JP5487931B2 - Position correction apparatus and method, and driving support apparatus - Google Patents

Description

  TECHNICAL FIELD The present invention relates to a position correction apparatus and method that corrects a travel position of a vehicle and corrects the travel position of the vehicle, and a driving support apparatus that performs drive support control using position information of the vehicle. About.

  As this type of device, there is known a device that detects the current position of a vehicle that is about to enter a curve or an intersection with high accuracy. For example, Patent Document 1 proposes a technique of correcting the position of the vehicle by comparing the shape of the road and the direction of the vehicle so that the position entering the curve is grasped at the initial turning stage.

JP-A-8-292040

  However, there is a possibility that the accuracy of the detected position may be lowered because there is some variation in the timing of starting the turning action at a curve, an intersection or the like depending on various conditions such as the driver and the place. That is, the above-described technique has a technical problem that it is difficult to detect an accurate position of the traveling vehicle.

  The present invention has been made in view of the above-described problems, for example, and an object of the present invention is to provide a position correction device capable of accurately detecting the position of a traveling vehicle.

In order to solve the above-described problem, the position correction apparatus of the present invention learns the driver's deceleration behavior in association with the location information where the vehicle is traveling, and the location information among the learned deceleration behaviors. If the vehicle travels at a point where the location information is the same, an extraction unit that extracts an action with variation within a predetermined range as a feature point if the vehicle travels a point where the location information is the same. When it is determined that the detecting means for detecting, the extractable determining means for determining whether or not the deceleration action and the location information are learned before the feature point can be extracted, and the feature point can be extracted, Compensating means for correcting the position of the vehicle by comparing the detected feature points and the feature points extracted from the learned deceleration behavior.

  The position correction apparatus of the present invention is mounted on a vehicle such as an automobile. During the operation, the driver's deceleration action is first learned in association with the location information where the vehicle is traveling. That is, it is learned (in other words, memorized) at which point during the driving the driver's deceleration action was performed. Here, “deceleration behavior” is a concept that encompasses various operations or behaviors performed by the driver when decelerating, and includes, for example, the operation timing and operation amount of a brake pedal and a steering wheel. The “location information” indicates the travel position of the vehicle, and is information that is embodied to the extent that it can be associated with a deceleration action. The location information includes, for example, the conditions of the road on which the vehicle is traveling (for example, a curve or an intersection) in addition to the coordinates where the vehicle is traveling.

  According to the learning described above, it is possible to know what kind of deceleration action the driver of the vehicle will take when traveling. Specifically, it is possible to learn at which position on the curve, how much the steering wheel and brake are operated, and at which position of the intersection the vehicle is decelerated so as to minimize the vehicle speed. Note that learning of the deceleration action is typically started from a point where the driver is expected to take a deceleration action, such as at the entrance of a curve or before an intersection.

  Subsequently, in the position correction apparatus of the present invention, among the learned deceleration actions, if the location information is the same, an action whose variation is within a predetermined range is extracted as a feature point. Note that “the place information is the same” here typically refers to a state in which the vehicle is traveling in the same place, but also includes cases where the road conditions are the same, such as the degree of curve and the vehicle width. . “Same” means that the road conditions are close enough to be regarded as the same (that is, the effect of the present invention to be described later can be obtained), including the case where the location information is completely the same. ing. That is, even if the location information does not completely match, if the location information is close, a corresponding effect can be obtained. The “predetermined range” is a threshold value for determining that one of a certain deceleration action is small enough to be extracted as a feature point of the driver, and is obtained theoretically, experimentally, or empirically. Is set.

  In the feature point extraction, the learned deceleration action timing, operation amount, or location information may include an error. However, such errors can be reduced by increasing the number of learnings. Therefore, it is desirable that the above-described learning be performed a plurality of times until an effective feature point can be extracted. Further, a plurality of feature points may be extracted.

  After the feature point is extracted, the feature point of the current driver is detected when the vehicle travels at a point having the same location information (that is, a point that has already been learned). When a feature point is detected, the detected feature point and the feature point extracted from the learned deceleration action are compared with each other. That is, the deceleration action that is a feature point currently performed by the driver is compared with the learned deceleration action. More specifically, parameters of feature points such as operation timing and operation amount are compared with each other.

  Here, if the feature point is the same location information, the variation is within a predetermined range, so the current vehicle position is derived very accurately by comparing the detected feature point with the learned feature point. be able to. In other words, it can be determined from the detected feature point parameter value that the vehicle is currently traveling at a point where the learned feature point has the same parameter. Therefore, for example, even if a means for acquiring position information with extremely high accuracy such as high-precision GPS (Global Positioning System) is not provided, the position information of the vehicle is corrected to an accurate position.

  As described above, according to the position correction apparatus of the present invention, it is possible to accurately detect the position of a running vehicle by extracting feature points that are stable parameters if the place is the same. is there.

In one aspect of the vehicle control apparatus of the present invention, the extraction means extracts at least a steering angle by the driver as the feature point. According to this aspect, the present inventor's research tends to be relatively stable. Since the known steering angle (that is, the amount of operation of the steering wheel) is extracted as the feature point, more accurate position correction can be performed. If there is a feature point other than the steering angle, it is used in combination with the steering angle, so that more accurate position information can be corrected.

  In another aspect of the vehicle control apparatus of the present invention, the vehicle control device further includes driver recognition means for recognizing the driver, and the learning means learns a deceleration action for each driver.

  According to this aspect, when a plurality of drivers use the same vehicle, the driver driving the vehicle can be recognized by, for example, face authentication or key registration. And learning of deceleration action is performed for every driver. Therefore, subsequent feature point extraction and position correction are also performed for each driver, and data learned while one driver is driving can be used for position correction while another driver is driving. Absent.

  According to the research of the present inventor, it has been found that the operation timing and the operation amount of the deceleration action, which is a feature point, are different for each driver. Therefore, as described above, if the deceleration action is learned for each driver, the vehicle position can be corrected more suitably.

In order to solve the above-described problem, the position correction method of the present invention is a position correction method by a position correction apparatus including a learning unit, an extraction unit, a detection unit, an extractable determination unit, and a correction unit. A learning step in which a person's deceleration action is learned in association with location information where the vehicle is traveling, and a variation within a predetermined range if the location information is the same among the learned deceleration behaviors by the extraction means An extraction step of extracting the behavior as a feature point, and a detection step of detecting the feature point of the driver when the vehicle travels a point where the location information is the same by the detection means , by the extractable determining means, and extractable determination step of determining whether the reduction action and the location information to the feature point can be extracted have been learned, the correction means , When it is determined that it extracts the feature points, said feature points detected and the learned feature points extracted from the deceleration action by comparing with each other, a correction step of performing position correction of the vehicle Is provided.

  According to the position correction method of the present invention, since the feature point that is a stable parameter is extracted from the learned deceleration action if the place is the same, the traveling vehicle as in the position correction apparatus of the present invention is extracted. It is possible to detect the position accurately.

  Note that the position correction method of the present invention can also adopt various aspects similar to the various aspects of the position correction apparatus of the present invention described above.

  In order to solve the above-described problem, the driving support device of the present invention is based on the position correction device of the present invention described above (including various aspects thereof) and the position corrected by the position correction device. A support control unit that performs driving support control.

  According to the driving support apparatus of the present invention, since the position information of the vehicle is corrected to be extremely accurate by the position correction apparatus of the present invention, for example, driving support such as automatic deceleration control can be suitably executed.

  The effect | action and other gain of this invention are clarified from the form for implementing invention demonstrated below.

It is a block diagram which shows the whole structure of the position correction apparatus which concerns on embodiment. It is a flowchart which shows the flow of learning operation | movement by the position correction apparatus which concerns on embodiment. It is a top view which shows roughly the motion of the vehicle at the time of turning left at an intersection. It is a graph which shows the change of various parameters when vehicles turn left at an intersection. It is a graph which shows the timing of steering operation according to a driver. It is a graph which shows the timing of brake operation according to a driver. It is a flowchart which shows the flow of the position correction operation | movement by the position correction apparatus which concerns on embodiment.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. Hereinafter, a case where the position correction device of the present invention is applied to a driving support control device that performs driving support control of a vehicle will be described.

  First, the configuration of the position correction apparatus according to the present embodiment will be described with reference to FIG. FIG. 1 is a block diagram showing the overall configuration of the position correction apparatus according to this embodiment.

  In FIG. 1, a position correction apparatus 100 according to the present embodiment includes a driver recognition unit 110, a learning unit 120, a location information detection unit 130, a deceleration action detection unit 140, a feature point extraction unit 150, and a position correction unit. 160.

  The driver recognition unit 110 is an example of the “driver recognition means” in the present invention, and recognizes the driver by face authentication, key registration, or the like. By providing the driver recognition unit 110, it is possible to recognize the current driver when multiple people use the same vehicle.

  The learning unit 120 is an example of the “learning unit” of the present invention, and associates the location information detected by the location information detection unit 130 described later and the driver's deceleration behavior detected by the deceleration behavior detection unit 140 with each other. learn. The learning unit 120 stores the learning content in, for example, an HDD (Hard Disc Drive) included in a car navigation system mounted on a vehicle, a server on the center side connected through communication, or the like.

  The location information detection unit 130 includes, for example, a GPS, an in-vehicle camera, and the like, and detects location information including a position where the vehicle travels or road conditions.

  The deceleration action detection unit 140 includes sensors or the like attached to various parts that are operated during the deceleration action by the driver, and detects, for example, the timing and the operation amount of the brake operation and the handle operation.

  The feature point extraction unit 150 is an example of the “extraction unit” in the present invention, and includes, for example, an arithmetic circuit and a memory. The feature point extraction unit 150h detects a feature point that does not change depending on the location from the deceleration action learned by the learning unit 120.

  The position correction unit 160 is an example of the “correction unit” in the present invention, and includes, for example, an arithmetic circuit and a memory. The position correction unit 160 compares the location information detected by the location information detection unit 130 and the driver's deceleration behavior detected by the deceleration behavior detection unit 140 with the feature points extracted by the feature point extraction unit 150. Thus, the position of the vehicle is corrected. This position correction operation will be described in detail later.

  The position information corrected by the position correction apparatus 100 according to the present embodiment is output to the driving support control unit 200 in the driving support apparatus. The driving support control unit 200 is configured to be able to perform control related to driving support of the vehicle using the position information corrected by the position correction device 100.

  Next, operations and effects of the position correction apparatus according to this embodiment will be described with reference to FIGS. 2 to 7 in addition to FIG. Here, of the operation of the position correction apparatus according to the present embodiment, a part characteristic of the effect of the present embodiment will be described in detail, and description of other operations will be omitted as appropriate.

  First, the learning operation of the position correction apparatus according to the present embodiment will be described with reference to FIGS. FIG. 2 is a flowchart showing the flow of the learning operation performed by the position correction apparatus according to this embodiment. FIG. 3 is a top view schematically showing the movement of the vehicle when turning left at the intersection, and FIG. 4 is a graph showing changes in various parameters when the vehicle turns left at the intersection. FIG. 5 is a graph showing the steering operation timing for each driver, and FIG. 6 is a graph showing the braking operation timing for each driver.

  In FIG. 2, during the learning operation of the position correction apparatus 100 according to the present embodiment, the driver recognition unit 110 first recognizes the driver of the vehicle (step S01). That is, it is recognized who is currently driving the vehicle. When the driver is recognized, the information is output to the learning unit 120, and thereafter the recognized learning operation for each driver is performed. Such driver recognition operation is performed, for example, when the engine is started.

  Subsequently, the location information detection unit 130 determines whether or not the traveling vehicle is approaching the deceleration target point (step S02). The “deceleration target point” is a point where it is estimated that the driver will be decelerating like a curve or an intersection, and what kind of place is set as the deceleration target point is set in advance. Has been. In addition, an appropriate threshold value may be set in advance to determine how close the deceleration target point is to “approach”. Below, as shown in FIG. 3, the case where the vehicle 500 is going to turn left at an intersection is demonstrated.

  In FIG. 4, when the vehicle 500 makes a left turn at an intersection, the brake operation by the driver is started before the vehicle 500 reaches the intersection, and the vehicle speed temporarily decreases. Specifically, the brake operation by the driver is performed at a point where the travel distance of the vehicle 500 exceeds 150 m, and the vehicle speed is greatly reduced.

  The driver performs the above-described brake operation and also performs the steering wheel operation in parallel. Specifically, the steering wheel is greatly turned to the left at a point where the vehicle 500 is sufficiently decelerated and the travel distance exceeds 200 m. Thereafter, the brake operation and the steering wheel operation are finished, and the accelerator operation or the like is performed. The vehicle 500 that has finished turning left at the intersection is accelerated.

  Returning to FIG. 2, in the position correction apparatus 100 according to the present embodiment, when it is determined that the vehicle 500 is approaching the deceleration target point (step S02: YES), the learning unit 120 is where the vehicle 500 travels. Information and the deceleration action by the driver are learned in association with each other (step S03). Specifically, the location information detected by the location information detection unit 130 (vehicle travel distance, coordinates, etc.) and the deceleration behavior detected by the deceleration behavior detection unit 140 (for example, the timing of brake operation or steering wheel operation) Are learned in association with each other.

  Here, according to the research of the present inventor, there are a driver's deceleration behavior described above that changes greatly when traveling in the same place and a driver that hardly changes.

  5 and 6, the timing of the steering wheel operation varies depending on the drivers A, B, and C, but becomes a relatively stable value if the driver is the same. For example, in the data shown in FIG. 5, the error is within 1 m in both cases of a right turn and a left turn. On the other hand, the timing of the brake operation varies depending on the drivers A, B, and C, and has large variations even if the drivers are the same.

  The position correction apparatus 100 according to the present embodiment extracts, as feature points, parameters that do not change significantly if the place is the same, such as the above-described steering operation.

  Returning to FIG. 2 again, when extracting feature points, first, the feature point extracting unit 150 determines whether or not the location information and the deceleration action are learned to such an extent that the feature points can be extracted (step S04). . That is, it is determined whether or not a sufficient amount of data has been learned so that the feature points can be determined. Specifically, for example, it is determined that a feature point can be extracted when the average and variance do not change more than a threshold value even when N + 1 data is taken into account.

  When it is determined that the feature point can be extracted (step S04: YES), the feature point extraction unit 150 extracts the feature point (step S05), and the learning operation ends. Note that one or more feature points may be extracted.

  On the other hand, when it is determined that the feature point cannot be extracted (step S04: NO), the location information and the deceleration action in step S03 are learned again, and data is accumulated. That is, the learning of the location information and the deceleration action is repeated until the feature point can be detected.

  Next, the position correction operation of the position correction apparatus according to the present embodiment will be described with reference to FIG. FIG. 7 is a flowchart showing the flow of the position correction operation by the position correction apparatus according to this embodiment. In the following description, the case where the driving of the vehicle is started in the state where the learning operation described above is completed will be described as an example.

  In FIG. 7, during the position correction operation of the position correction apparatus 100 according to the present embodiment, first, the driver recognition unit 110 recognizes the driver of the vehicle 500 (step S11). That is, as in step S01 during the learning operation, it is recognized who is currently driving the vehicle 500. When the driver is recognized, the information is output to the position correction unit 160, and then the recognized position correction operation for each driver is performed.

  Subsequently, the location information detection unit 130 determines whether or not the traveling vehicle 500 is approaching the deceleration target point (step S12). Note that the deceleration target point here is a deceleration target point that has been learned (that is, the driver's feature point has been extracted).

  If it is determined that the vehicle 500 is approaching the deceleration target point (step S12: YES), the deceleration behavior detection unit 140 detects the deceleration behavior extracted as the feature point (step S13). In other words, the deceleration action detection unit 140 here is an example of the “detection means” in the present invention.

  When a feature point is detected, the detected feature point and the feature point extracted from the learned deceleration action are compared with each other. That is, the deceleration action that is a feature point currently performed by the driver is compared with the learned deceleration action.

  Here, since the feature point is a deceleration action that becomes a stable value if the location information is the same, the current position of the vehicle 500 can be derived very accurately by comparing the detected feature point with the learned feature point. Can do. That is, it can be determined from the detected parameter value of the feature point that the vehicle 500 is currently traveling at a point where the learned feature point has the same parameter. Using this, the position correction unit 160 corrects the position of the vehicle 500 (step S14).

  According to the position correction operation described above, the position information of the vehicle 500 can be detected in the cm order. Although accurate position information can be detected by a highly accurate GPS or the like, the cost is very high in this case. On the other hand, the position correction apparatus 100 according to the present embodiment can be realized at a relatively low cost.

  The position information of the vehicle 500 corrected by the position correction apparatus 100 of the present invention is output to the driving support control unit 200. Then, the driving support control unit 200 performs various driving support controls using the corrected position information of the vehicle 500 (step S15). Since the corrected position information of the vehicle 500 has a very accurate value by using the feature points, the driving support control is suitably executed.

  As described above, according to the position correction apparatus according to the present embodiment, if the location is the same, it is possible to accurately detect the position of the traveling vehicle by extracting feature points that are stable parameters. Is possible.

  The present invention is not limited to the above-described embodiments, and can be appropriately changed without departing from the gist or concept of the invention that can be read from the claims and the entire specification, and control of the vehicle accompanying such changes. The apparatus is also included in the technical scope of the present invention.

  DESCRIPTION OF SYMBOLS 100 ... Position correction apparatus, 110 ... Driver recognition part, 120 ... Learning part, 130 ... Location information detection part, 140 ... Deceleration action detection part, 150 ... Feature point detection part, 160 ... Position correction part, 200 ... Driving assistance control 500, vehicle

Claims (5)

Learning means for learning the driver's deceleration behavior in association with the location information where the vehicle is traveling;
An extracting means for extracting, as a feature point, an action whose variation is within a predetermined range if the location information is the same among the learned deceleration actions;
Detecting means for detecting the feature point of the driver when the vehicle travels a point where the location information is the same;
Extractable determination means for determining whether or not the deceleration action and the location information have been learned before the feature points can be extracted;
Correction means for correcting the position of the vehicle by comparing the detected feature points and the feature points extracted from the learned deceleration action when it is determined that the feature points can be extracted; A position correction apparatus comprising:   The position correcting apparatus according to claim 1, wherein the extraction unit extracts at least a steering angle by the driver as the feature point. A driver recognition means for recognizing the driver;
The position correction apparatus according to claim 1, wherein the learning unit learns a deceleration action for each driver. A position correction method by a position correction apparatus comprising learning means, extraction means, detection means, extractable determination means and correction means,
A learning step of learning the driver's deceleration behavior in association with location information where the vehicle is traveling by the learning means ;
An extraction step of extracting, as a feature point, an action in which variation is within a predetermined range if the location information is the same among the learned deceleration actions by the extraction means ;
A detection step of detecting the feature point of the driver when the vehicle travels a point where the location information is the same, by the detection means ;
An extractable determination step for determining whether or not the deceleration action and the location information are learned before the feature point can be extracted by the extractable determination means ;
When it is determined that the feature point can be extracted by the correcting means, the detected feature point and the feature point extracted from the learned deceleration action are compared with each other, thereby correcting the position of the vehicle. A position correction method comprising: a correction step to perform. The position correction device according to claim 1,
A driving support device, comprising: a support control unit that performs driving support control of the vehicle based on the position corrected by the position correction device.

Images