JP5573780B2 - Course evaluation device and course evaluation method - Google Patents

Description

  The present invention relates to a route evaluation apparatus and a route evaluation method for evaluating a route candidate of a host vehicle.

  Conventionally, what was disclosed by patent document 1 as a technique of such a field | area is known. The apparatus described in Patent Document 1 constitutes an evaluation function that calculates the desirability of driving operation of the host vehicle from the state of the host vehicle and the state of other surrounding vehicles, and the evaluation function and the prediction result of the behavior of the other vehicle Based on this, the desired operation for the vehicle is calculated.

JP 2004-152125 A

  However, just paying attention to the relationship between the host vehicle and the other vehicle as in the above-described device, there is a risk of selecting an operation that disturbs the traffic flow in a crowded situation or the like. Specifically, when the own vehicle changes lanes in a crowded situation, the lane change cannot be completed and the traffic of other vehicles is obstructed unless the lane change destination has enough distance between the vehicles. Therefore, if the lane change is performed even at a short inter-vehicle distance, the traffic flow may be disturbed by forcibly changing the lane.

  Therefore, the present invention can perform route evaluation in consideration of traffic flow by performing route evaluation based on the similarity between the standard locus obtained by standardizing the movement locus of the preceding vehicle and the route candidate of the host vehicle. It is an object to provide a course evaluation apparatus and a course evaluation method.

  In order to solve the above-described problem, a course evaluation apparatus according to the present invention obtains course candidate generation means for generating a course candidate of the own vehicle and movement trace information related to the movement trace of a preceding vehicle on the running road of the own vehicle. Based on the movement trajectory information, the movement trajectory information acquisition means, the standard trajectory calculation means for calculating the standard trajectory that standardizes the movement trajectory of the preceding vehicle, and the calculation of the similarity between the course candidate of the host vehicle and the standard trajectory Similarity calculation means for performing, and evaluation means for evaluating a course candidate of the host vehicle based on the similarity.

  According to the route evaluation apparatus according to the present invention, since the route candidate close to the movement of the preceding vehicle is likely to be along the flow of surrounding traffic, the standard locus obtained by standardizing the moving locus of the preceding vehicle and the vehicle itself are considered. By evaluating the route candidates based on the similarity to the vehicle route candidates, it is possible to perform the route evaluation in consideration of the traffic flow.

In the course evaluation device according to the present invention, the movement trajectory information acquisition means may acquire movement trajectory information related to the movement trajectory of the preceding vehicle on the traveling lane of the own vehicle based on road map information around the own vehicle. preferable.
According to the course evaluation apparatus according to the present invention, by referring to the road map information, it is possible to avoid the acquisition of unnecessary movement trajectory information such as other vehicles located outside the road. Information can be narrowed down, and the speed of calculation processing related to the course evaluation can be improved.

In the course evaluation apparatus according to the present invention, course prediction means for predicting the course of other vehicles around the host vehicle, and determination means for determining an interference mode between the course candidate of the host vehicle and the predicted course of the other vehicle are further provided. Preferably, the evaluation means evaluates the course candidate of the host vehicle based on the determination result of the determination means.
According to the route evaluation apparatus according to the present invention, the route candidate of the own vehicle is evaluated based on the form of interference between the route candidate of the own vehicle and the predicted route of the other vehicle, so that the own vehicle interferes with the other vehicle. The course evaluation can be performed in consideration of the situation to be avoided.

In the course evaluation apparatus according to the present invention, the determination means determines whether the other vehicle has its own vehicle depending on whether an interference parameter obtained by quantifying the form of interference between the course candidate of the host vehicle and the predicted course of the other vehicle exceeds a determination threshold. It is preferable to determine whether or not the vehicle interferes with the vehicle or whether the host vehicle interferes with another vehicle and change the determination threshold based on the similarity.
According to the course evaluation device of the present invention, when the degree of similarity between the course candidate of the own vehicle and the standard trajectory of the preceding vehicle is high, the course candidate follows the traffic flow and the own vehicle interferes with other vehicles. Therefore, by changing the determination threshold based on the degree of similarity, it is possible to determine the interference mode in consideration of the traffic flow.

The present invention is a course evaluation method using the above-described course evaluation apparatus, wherein a course candidate generation step of generating a course candidate of the host vehicle by a path candidate generation unit and a movement trajectory information acquisition unit on a traveling road of the host vehicle. A standard that calculates a standard trajectory that standardizes the travel trajectory of the preceding vehicle based on the travel trajectory information by a travel trajectory information acquisition step that acquires travel trajectory information related to the travel trajectory of the preceding vehicle. A trajectory calculation step, a similarity calculation step for calculating the similarity between the course candidate of the host vehicle and the standard trajectory by the similarity calculation means, and an evaluation means for evaluating the course candidate of the host vehicle based on the similarity. An evaluation step to be performed.

  According to the course evaluation method according to the present invention, a course candidate close to the movement of the preceding vehicle is likely to be along the flow of surrounding traffic. By evaluating the route candidates based on the similarity to the vehicle route candidates, it is possible to perform the route evaluation in consideration of the traffic flow.

  According to the present invention, the route evaluation can be performed in consideration of the traffic flow by performing the route evaluation based on the similarity between the standard locus obtained by standardizing the movement locus of the preceding vehicle and the route candidate of the host vehicle. .

It is a block diagram which shows the course evaluation apparatus which concerns on 1st Embodiment. It is a figure which shows the example of the movement locus | trajectory of the preceding vehicle on the one-side three lane road, and the example of a standard locus | trajectory. It is a figure which shows the course candidate and standard locus | trajectory in case a own vehicle changes lanes under congestion conditions. It is a figure for demonstrating the interference parameter and interference form determination map which are used in 1st Embodiment. It is a flowchart which shows the course evaluation method which concerns on 1st Embodiment. It is a figure which shows the example of the driving assistance implement | achieved based on the evaluation result of a course evaluation apparatus. It is a figure which shows the other example of the driving assistance implement | achieved based on the evaluation result of a course evaluation apparatus. It is a figure which shows the other example of the driving assistance implement | achieved based on the evaluation result of a course evaluation apparatus. It is a block diagram which shows the course evaluation apparatus which concerns on 2nd Embodiment. It is a flowchart which shows the course evaluation method which concerns on 2nd Embodiment.

  DESCRIPTION OF EMBODIMENTS Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings. In addition, in each figure, the same code | symbol is attached | subjected to the same or an equivalent part, and the overlapping description is abbreviate | omitted.

[First embodiment]
As shown in FIG. 1, the course evaluation device 1 according to the first embodiment performs course evaluation of the host vehicle M based on traffic conditions around the host vehicle. The course evaluation apparatus 1 enables driving support of the host vehicle M based on the course evaluation result by being incorporated in the driving support apparatus of the host vehicle M, for example.

  The course evaluation device 1 uses the movement trajectory information regarding the movement trajectory K of the preceding vehicle to realize the course evaluation in consideration of the traffic flow. The preceding vehicle means a vehicle that has traveled before the host vehicle M on the traveling road on which the host vehicle M travels.

  The course evaluation device 1 first generates a course candidate R for the host vehicle M. The route evaluation device 1 generates a route candidate R based on a preset destination, surrounding traffic conditions, and the like. The course evaluation device 1 generates a plurality of course candidates R depending on the situation. The “course” here refers to a concept including temporal elements such as time and speed, and is different from a “route” not including these temporal elements.

  When the route evaluation device 1 generates the route candidate R, the route evaluation device 1 acquires the movement track information of the preceding vehicle on the traveling road of the host vehicle M.

  The movement trajectory information of the preceding vehicle can be acquired by tracking the movement of the preceding vehicle using an in-vehicle device such as a camera or a radar sensor. When using an in-vehicle device, it is possible to acquire movement trajectory information by tracking the preceding vehicle for about 100 m in distance or about 1 minute in time. In particular, when the movement trajectory information is acquired from the image information of the camera, it is preferable that the same position of the front vehicle (for example, the upper right corner of the rear license plate) is a tracking target in order to perform accurate tracking. A method using estimation by a Kalman filter or a particle filter is known for tracking a preceding vehicle. In addition, the movement trajectory information can also be acquired from the traveling history of the preceding vehicle received by inter-vehicle communication.

  Moreover, the course evaluation apparatus 1 acquires the movement trajectory information of the preceding vehicle using road-to-vehicle communication with a traffic information center that handles road traffic information. In the traffic information center, the movement trajectory information of the preceding vehicle is generated based on the road traffic information of the traveling road on which the host vehicle M is traveling. The traffic information center generates movement trajectory information by detecting the movement trajectory of the preceding vehicle from image information of a traffic monitoring camera installed on the road, detection information of a loop coil embedded in the road, and the like. When generating the movement trajectory information from the image information of the traffic monitoring camera, it is preferable that the part where the front vehicle is in contact with the road surface (for example, the right rear wheel) is to be tracked. The course evaluation device 1 acquires the movement trajectory information generated by the traffic information center through road-to-vehicle communication. Note that the traffic information center may provide only road traffic information, and the route evaluation device 1 may generate (acquire) movement trajectory information.

  Next, the course evaluation apparatus 1 sequentially stores the acquired movement trajectory information. The course evaluation apparatus 1 stores the movement trajectory K as a parameter discretized at a constant distance interval. For example, the course evaluation apparatus 1 stores each point obtained by discretizing the movement trajectory K as a parameter including three elements of the position X, Y on the plane and the direction θ of the preceding vehicle, thereby greatly increasing the storage capacity. Save money. Alternatively, the course evaluation apparatus 1 may realize saving of the storage capacity by storing the movement locus K as a combination of line segments such as a straight line, an arc, and a clothoid curve. In this case, the course evaluation apparatus 1 can store the movement locus K as a parameter of a matrix in which two elements of the curvature c and the line segment length d are arranged in order.

  Furthermore, the course evaluation apparatus 1 saves storage capacity by deleting unnecessary movement trajectory information. Unnecessary movement trajectory information is information on a location where the host vehicle M has passed or information that has become out of date. Note that the movement trajectory information is preferably stored and deleted in units of segments, not in units of trajectories. With respect to the movement trajectory information of a certain preceding vehicle, the movement trajectory of the section where the host vehicle has passed is deleted, and newly acquired movement trajectories are sequentially added.

  Subsequently, the course evaluation apparatus 1 calculates a standard trajectory H obtained by standardizing the travel trajectory K of the preceding vehicle based on the acquired travel trajectory information of the preceding vehicle. The movement trajectory K is standardized by, for example, a standardization process (average process) of the movement trajectory K by a statistical method or a process of adjusting the movement trajectory K to be a smooth line.

  As a statistical method for standardizing the movement locus K, for example, a clustering method is used. Hereinafter, a method using the k-means method among the clustering methods will be described. Here, FIG. 2 (a) is a diagram showing an example of the movement trajectory of a preceding vehicle on a one-side three-lane road, and FIG. 2 (b) is a standard trajectory H calculated from the movement trajectory K of FIG. 2 (a). It is a figure which shows the example of. The movement trajectory of the preceding vehicle is shown as a thin line K, and the standard trajectory is shown as a black circle H.

  In the k-means method, the upper limit of the number of clusters is first determined. The upper limit of the number of clusters is determined as a value greater than or equal to the number of lanes on the traveling road ahead of the host vehicle. In the case of the one-side three lane shown in FIG. 2A, the upper limit of the number of clusters is determined to be 5, for example. When the number of lanes is unknown, the upper limit of the number of clusters is determined assuming a general road.

  Subsequently, classification by the k-means method is sequentially performed on the number of clusters from 1 to the upper limit value. Thereafter, the classification results of the k-means method for different numbers of clusters are compared, and the number of clusters having the smallest sum or maximum value of intra-cluster variance is obtained as the number of standard trajectories. In the case shown in FIG. 2 (a), the variance within the cluster is minimized when it is classified into five clusters, and the number of standard trajectories is six.

  In this k-means method, the center (average) of each cluster is calculated as a parameter indicating a standard locus. The bundle of movement trajectories K shown in FIG. 2A is classified into five clusters, and six standard trajectories H shown in FIG. 2B are calculated as the centers of the clusters. The standard trajectory H is calculated in an appropriate data format. For example, the standard trajectory H may be expressed as a collection of points discretized at a constant distance interval, or may be expressed as a combination of continuous line segments.

  After the calculation of the standard trajectory H, the course evaluation device 1 calculates the similarity between the standard trajectory H of the preceding vehicle and the course candidate R of the host vehicle M. Here, Fig.3 (a) is a figure explaining the condition where the own vehicle M changes lanes under congestion conditions. FIG. 3B is a diagram for explaining the standard trajectory H calculated under the situation of FIG. In FIG. 3B, three standard trajectories H1 to H3 are calculated based on the movement trajectory K of the preceding vehicle. In FIG. 3, the preceding vehicle on the traveling lane is indicated as F, and the other vehicle traveling in the adjacent lane is indicated as N.

  In the situation shown in FIG. 3A and FIG. 3B, the course evaluation device 1 compares the course candidate R of the host vehicle M with the standard trajectories H1 to H3, and the standard trajectory H2 closest to the course candidate R. The similarity is calculated for. Note that the similarity may be calculated for all the standard trajectories H1 to H3. Moreover, it is preferable that the course candidate R and the standard trajectory H are generated in a data format that can be compared in advance. If the data formats are different, it is necessary to convert either to a comparable data format.

  When the standard trajectory H is calculated by statistical processing, the course evaluation device 1 can calculate the similarity between the path candidate R and the standard trajectory H using a so-called Mahalanobis distance. By calculating the Mahalanobis distance for each standard trajectory H, the standard trajectory H of the cluster having the highest probability that the course candidate R belongs can be obtained, and the similarity is calculated from the Mahalanobis distance between the standard trajectory H and the course candidate R. can do.

  Further, when the course candidate R and the standard trajectory H are discretized at a constant distance interval, the course evaluation device 1 determines the difference in the Euclidean distance and the angle (orientation) between the closest points for each discretized point. The similarity may be calculated using the difference between the two. In this case, for example, by applying an appropriate coefficient to the angle difference to match the Euclidean distance and the unit, the calculation of the similarity according to the sum of the Euclidean distance difference and the angle difference can be realized.

  The course evaluation device 1 may determine whether the calculated similarity is lower than a predetermined threshold. When the similarity is lower than a predetermined threshold, the course evaluation device 1 determines that the influence of the similarity on the course evaluation is small and does not perform the course evaluation based on the similarity, thereby performing an extra calculation processing work. Can be avoided. When the Mahalanobis distance is used for the similarity calculation, for example, the route evaluation is not performed when the route candidate R is not within the standard deviation 3σ within the cluster of the standard trajectory H. Calculation processing work can be avoided.

  The course evaluation device 1 also predicts courses for other vehicles around the host vehicle M. The course evaluation device 1 predicts the course of the other vehicle based on the position and the traveling direction of the other vehicle. The course evaluation device 1 determines the interference mode between the predicted course of the other vehicle and the course candidate of the host vehicle when the course of the other vehicle is predicted. Here, “interference” means that the courses of the host vehicle and the other vehicle intersect in a plane in consideration of the vehicle width and the vehicle length.

  Next, the course evaluation device 1 determines an interference mode between the course candidate R of the host vehicle M and the predicted course of another vehicle. The course evaluation device 1 determines whether the other vehicle interferes with the own vehicle M or whether the own vehicle M interferes with the other vehicle. The course evaluation device 1 determines the degree of interference between the host vehicle M and other vehicles.

  The form in which the other vehicle interferes with the own vehicle M is, for example, a case in which the other vehicle enters the front of the own vehicle M by changing the lane, and the form in which the own vehicle M interferes with the other vehicle. Is a case where the host vehicle M enters the front of another vehicle by changing the lane.

  The determination of the interference form is preferably performed using, for example, an interference form parameter. The interference mode parameter is a parameter obtained by quantifying the interference mode between the host vehicle M and the other vehicle. For example, the parameter increases as the degree of interference between the host vehicle M increases, or increases as the degree of interference between the other vehicle increases. Will be used. As the interference mode parameter, a vehicle travel information value corresponding to the interference mode is used. For example, the distance between the host vehicle M and the other vehicle, the relative speed between the host vehicle M and the other vehicle, or a value using those values. Is used.

  A plurality of different interference pattern determination parameters may be used as the interference pattern determination parameter. For example, the inter-vehicle distance and the relative speed may be used as the interference form parameters.

  The determination of the interference form using the interference form parameter is preferably performed using an interference form determination map. As shown in FIG. 4, the interference form determination map is a map for determining whether the own vehicle is interfering with another vehicle or whether the other vehicle is interfering with the own vehicle based on the interference form parameter. is there. In this interference form determination map, a determination threshold T for determining the interference form is set in advance, and the interference form is determined depending on whether or not the interference form determination parameter exceeds the determination threshold T. .

  The determination threshold value T is used as a determination reference value when determining the interference mode between the host vehicle M and another vehicle. The initial value of the determination threshold T is set according to the positional relationship between the host vehicle M and another vehicle.

  When the course evaluation device 1 determines the similarity between the course candidate R of the host vehicle and the standard trajectory H of the preceding vehicle, the course evaluation apparatus 1 changes the determination threshold T based on the similarity. For example, the higher the similarity between the standard trajectory H of the preceding vehicle and the route candidate R of the host vehicle, the higher the possibility that the route candidate R will follow the traffic flow and the host vehicle may interfere with other vehicles. Therefore, by changing the determination threshold value T to the value on the right side of FIG. 3, it is easy to determine that another vehicle interferes with the host vehicle M.

  The course evaluation device 1 evaluates the course candidate R based on the above-described similarity and interference mode determination results. The course evaluation device 1 evaluates the course candidate R higher as the similarity between the course candidate R of the host vehicle and the standard trajectory H of the preceding vehicle is higher. In addition, the course evaluation device 1 evaluates the course candidate R higher when it is determined that the other vehicle interferes with the own vehicle than when the own vehicle determines that the other vehicle interferes with the other vehicle. Moreover, the course evaluation apparatus 1 evaluates the course candidate R higher as the degree of interference of the host vehicle with other vehicles is smaller.

  Next, the configuration of the course evaluation apparatus 1 according to the first embodiment will be described.

  As shown in FIG. 1, the course evaluation device 1 includes a course evaluation ECU [Electronic Control Unit] 2 that comprehensively controls the apparatus. The course evaluation ECU 2 is an electronic control unit including a CPU [Central Processing Unit], a ROM [Read Only Memory], a RAM [Random Access Memory], and the like.

  The course evaluation ECU 2 is connected to the radar sensor 3, the camera 4, the road-vehicle communication unit 5, and the vehicle state detection unit 6.

  The radar sensor 3 is a sensor that detects an object using radio waves. The radar sensor 3 detects an object by emitting a radio wave every predetermined time and receiving a reflected wave from the object. The radar sensor 3 outputs object detection information related to the object detection result to the course evaluation ECU 2. The camera 4 is an in-vehicle camera that captures an image of the surroundings of the host vehicle M. The camera 4 outputs the captured image information around the host vehicle to the course evaluation ECU 2.

  The road-to-vehicle communication unit 5 performs road-to-vehicle communication with a traffic information center that handles road traffic information using a wireless communication network. The road-to-vehicle communication unit 5 acquires the movement locus information of the preceding vehicle generated by the traffic information center by road-to-vehicle communication. The road-vehicle communication unit 5 outputs the acquired movement trajectory information to the course evaluation ECU 2.

  The vehicle state detection unit 6 detects vehicle state information such as position information and vehicle speed information of the host vehicle M, and for example, a GPS [Global Positioning System] or a wheel speed sensor is used. The GPS acquires vehicle position information. The wheel speed sensor is attached to a wheel portion of the vehicle, for example, and acquires vehicle speed information from the rotational speed of the wheel. The vehicle state detection unit 6 outputs the acquired vehicle state information to the course evaluation ECU 2.

  As shown in FIG. 1, the course evaluation ECU 2 includes a course candidate generation unit (course candidate generation unit) 11, a movement track information acquisition unit (movement track information acquisition unit) 12, an information storage unit 13, and a standard track calculation unit ( Standard trajectory calculation means) 14 is provided. Further, the course evaluation ECU 2 includes a similarity calculation unit (similarity calculation unit) 15, a course prediction unit 16, an interference form determination unit (determination unit) 17, and a course evaluation unit (path evaluation unit) 18.

  The route candidate generation unit 11 generates a route candidate R for the host vehicle. The generation of the route candidate R can be performed by various well-known methods. The route candidate generation unit 11 generates a route candidate R based on destination information input in advance and vehicle state information of the vehicle state detection unit 6. The route candidate generation unit 11 preferably generates the route candidate R in consideration of the safety of the host vehicle M, traffic rules, passenger comfort, and the like.

  The movement locus information acquisition unit 12 acquires movement locus information of the preceding vehicle. The movement trajectory information acquisition unit 12 acquires the movement trajectory information by tracking the preceding vehicle using the object detection information of the radar sensor 3 and the image information of the camera 4. For example, it is possible to detect and track the preceding vehicle by extracting information on the intensity of the radio wave and the speed of the object from the object detection information of the radar sensor 3. Further, for the image information of the camera 4, it is possible to detect and track the preceding vehicle by template matching. Alternatively, the preceding vehicle may be defined by defining the preceding vehicle as a collection of feature amounts (corner or edge of the preceding vehicle outline, difference in luminance, etc.) and extracting the feature amount from the image.

  Further, the movement trajectory information acquisition unit 12 acquires the movement trajectory information of the preceding vehicle by the output from the road-vehicle communication unit 5.

  The information storage unit 13 is a database that stores movement trajectory information. The information storage unit 13 stores information in an appropriate data format for saving storage capacity. The information storage unit 13 also saves storage capacity by deleting unnecessary information such as movement trajectory information of a location where the host vehicle M has passed based on the vehicle state information of the vehicle state detection unit 6 every predetermined time. Do.

  The standard trajectory calculation unit 14 calculates a standard trajectory H obtained by standardizing the travel trajectory of the preceding vehicle based on the travel trajectory information stored in the information storage unit 13. The standard trajectory calculation unit 14 calculates the standard trajectory H by standardizing the movement trajectory using the clustering method described above.

  The similarity calculation unit 15 calculates the similarity between the route candidate R of the host vehicle M generated by the route candidate generation unit 11 and the standard locus H of the preceding vehicle calculated by the standard locus calculation unit 14. The similarity calculation unit 15 calculates the similarity by, for example, comparing parameters for each point obtained by discretizing the course candidate R and the standard trajectory H at a constant distance interval.

  The course prediction unit 16 performs course prediction of other vehicles that travel around the host vehicle M. The course prediction unit 16 detects other vehicles around the host vehicle M based on the object detection information of the radar sensor 3 and the image information of the camera 4. The course prediction unit 16 predicts a course based on the detected position and orientation of the other vehicle. The course prediction of other vehicles can be performed by various known methods.

  The interference form determination unit 17 determines an interference form between the predicted course of the other vehicle and the course candidate of the host vehicle M when the course prediction unit 16 predicts the course of the other vehicle. Here, “interference” means that the course of the host vehicle M and the other vehicle intersects in a plane in consideration of the vehicle width and the vehicle length.

  The interference form determination unit 17 determines the interference form between the route candidate R of the host vehicle M and the predicted route of the other vehicle. The interference form determination unit 17 calculates the above-described interference form parameter based on the course candidate R of the host vehicle M and the predicted course of the other vehicle.

  Next, the interference form determination unit 17 acquires the initial value of the determination threshold T described above based on the positional relationship between the host vehicle M and another vehicle. The interference form determination unit 17 changes the determination threshold T based on the similarity calculated by the similarity calculation unit 15. The interference form determination unit 17 determines the interference form based on the interference form parameter and the determination threshold value T.

  The course evaluation unit 18 evaluates the course candidate R generated by the course candidate generation unit 11 based on the similarity calculated by the similarity calculation unit 15 and the determination result of the interference form determination unit 17. The course evaluation method can be performed by various known methods.

  Next, the course evaluation method by the course evaluation apparatus 1 will be described.

  As shown in FIG. 5, first, in step S <b> 1, the route candidate generation unit 11 generates a route candidate R (route candidate generation step). Next, in step S2, the movement locus information acquisition unit 12 acquires the movement locus information of the preceding vehicle, and the information storage unit 13 stores the movement locus information (movement locus information acquisition step). The movement trajectory information acquisition unit 12 acquires the movement trajectory information by tracking the preceding vehicle with the radar sensor 3 and the camera 4. Further, the movement trajectory information acquisition unit 12 acquires the movement trajectory information based on the output from the road-vehicle communication unit 5. The information storage unit 13 stores the movement trajectory information in an appropriate data format. The conversion of the data format of the movement trajectory information may be performed in advance by the movement trajectory information acquisition unit 12.

  In step S3, the standard trajectory calculation unit 14 calculates the standard trajectory H of the preceding vehicle (standard trajectory calculation step). The standard trajectory calculation unit 14 calculates the standard trajectory H by standardizing the movement trajectory K using a clustering method or the like. Thereafter, in step S4, the similarity calculation unit 15 calculates the similarity (similarity calculation step).

  In step S5, the course prediction unit 16 performs the course prediction of the other vehicle. The course prediction unit 16 performs course prediction on other vehicles detected by the object detection information of the radar sensor 3 and the image information of the camera 4. This step S5 may be performed at another stage of the flow.

  In step S6, the determination threshold value T is changed by the interference pattern determination unit 17. The interference form determination unit 17 first acquires an initial value of the determination threshold value T. The interference form determination unit 17 changes the determination threshold value T based on the similarity calculated by the similarity calculation unit 15.

  In step S7, the interference mode determination unit 17 determines the interference mode. The interference form determination unit 17 calculates an interference form parameter based on the course candidate R of the host vehicle M and the predicted course of the other vehicle. The interference pattern determination unit 17 determines the interference pattern between the route candidate R of the host vehicle M and the predicted track of the other vehicle using the calculated interference parameter and the determination threshold value T.

  Thereafter, in step S8, the route evaluation unit 18 evaluates the route candidate R of the host vehicle M (route evaluation step). The course evaluation unit 18 evaluates the course candidate R of the host vehicle M based on the similarity calculated by the similarity calculation unit 15 and the determination result of the interference form determination unit 17. Thereafter, the course evaluation for the course candidate R is terminated. The course evaluation described above is repeatedly performed for each course candidate R. In addition, the aspect which evaluates collectively about several course candidate R may be sufficient.

  According to the course evaluation apparatus 1 according to the first embodiment described above, the course candidate R close to the movement of the preceding vehicle is likely to follow the flow of surrounding traffic. By evaluating the route candidate R on the basis of the similarity between the standard track H obtained by standardizing the track K and the route candidate R of the host vehicle M, it is possible to perform the route evaluation in consideration of the traffic flow.

  In addition, according to the route evaluation apparatus 1, by using a standard locus obtained by standardizing the movement locus of the preceding vehicle as a target for similarity determination, an appropriate route evaluation based on a large number of movement locus can be performed in a short time. It becomes possible.

  Furthermore, according to this course evaluation device 1, the course candidate R of the host vehicle M is evaluated based on the form of interference between the course candidate R of the host vehicle M and the predicted course of the other vehicle. On the other hand, the course evaluation can be performed in consideration of the situation to be avoided such as interference.

  Further, according to the route evaluation apparatus 1, when the similarity between the route candidate R of the host vehicle M and the standard trajectory H of the preceding vehicle is high, the route candidate R follows the traffic flow and the host vehicle M Since the probability that the vehicle interferes with other vehicles is considered to be low, it is possible to realize the determination of the interference form in consideration of the traffic flow by changing the determination threshold T based on the similarity. Therefore, according to the route evaluation apparatus 1, since it is possible to determine the interference mode in consideration of the traffic flow, it is possible to perform an appropriate route evaluation corresponding to the traffic situation in the real world based on the determination result.

  Here, an example of driving support using the evaluation result of the course evaluation device 1 will be described with reference to the drawings. FIGS. 6-8 has shown the example of the driving assistance which concerns on the lane change of the own vehicle M. FIG.

FIG. 6A is a diagram illustrating a situation in which one other vehicle N exists on the adjacent lane to which the lane is changed. FIG. 6B is a diagram showing a driving support result under the situation of FIG. In the situation shown in FIG. 6A, the course evaluation device 1 determines that the course candidate R1 of the host vehicle M related to the lane change interferes with the predicted course L of the other vehicle N.

  In this case, the course evaluation device 1 can avoid the driving support based on the course candidate R1 by evaluating the course candidate R1 low based on the determination result. Thereafter, as illustrated in FIG. 6B, the course evaluation device 1 generates a course candidate R <b> 2 that changes the lane at the timing when the other vehicle N has overtaken the host vehicle M. The course evaluation device 1 determines that the course candidate R2 does not interfere with the predicted course L of the other vehicle N, and evaluates the course candidate R2 higher than the course candidate R1. Thereby, the course evaluation device 1 can avoid the driving support based on the path candidate R1 that interferes with the other vehicle N, and can execute the driving support based on the path candidate R2 that waits for the other vehicle N to pass and changes the lane. .

FIG. 7A is a diagram illustrating a situation in which another vehicle N is congested on the adjacent lane to which the lane is changed. FIG.7 (b) is a figure which shows the driving | running | working assistance result under the condition of Fig.7 (a). In the situation shown in FIG. 7A, the course evaluation device 1 calculates standard trajectories H4 and H5 based on the acquired movement standard information of the preceding vehicle. The standard trajectory H4 is a trajectory that goes straight in the current travel lane, and the standard trajectory H5 is a trajectory for changing the lane from the travel lane to the adjacent lane.

  In this case, the course evaluation device 1 calculates the similarity with the course candidate R3 for the standard trajectory H5 close to the course candidate R3 of the host vehicle M related to the lane change. The course evaluation device 1 is configured such that the course candidate R3 interferes with the predicted course of the other vehicle N. However, since the degree of similarity between the course candidate R3 and the standard locus H5 is high, the course candidate R3 is highly evaluated. Thereby, the course evaluation device 1 can carry out the driving support based on the course candidate R3 without giving up the lane change just because the own vehicle M interferes with the other vehicle N. In addition, by changing the lane on a route similar to the standard trajectory H5 of the preceding vehicle, it is possible to realize a lane change with little influence on the other vehicle N without disturbing the traffic flow.

FIG. 8A is a diagram illustrating a situation where another vehicle N is congested on the adjacent lane to which the lane is changed. FIG. 8A differs from FIG. 7A in that the adjacent lane is a right-turn exclusive lane. FIG. 8B is a diagram showing a driving support result under the situation of FIG. In the situation shown in FIG. 8A, the course evaluation device 1 calculates the standard trajectory H6 based on the acquired movement standard information of the preceding vehicle. The standard trajectory H6 is a trajectory that goes straight in the current travel lane.

  In this case, the course evaluation device 1 calculates the similarity between the course candidate R4 of the host vehicle M related to the lane change and the standard locus H6. The course evaluation device 1 calculates the similarity low because the course candidate R4 and the standard locus H6 are not similar. The course evaluation device 1 is a form in which the course candidate R4 interferes with the predicted course of the other vehicle N and has a low similarity to the standard locus H6. Therefore, the course evaluation apparatus 1 evaluates the course candidate R4 low, and is based on the course candidate R4. Driving assistance can be avoided.

  Thereafter, as shown in FIG. 8B, the course evaluation device 1 generates a course candidate R5 along the travel lane of the host vehicle M. The course evaluation device 1 highly evaluates the course candidate R5 because the course candidate R5 does not interfere with the predicted course L of the other vehicle N and has a high similarity to the standard locus H6. As a result, the route evaluation device 1 can avoid the driving support for performing an unreasonable lane change that disturbs the traffic flow, and the driving support based on the route candidate R5 that goes straight in the driving lane like the other preceding vehicles. Can be implemented.

[Second Embodiment]
As shown in FIG. 9, the course evaluation apparatus 20 according to the second embodiment includes a road map information storage unit 23 that stores road map information, and a movement trajectory information acquisition unit 24 uses road map information. Thus, the movement trajectory information is acquired and the standard trajectory calculation unit 25 calculates the standard trajectory H using the road map information.

  Specifically, the course evaluation ECU 21 according to the course evaluation apparatus 20 includes a road map information storage unit 23. The road map information storage unit 23 stores road map information including information such as the number of road lanes, road shape, and direction of travel of the lane. The road map information storage unit 23 acquires road map information around the host vehicle based on the position information of the host vehicle M included in the vehicle state information from the vehicle state detection unit 6.

  The movement track information acquisition unit 24 acquires the movement track information of the preceding vehicle based on the road map information around the host vehicle. The movement locus information acquisition unit 24 refers to road map information around the host vehicle, thereby avoiding unnecessary movement locus information acquisition. The unnecessary movement trajectory information mentioned here includes movement trajectory information of a vehicle traveling on the opposite lane, movement trajectory information of a vehicle traveling off the road, and the like. The movement trajectory information acquisition unit 24 selectively acquires necessary movement trajectory information based on road map information around the host vehicle.

  The standard trajectory calculation unit 25 calculates a standard trajectory H that standardizes the travel trajectory of the preceding vehicle based on the road map information around the host vehicle and the travel trajectory information stored in the information storage unit 13. By referring to the road map information around the host vehicle in this way, the upper limit of the number of clusters can be appropriately determined according to the number of lanes in the standardization using the k-means method of the clustering method described above. In general, the movement trajectory K is formed along the lane, so that it can be used as an appropriate initial value in clustering by recognizing the lane shape and the like from the road map information. Thereby, the stability of the statistical process by a k-means method can be improved, and the reliability which concerns on the calculation of the standard locus | trajectory H can be aimed at.

  Next, a course evaluation method by the course evaluation apparatus 21 according to the second embodiment will be described.

  As shown in FIG. 10, first, in step S11, a route candidate R is generated by the route candidate generation unit 11 (track candidate generation step). Next, in step S12, road map information is acquired by the road map information storage unit 23. The road map information storage unit 23 acquires road map information around the host vehicle based on the position information of the host vehicle M included in the vehicle state information from the vehicle state detection unit 6.

  Subsequently, in step S13, the movement trajectory information of the preceding vehicle is acquired by the movement trajectory information acquisition unit 24, and the movement trajectory information is stored by the information storage unit 13 (movement trajectory information acquisition step). The movement trajectory information acquisition unit 24 selectively acquires necessary movement trajectory information based on road map information around the host vehicle.

  In step S14, the standard trajectory calculation unit 25 calculates the standard trajectory H of the preceding vehicle (standard trajectory calculation step). The standard trajectory calculation unit 25 calculates the standard trajectory H by standardizing the movement trajectory K using a clustering method or the like. The standard trajectory calculation unit 25 standardizes the movement trajectory K with reference to road map information around the host vehicle. Thereafter, in step S15, the similarity calculation unit 15 calculates the similarity (similarity calculation step).

  In step S <b> 16, the course prediction unit 16 performs the course prediction of the other vehicle. The course prediction unit 16 performs course prediction on other vehicles detected by the object detection information of the radar sensor 3 and the image information of the camera 4. This step S16 may be performed at another stage.

  In step S17, the determination threshold value T is changed by the interference pattern determination unit 17. The interference form determination unit 17 first acquires an initial value of the determination threshold value T. The interference form determination unit 17 changes the determination threshold value T based on the similarity calculated by the similarity calculation unit 15.

  In step S18, the interference form determination unit 17 determines the interference form. The interference form determination unit 17 calculates an interference form parameter based on the course candidate R of the host vehicle M and the predicted course of the other vehicle. The interference pattern determination unit 17 determines the interference pattern between the route candidate R of the host vehicle M and the predicted track of the other vehicle using the calculated interference parameter and the determination threshold value T.

  Thereafter, in step S19, the route evaluation unit 18 evaluates the route candidate R of the host vehicle M (route evaluation step). The course evaluation unit 18 evaluates the course candidate R of the host vehicle M based on the similarity calculated by the similarity calculation unit 15 and the determination result of the interference form determination unit 17. Thereafter, the course evaluation for the course candidate R is terminated. The course evaluation described above is repeatedly performed for each course candidate R. In addition, the method of evaluating collectively about several course candidate R may be sufficient.

  According to the course evaluation apparatus 21 according to the second embodiment described above, it is possible to obtain the same effects as those of the course evaluation apparatus 1 according to the first embodiment. Further, according to the course evaluation device 21, the movement trajectory information acquisition unit 24 refers to the road map information, and can selectively acquire necessary movement trajectory information while avoiding acquisition of unnecessary movement trajectory information. Therefore, it is possible to narrow down the information that is subject to the calculation processing of the standard trajectory H, and it is possible to improve the speed of the calculation processing related to the course evaluation. In addition, since the calculation load of the arithmetic processing related to the course evaluation can be reduced, it is possible to obtain a sufficient result even with a small calculation processing capability.

  Further, according to the course evaluation device 21, the standard trajectory calculation unit 25 can improve the stability of statistical processing by referring to the road map information, and the reliability related to the calculation of the standard trajectory H can be improved. Can be achieved.

  The present invention is not limited to the embodiment described above.

  For example, the traffic information center may have the functions of the movement trajectory information acquisition unit 12, the information storage unit 13, and the standard trajectory calculation unit 14. In this case, the own vehicle M acquires the calculation result of the standard trajectory H by road-to-vehicle communication. Thus, by calculating the standard trajectory H at a traffic information center or the like, it is possible to improve the calculation speed and reduce the calculation load of the in-vehicle device.

  In addition, the course evaluation apparatus 1 does not necessarily need to perform both the acquisition of movement trajectory information by an in-vehicle device such as the radar sensor 3 and the acquisition of movement trajectory information by road-to-vehicle communication, and only one of them may be used. .

DESCRIPTION OF SYMBOLS 1 ... Course evaluation apparatus 2 ... Course evaluation ECU 3 ... Radar sensor 4 ... Camera 5 ... Road-to-vehicle communication part 6 ... Vehicle state detection part 11 ... Course candidate generation part (course candidate generation means) 12, 24 ... Movement track information acquisition part (Movement trajectory information acquisition means) 13 ... information storage section 14, 25 ... standard trajectory calculation section (standard trajectory calculation means) 15 ... similarity calculation section <similarity calculation means> 16 ... course prediction section (course prediction means) 17 ... Interference form determination unit (interference form determination unit) 18 ... Course evaluation unit (Course evaluation unit) 23 ... Road map information storage unit H ... Standard trajectory K ... Moving trajectory M ... Own vehicle R ... Course candidate T ... Determination threshold

Claims (5)

A route candidate generating means for generating a route candidate of the host vehicle;
Movement trajectory information acquisition means for acquiring movement trajectory information related to the movement trajectory of the preceding vehicle on the traveling road of the host vehicle;
Standard trajectory calculating means for calculating a standard trajectory that standardizes the moving trajectory of the preceding vehicle based on the moving trajectory information;
Similarity calculating means for calculating similarity between the course candidate of the host vehicle and the standard locus;
Evaluation means for evaluating a course candidate of the host vehicle based on the similarity;
A course evaluation apparatus comprising:   2. The course evaluation according to claim 1, wherein the movement locus information acquisition unit acquires movement locus information related to a movement locus of a preceding vehicle on a traveling lane of the own vehicle based on road map information around the own vehicle. apparatus. Course prediction means for predicting the course of other vehicles around the host vehicle;
Determination means for determining an interference mode between the course candidate of the host vehicle and the predicted course of the other vehicle;
The route evaluation apparatus according to claim 1, wherein the evaluation unit evaluates a route candidate of the host vehicle based on a determination result of the determination unit.   The determination means causes the other vehicle to interfere with the own vehicle depending on whether an interference parameter obtained by quantifying the form of interference between the route candidate of the own vehicle and the predicted route of the other vehicle exceeds a determination threshold value. 4. The course evaluation according to claim 3, wherein it is determined whether or not the vehicle is in a form or whether or not the own vehicle interferes with the other vehicle, and the determination threshold value is changed based on the similarity. apparatus. A route evaluation method by the route evaluation device according to claim 1,
A route candidate generation step of generating a route candidate of the host vehicle by the route candidate generation means ;
A movement trajectory information acquisition step of acquiring movement trajectory information related to the movement trajectory of the preceding vehicle on the travel road of the host vehicle by the movement trajectory information acquisition means ;
A standard trajectory calculating step of calculating a standard trajectory by standardizing the moving trajectory of the preceding vehicle based on the moving trajectory information by the standard trajectory calculating means ;
A similarity calculation step of calculating similarity between the course candidate of the host vehicle and the standard trajectory by the similarity calculation means ;
An evaluation step of evaluating a course candidate of the host vehicle based on the similarity by the evaluation means ;
A course evaluation method comprising:

Images