectory curve curve_2.

[0091] Using the map sizes of the first trajectory curve curve_1 and the second trajectory curve curve_2, that is, the area range of the first trajectory curve and the second trajectory curve, the corresponding road section within the area range is intercepted in the planned path before the target vehicle deviates, and the latitude and longitude values ​​of all trajectory points corresponding to the road section are obtained, and the latitude and longitude points are connected in sequence, so as to obtain the third trajectory curve curve_3.

[0092] Then, the first trajectory curve curve_1, the second trajectory curve curve_2, and the third trajectory curve curve_3 can be calculated for similarity in pairs, for example, similarity can be calculated based on curvature similarity algorithm, dynamic time warping (DTW) algorithm, etc., and the similarity value Similar between each two groups of trajectory curves can be obtained. Then these similarity values ​​Similar are compared with the pre-set similarity threshold. If they are all greater than the set similarity threshold, they are determined to be similar. Based on the comparison result, it can be determined whether the target vehicle is in an error-prone scenario.

[0093] In an embodiment of the present application, determining whether the target vehicle is in an error-prone scenario based on the comparison result includes:

[0094] If the second trajectory curve does not overlap with the third trajectory curve, the similarity between any two of the first trajectory curve, the second trajectory curve and the third trajectory curve is greater than a preset similarity threshold, and the similarity between the first trajectory curve and the second trajectory curve is greater than the similarity between the first trajectory curve and the third trajectory curve, it is determined that the target vehicle is in an error-prone scenario.

[0095] Optionally, if the third trajectory curve curve_3 does not overlap with the second trajectory curve curve_2, the similarity value can be calculated to determine that the two trajectory curves are similar.

[0096] Similar(curve_1,curve_3)>Similar(curve_1,curve_2), then it is an error-prone scenario. It is possible that the original trajectory point of the target vehicle just drifts to other paths that can be connected, there is an equidistant deviation between the road network data and the trajectory point of the target vehicle, or parallel roads (including main and auxiliary roads / viaducts, etc.) and other scenarios lead to matching errors.

[0097] In one embodiment of the present application, if the target vehicle is in an error-prone scene, determining whether a deviation misjudgment occurs includes:

[0098] If the target vehicle is in an error-prone scenario, determining the comprehensive probability between different roads in the actual driving path;

[0099] Determining the best matching path other than the actual driving path from the candidate paths based on the product of the comprehensive probability and the preset penalty weight;

[0100] Comparing the optimal matching path with the planned path of the target vehicle before the deviation;

[0101] Based on the comparison results, it is determined whether a yaw misjudgment occurs.

[0102] It should be noted that when the target vehicle is in an error-prone scenario, the actual driving path of the target vehicle obtained by the offline matching method may be inaccurate. Therefore, it is necessary to optimize the forward process and then perform a second round of matching to obtain the second matching result. For example, if the result of the first round of matching is M, that is, the actual driving path obtained by offline matching is M, in the forward process, the comprehensive probability between Link and Link in the matching result M is multiplied by a penalty weight, such as 0.3, and the matching result M is not selected first. In the end, an optimal matching path solution M other than the matching result M will be obtained. opt As a selection of the second matching result, it is used for mis-deviation judgment in error-prone scenarios.

[0103] It should be noted that the matching result M may be composed of multiple links. When matching routes offline through an offline algorithm, for example, taking the HMM algorithm as an example, each link can be regarded as a state. For the starting point of the matching result M, its initial probability can be determined first, and the initial probability can be determined based on prior knowledge or historical data. For each state, the observation probability can be calculated based on the observation data related to it. The observation data can be position, speed, etc. Taking the position as an example, the observation probability can be determined based on the matching degree between the current link and the observed position. Then, the probability of a link transferring to the next link, that is, the transition probability, can be calculated based on the topological structure of the road network, traffic rules, historical driving data, etc. Then, the comprehensive probability can be calculated based on the initial probability, transition probability and observation probability.

[0104] The optimal matching path solution M obtained in the second round of matching opt Afterwards, the M opt It can be composed of multiple links. The planned path before the target vehicle deviates can also be composed of multiple links. By comparing M opt The similarity between the link of the planned path before the target vehicle deviates is calculated. If the similarity is greater than the preset similarity threshold, the comparison is considered consistent. At this time, it can be considered that a misjudgment of yaw has occurred. Otherwise, it is considered that the target vehicle has yawed, that is, the yaw judgment result of the navigation engine is correct.

[0105] In an embodiment of the present application, after determining whether the target vehicle is in an error-prone scenario, the method further includes:

[0106] If the target vehicle is in a non-error-prone scenario, it is determined whether a yaw misjudgment occurs.

[0107] It is understandable that in different scenarios, different methods can be used to determine whether the navigation engine has erroneously deviated. For example, if the target vehicle is in a non-error-prone scenario, the path M matching the yaw trajectory and the planned route R of the target vehicle before the yaw can be obtained according to the yaw trajectory [S1->S2] obtained above for judgment. That is, the paths M and R can be compared. If the comparison results are different, it can be considered that the target vehicle actually deviated, that is, the navigation engine's judgment is correct; if the comparison results are the same, it can be considered that the target vehicle did not deviate, so the navigation engine has a yaw judgment error.

[0108] In an embodiment of the present application, determining whether a deviation misjudgment occurs includes:

[0109] determining a yaw path corresponding to the yaw trajectory;

[0110] comparing the yaw path with a planned path of the target vehicle before the yaw;

[0111] Based on the comparison results, determine whether a yaw misjudgment occurs.

[0112] Optionally, based on the timestamps corresponding to the starting point S1 and the ending point S2 of the yaw trajectory, a corresponding path can be intercepted in the actual driving path, and the path is the yaw path M. The yaw path refers to the path that the target vehicle travels when the yaw occurs. The yaw path M can be composed of multiple links, which can be combined into a link string in time order or position order, for example, M: Link m-2 →Link m-1 →Link mThe planned path R before the target vehicle deviates can also be composed of multiple links, for example, R: Link1→Link2→...→Link k , the similarity comparison is performed using the Link string of the yaw path M and the planned path R before the target vehicle yaws. If the similarity is greater than the preset similarity threshold, it means that the target vehicle has not yawed, and the yaw judgment of the navigation engine is incorrect. Otherwise, it means that the target vehicle has yawed, and the yaw judgment of this navigation engine is accurate.

[0113] In an embodiment of the present application, after determining whether a yaw misjudgment occurs, the method further includes:

[0114] When no yaw misjudgment occurs, determining a yaw misjudgment rate of the navigation engine;

[0115] Determine the yaw response time for this navigation determination result;

[0116] The yaw determination effect of the navigation engine is evaluated based on the yaw error rate and the yaw response time.

[0117] Among them, the rate of false deviation = the total number of false deviations / the total number of deviations;

[0118] Yaw response time = timestamp of the front-end replanning the route after obtaining the yaw - timestamp of the driver's actual yaw point;

[0119] Optionally, if no misjudgment of yaw occurs, that is, the navigation engine's yaw judgment is correct, the navigation engine's misjudgment rate and yaw response time can be calculated. In a non-error-prone scenario, if the actual driving path M is different from the Link string comparison of the planned route R before the target vehicle yaws, the navigation engine's yaw is correct, and the actual driving route M is still selected for yaw point determination.

[0120] According to the yaw trajectory information [S1->S2] during the yaw period, and the corresponding matching path result M (i.e., the actual driving path M is matched using the timestamps corresponding to the trajectory points S1 and S2) total Intercept), set M:Link m-2 →Link m-1 →Link m Compare the Link strings of M and R, and find the earliest Link in M ​​that is different from R yaw , and find the yaw trajectory segment [S1->S2] matching the Link yaw The earliest trajectory point S yaw , the timestamp T of the trajectory point yaw This is the actual deviation time of the driver. Then the deviation response time = the timestamp T of the route replanned after the deviation is obtained. B - The timestamp T of the driver's actual deviationyaw It should be noted that the yaw response time is an important indicator for measuring the yaw sensitivity of the navigation engine, and the yaw response time calculation is only performed for correct yaw judgment.

[0121] For the false deviation rate, the total number of false deviations of the navigation engine and the total number of deviations can be counted. If this is a false deviation, the total number of false deviations is increased by 1, otherwise it remains unchanged.

[0122] After obtaining the false yaw rate and the yaw response time, if the yaw response time is less than the preset time threshold, it is considered that the navigation engine has a high sensitivity, and if the false yaw rate is low, it is considered that the navigation engine has a high accuracy, so that the yaw judgment effect of the navigation engine can be evaluated, and the navigation engine can be adjusted based on the evaluation result to improve the yaw judgment accuracy of the navigation engine. Moreover, the above-mentioned automatic yaw response time calculation method can replace the existing manual road test marking calculation method, saving labor costs and improving efficiency.

[0123] In an embodiment of the present application, when the navigation engine detects that the vehicle has deviated, the vehicle's trajectory points, deviation time, re-planned route and re-planned route time can be obtained to determine the vehicle's actual driving path and deviated trajectory, and based on the vehicle's current scenario, determine whether a deviated misjudgment problem has occurred. This can not only more accurately restore the driver's driving path, but also be compatible with complex scenarios such as original trajectory point drift, road network equidistant deviation, parallel roads, etc., thereby determining whether the navigation engine's current deviated judgment is incorrect, thereby improving the accuracy and completeness of the navigation engine's deviated judgment.

[0124] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0125] In one embodiment, a yaw determination device is provided, and the yaw determination device corresponds one-to-one to the yaw determination method in the above embodiment. Figure 4 As shown, the yaw determination device includes a yaw information acquisition unit 10, an actual driving path determination unit 20, a yaw trajectory determination unit 30, an error-prone scene judgment unit 40, and a yaw determination unit 50. The functional modules are described in detail as follows:

[0126] The deviation information acquisition unit 10 is used to obtain the original track point information, actual deviation time, re-planned path information and re-planned path time of the target vehicle when the target vehicle deviates;

[0127] An actual driving path determining unit 20, configured to determine the actual driving path of the target vehicle based on the original trajectory point information;

[0128] A yaw trajectory determination unit 30, configured to determine a yaw trajectory based on the actual yaw time, the time for re-planning the path, and the actual driving path;

[0129] An error-prone scene judgment unit 40, used to determine whether the target vehicle is in an error-prone scene based on the yaw trajectory and the planned path of the target vehicle before the yaw;

[0130] The yaw determination unit 50 is used to determine whether a yaw misjudgment occurs if the target vehicle is in an error-prone scenario.

[0131] In one embodiment of the present application, the error-prone scene judgment unit 40 is further configured to:

[0132] Determining a first trajectory curve and a yaw path corresponding to the yaw trajectory;

[0133] Based on the yaw path, determining a second trajectory curve;

[0134] Based on the sizes of the first trajectory curve and the second trajectory curve, intercepting a corresponding third trajectory curve in the planned path before the target vehicle deviates;

[0135] Comparing the first trajectory curve, the second trajectory curve and the third trajectory curve in pairs;

[0136] Based on the comparison result, it is determined whether the target vehicle is in an error-prone scenario.

[0137] In one embodiment of the present application, the error-prone scene judgment unit 40 is further configured to:

[0138] If the second trajectory curve does not overlap with the third trajectory curve, the similarity between any two of the first trajectory curve, the second trajectory curve and the third trajectory curve is greater than a preset similarity threshold, and the similarity between the first trajectory curve and the second trajectory curve is greater than the similarity between the first trajectory curve and the third trajectory curve, it is determined that the target vehicle is in an error-prone scenario.

[0139] In one embodiment of the present application, the yaw determination unit 50 is further configured to:

[0140] If the target vehicle is in an error-prone scenario, determining the comprehensive probability between different roads in the actual driving path;

[0141] Determining the best matching path other than the actual driving path from the candidate paths based on the product of the comprehensive probability and the preset penalty weight;

[0142] Comparing the optimal matching path with the planned path of the target vehicle before the deviation;

[0143] Based on the comparison results, it is determined whether a yaw misjudgment occurs.

[0144] In one embodiment of the present application, the yaw determination unit 50 is further configured to:

[0145] If the target vehicle is in a non-error-prone scenario, it is determined whether a yaw misjudgment occurs.

[0146] In one embodiment of the present application, the yaw determination unit 50 is further configured to:

[0147] determining a yaw path corresponding to the yaw trajectory;

[0148] comparing the yaw path with a planned path of the target vehicle before the yaw;

[0149] Based on the comparison results, determine whether a yaw misjudgment occurs.

[0150] In one embodiment of the present application, the yaw trajectory determination unit 30 is further configured to:

[0151] Determining a target yaw trajectory point based on the actual yaw time;

[0152] Determining a starting trajectory point and an ending trajectory point in the actual driving path based on the target off-course trajectory point;

[0153] The yaw trajectory is obtained based on the start trajectory point and the end trajectory point.

[0154] In one embodiment of the present application, the apparatus further includes a navigation engine evaluation unit, configured to:

[0155] When no yaw misjudgment occurs, determining a yaw misjudgment rate of the navigation engine;

[0156] Determine the yaw response time for this navigation determination result;

[0157] The yaw determination effect of the navigation engine is evaluated based on the yaw error rate and the yaw response time.

[0158] In an embodiment of the present application, when the navigation engine detects that the vehicle has deviated, the vehicle's trajectory points, deviation time, re-planned route and re-planned route time can be obtained to determine the vehicle's actual driving path and deviated trajectory, and based on the vehicle's current scenario, determine whether a deviated misjudgment problem has occurred. This can not only more accurately restore the driver's driving path, but also be compatible with complex scenarios such as original trajectory point drift, road network equidistant deviation, parallel roads, etc., thereby determining whether the navigation engine's current deviated judgment is incorrect, thereby improving the accuracy and completeness of the navigation engine's deviated judgment.

[0159] For the specific definition of the yaw determination device, please refer to the definition of the yaw determination method above, which will not be repeated here. Each module in the above yaw determination device can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each of the above modules.

[0160] In one embodiment, a computer device is provided. The computer device may be a terminal device, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a readable storage medium. The readable storage medium stores computer-readable instructions. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer-readable instructions are executed by the processor, a yaw determination method is implemented. The readable storage medium provided in this embodiment includes a non-volatile readable storage medium and a volatile readable storage medium.

[0161] In an embodiment of the present application, a computer device is provided, including a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor. When the processor executes the computer-readable instructions, the steps of the above-mentioned yaw determination method are implemented.

[0162] In an embodiment of the application, a readable storage medium is provided, wherein the readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the above-mentioned yaw determination method are implemented.

[0163] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware through computer-readable instructions, and the computer-readable instructions can be stored in a non-volatile readable storage medium or a volatile readable storage medium. When the computer-readable instructions are executed, they may include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0164] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0165] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for determining yaw, characterized in that: The method comprises: When the target vehicle deviates, the original track point information, actual deviation time, re-planned path information and re-planned path time of the target vehicle are obtained; Based on the original trajectory point information, determining the actual driving path of the target vehicle; Determining a yaw trajectory based on the actual yaw time, the re-route time, and the actual travel path; Determining whether the target vehicle is in an error-prone scenario based on the yaw trajectory and the planned path of the target vehicle before the yaw; If the target vehicle is in an error-prone scene, determine whether a yaw misjudgment occurs.

2. The yaw determination method according to claim 1, characterized in that: The determining whether the target vehicle is in an error-prone scenario based on the yaw trajectory and the planned path of the target vehicle before the yaw includes: Determining a first trajectory curve and a yaw path corresponding to the yaw trajectory; Based on the yaw path, determining a second trajectory curve; Based on the sizes of the first trajectory curve and the second trajectory curve, intercepting a corresponding third trajectory curve in the planned path before the target vehicle deviates; Comparing the first trajectory curve, the second trajectory curve and the third trajectory curve in pairs; Based on the comparison result, it is determined whether the target vehicle is in an error-prone scenario.

3. The yaw determination method according to claim 2, characterized in that: The determining, based on the comparison result, whether the target vehicle is in an error-prone scenario includes: If the second trajectory curve does not overlap with the third trajectory curve, the similarity between any two of the first trajectory curve, the second trajectory curve and the third trajectory curve is greater than a preset similarity threshold, and the similarity between the first trajectory curve and the second trajectory curve is greater than the similarity between the first trajectory curve and the third trajectory curve, it is determined that the target vehicle is in an error-prone scenario.

4. The yaw determination method according to claim 1 or 2, characterized in that: The determining whether a deviation misjudgment occurs includes: If the target vehicle is in an error-prone scenario, determining the comprehensive probability between different roads in the actual driving path; Determining the best matching path other than the actual driving path from the candidate paths based on the product of the comprehensive probability and the preset penalty weight; Comparing the optimal matching path with the planned path of the target vehicle before the deviation; Based on the comparison results, it is determined whether a yaw misjudgment occurs.

5. The yaw determination method according to claim 1, characterized in that: After determining whether the target vehicle is in an error-prone scenario, the method further includes: If the target vehicle is in a non-error-prone scenario, it is determined whether a yaw misjudgment occurs.

6. The yaw determination method according to claim 5, characterized in that: The determining whether a deviation misjudgment occurs includes: determining a yaw path corresponding to the yaw trajectory; comparing the yaw path with a planned path of the target vehicle before the yaw; Based on the comparison results, determine whether a yaw misjudgment occurs.

7. The yaw determination method according to claim 1, characterized in that: The determining of the yaw trajectory based on the actual yaw time, the time of re-planning the path, and the actual driving path includes: Determining a target yaw trajectory point based on the actual yaw time; Determining a starting trajectory point and an ending trajectory point in the actual driving path based on the target off-course trajectory point; The yaw trajectory is obtained based on the start trajectory point and the end trajectory point.

8. The yaw determination method according to any one of claims 1 to 7, characterized in that: After determining whether a deviation misjudgment occurs, the method further includes: When no yaw misjudgment occurs, determining a yaw misjudgment rate of the navigation engine; Determine the yaw response time for this navigation determination result; The yaw determination effect of the navigation engine is evaluated based on the yaw error rate and the yaw response time.

9. A yaw determination device, characterized in that: The device comprises: A deviation information acquisition unit, used to acquire original track point information, actual deviation time, re-planned path information and re-planned path time of the target vehicle when the target vehicle deviates; An actual driving path determining unit, used to determine the actual driving path of the target vehicle based on the original trajectory point information; a yaw trajectory determination unit, configured to determine a yaw trajectory based on the actual yaw time, the time for re-planning the path, and the actual driving path; An error-prone scene judgment unit, used to determine whether the target vehicle is in an error-prone scene based on the yaw trajectory and the planned path of the target vehicle before the yaw; The yaw determination unit is used to determine whether a yaw misjudgment occurs if the target vehicle is in an error-prone scenario.

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

Images