WO2019012838A1

WO2019012838A1 - Obstruction detection device

Info

Publication number: WO2019012838A1
Application number: PCT/JP2018/020863
Authority: WO
Inventors: 誠也高辻; 加藤　智啓; 豊濱本; 田中　秀典; 洋一岩田; 御幸大森; 絵里子山崎; 裕之立花
Original assignee: 株式会社デンソー; トヨタ自動車株式会社
Priority date: 2017-07-11
Filing date: 2018-05-30
Publication date: 2019-01-17
Also published as: JP6844457B2; JP2019020769A

Abstract

This obstruction detection device (1) is mounted in a moving body and detects obstructions on the outside of the moving body. Specifically, the obstruction detection device is equipped with a distance measurement sensor (2) and a type determination unit (104). The distance measurement sensor emits probe waves toward the outside of the moving body and receives reflected waves resulting from the reflection of the probe waves by an obstruction, thereby repeatedly acquiring a detected distance corresponding to the distance to the obstruction while the moving body is moving. The type determination unit determines the type of obstruction on the basis of a frequency distribution for the amount of change in the detected distance.

Description

Obstacle detection device

CROSS-REFERENCE TO RELATED APPLICATIONS

This application is based on Japanese Patent Application No. 2017-135580 filed on Jul. 11, 2017, the contents of which are incorporated herein by reference.

The present disclosure relates to an obstacle detection device.

The apparatus described in Japanese Patent Laid-Open No. 2008-21039 includes a distance measuring sensor and shape estimation means. The distance measuring sensor detects the distance from the vehicle to the obstacle. Specifically, the distance measuring sensor transmits a sound wave or the like in a predetermined direction centered on the vehicle width direction, and receives the reflected wave to perform distance detection on an obstacle present on the side of the vehicle. The detection result of the distance measurement sensor is also referred to as point train data. The shape estimation means estimates the shape of the obstacle by performing curve approximation on point sequence data obtained from the distance measurement sensor.

In this type of device, it may be difficult to obtain point train data sufficiently because it becomes difficult to receive a reflected wave by the distance measurement sensor for a part of obstacles. Such a portion of the obstacle is typically the end of the obstacle in the direction of movement of the ranging sensor. In particular, when the obstacle is a parked vehicle, the corner portion of the parked vehicle has a complicated shape.

For example, when the obstacle is an object having a relatively simple shape such as a wall or a prism, the end position can be well estimated based on the acquired point sequence data. However, when the obstacle is a parked vehicle, the end position estimated based on the acquired point sequence data may largely deviate from the actual end position. Therefore, the type of obstacle is important information when estimating the end position of the obstacle. The present disclosure has been made in view of the circumstances and the like exemplified above.

According to one aspect of the present disclosure,
An obstacle detection device configured to be mounted on a mobile body and configured to detect an obstacle existing outside the mobile body,
The detection distance corresponding to the distance to the obstacle is transmitted along with the movement of the mobile body by transmitting the search wave toward the outside of the mobile body and receiving the reflected wave of the search wave due to the obstacle. A distance measuring unit provided to repeatedly acquire,
A frequency distribution acquisition unit provided to acquire the change amount of the detection distance for each acquisition of the detection distance in the distance measuring unit and to acquire the acquired frequency distribution of the change amount;
A type determination unit provided to determine the type of the obstacle based on the frequency distribution acquired by the frequency distribution acquisition unit;
Equipped with

In such a configuration, the ranging unit transmits the search wave toward the outside of the moving body while the moving body is moving, and receives the reflected wave by the obstacle of the search wave. Thus, the distance measuring unit repeatedly acquires the detection distance corresponding to the distance to the obstacle as the moving object moves. The frequency distribution acquisition unit acquires the frequency distribution of the change amount of the detection distance. That is, the frequency distribution acquisition unit acquires the amount of change in the detection distance for each acquisition of the detection distance in the distance measurement unit. Further, the frequency distribution acquisition unit acquires the frequency distribution of the amount of change based on the acquired amount of change.

The change amount of the detection distance may be acquired based on the detection distance acquired this time and the detection distance acquired at a point before this time. Specifically, for example, the plurality of acquired detection distances are indicated by a point sequence in a coordinate system in which the vertical axis is the detection distance and the horizontal axis is the position of the distance measurement sensor. In this case, the change amount of the detection distance can be typically indicated by the slope of a line segment connecting two adjacent points in the point sequence.

As a result of earnest research, the inventor focused on the frequency distribution of the change amount of the detection distance. That is, when frequency distribution of the amount of change of detection distance was acquired about obstacles of different types, it was found that the shape of frequency distribution was different for each type of obstacle. Therefore, the type determination unit determines the type of obstacle based on the frequency distribution acquired by the frequency distribution acquisition unit. Specifically, for example, the type determination unit determines the first condition based on the first number, which is the frequency of the small change region where the absolute value of the change amount is small, and the frequency of the large change region where the absolute value of the change amount is large. Based on the second condition based on a certain second frequency and the third condition based on the relationship between the first number and the second frequency, it is determined whether the obstacle is a vehicle.

As described above, in such a configuration, the type of obstacle can be favorably determined based on the frequency distribution of the change amount of the detection distance. Therefore, according to such a configuration, the end position of the obstacle can be estimated better than before based on the determined type. That is, for example, the end position can be favorably corrected based on the type determined by the type determination unit.

In addition, the parenthesized reference symbol attached to each element indicates an example of the correspondence between the element and the specific means described in the embodiments to be described later.

It is a block diagram showing a schematic structure of a parking assistance device. It is a conceptual diagram which shows the operation | movement outline | summary of the parking assistance apparatus shown by FIG. It is a conceptual diagram which shows the operation | movement outline | summary of the parking assistance apparatus shown by FIG. It is a conceptual diagram which shows the operation | movement outline | summary of the parking assistance apparatus shown by FIG. It is a conceptual diagram which shows the operation | movement outline | summary of the parking assistance apparatus shown by FIG. It is a histogram which shows the outline | summary of frequency distribution acquired by the ranging information processing part shown by FIG. It is a conceptual diagram which shows the operation | movement outline | summary of the parking assistance apparatus shown by FIG. It is a conceptual diagram which shows the operation | movement outline | summary of the parking assistance apparatus shown by FIG. It is a conceptual diagram which shows the operation | movement outline | summary of the parking assistance apparatus shown by FIG. It is a conceptual diagram which shows the operation | movement outline | summary of the parking assistance apparatus shown by FIG. It is a conceptual diagram which shows the operation | movement outline | summary of the parking assistance apparatus shown by FIG. It is a conceptual diagram which shows the operation | movement outline | summary of the parking assistance apparatus shown by FIG. It is a conceptual diagram which shows the operation | movement outline | summary of the parking assistance apparatus shown by FIG. It is a flowchart for demonstrating the operation example of the parking assistance apparatus shown by FIG. It is a flowchart for demonstrating the operation example of the parking assistance apparatus shown by FIG. It is a flowchart for demonstrating the operation example of the parking assistance apparatus shown by FIG.

Hereinafter, embodiments will be described based on the drawings.

(Constitution)
With reference to FIGS. 1 and 2, the parking assistance device 1 corresponding to an embodiment of the obstacle detection device of the present disclosure is mounted on a vehicle V, which is a mobile body, to provide parking assistance including an obstacle detection operation. It is configured to perform an action. That is, the parking assistance device 1 detects the obstacle B existing outside the vehicle V, detects the parking space PS based on the detection result of the obstacle B, and detects the parking space PS based on the detection result of the parking space PS. It is designed to support parking in the parking space PS.

Specifically, the parking assistance device 1 includes a distance measuring unit 2, a traveling state acquisition unit 3, a camera 4, an image display device 5, an audio output device 6, an operation unit 7, and a torque control ECU 8. A braking control ECU 9 and a parking assist ECU 10 are provided. ECU is an abbreviation of Electronic Control Unit.

The distance measuring unit 2 is provided to repeatedly acquire a detection distance corresponding to the distance to the obstacle B as the vehicle V on which the parking assistance device 1 is mounted moves. In addition, the distance measuring unit 2 is electrically connected to the parking assist ECU 10 so as to transmit the acquired detected distance to the parking assist ECU 10.

Specifically, the distance measuring unit 2 includes a distance measuring sensor 20. The distance measurement sensor 20 is provided to transmit a survey wave toward the outside of a vehicle V equipped with the parking assistance device 1 and to receive a reflected wave by an obstacle B of the survey wave. Specifically, the distance measurement sensor 20 is configured to repeat the transmission of the search wave and the reception of the reflected wave at a predetermined detection cycle. The predetermined detection cycle is, for example, 30 milliseconds. For example, sound waves, radio waves, or light waves may be used as the search waves. Specifically, for example, an ultrasonic wave, a millimeter wave, or a laser can be suitably used as a search wave. When the search wave is an ultrasonic wave, the distance measurement sensor 20 may also be referred to as an “ultrasound sensor”.

The distance measuring sensor 20 is fixed at a predetermined position on the vehicle V on which the parking assistance device 1 is mounted so as to face in a predetermined direction. Typically, as shown in FIG. 2, the distance measuring sensor 20 is provided on both sides of the vehicle V so as to transmit a survey wave in the vehicle width direction. The distance measuring sensor 20 has a predetermined detectable range CR which extends in a triangular pyramid shape to a predetermined distance and then extends along the vehicle width direction. That is, the distance measuring sensor 20 is configured to output an electric signal corresponding to the detected distance by receiving a reflected wave from the obstacle B located in the detectable range CR.

The traveling state acquisition unit 3 is provided to acquire the traveling state of the vehicle V on which the parking assistance device 1 is mounted. Hereinafter, the traveling state of the vehicle V will be simply referred to as “the traveling state”. The traveling state includes, for example, the wheel speeds of the wheels of the vehicle V, the traveling speed of the vehicle V, the steering angle of the vehicle V, the yaw rate acting on the vehicle V, and the like. That is, the traveling state acquisition unit 3 includes a wheel speed sensor, a steering angle sensor, a yaw rate sensor, an acceleration sensor, and the like. The traveling state acquisition unit 3 is electrically connected to the parking assistance ECU 10 so as to transmit the acquired traveling state to the parking assistance ECU 10. In addition, the traveling speed of the vehicle V which mounts the parking assistance apparatus 1 is only hereafter called "vehicle speed."

The camera 4 is mounted on the vehicle V so as to capture an image around the vehicle V on which the parking assistance device 1 is mounted. The image display device 5, the voice output device 6, and the operation unit 7 are disposed in the vehicle compartment of the vehicle V on which the parking assistance device 1 is mounted. The image display device 5 is a liquid crystal display panel, and is provided to perform image display accompanying the parking assistance operation under the control of the parking assistance ECU 10. The voice output device 6 is a speaker device, and is provided so as to perform voice output according to the parking assistance operation under the control of the parking assistance ECU 10. The operation unit 7 is at least one of a touch panel superimposed on the image display device 5, an operation switch disposed around the image display device 5, an audio microphone disposed near the driver's seat of the vehicle V, and the like. , And is provided to receive user input accompanying the parking assistance operation.

The torque control ECU 8 and the braking control ECU 9 are electrically connected to the parking assist ECU 10. The torque control ECU 8 controls the engine output of the vehicle V equipped with the parking assistance device 1 at the time of parking assistance operation by controlling the operation of an injector or the like (not shown) based on a signal received from the parking assistance ECU 10 It is provided. The braking control ECU 9 is provided to control the braking force of the vehicle V equipped with the parking assistance device 1 at the time of parking assistance operation by driving a brake actuator etc. (not shown) based on a signal received from the parking assistance ECU 10 It is done.

The parking assistance ECU 10 is configured to control the parking assistance operation based on inputs from the distance measuring unit 2, the traveling state acquisition unit 3, the camera 4, and the operation unit 7. Specifically, the parking assistance ECU 10 generates a drive control signal for controlling the operation of the image display device 5, the audio output device 6, the torque control ECU 8, and the braking control ECU 9 based on the various inputs described above, The generated drive control signal is output to each part.

In the present embodiment, the parking assist ECU 10 is a so-called microcomputer, and includes a CPU, a ROM, a RAM, a non-volatile RAM, an input / output interface, and the like (not shown). That is, the parking assist ECU 10 is configured to be able to realize various control operations by the CPU reading and executing the program from the ROM or the nonvolatile RAM. In the ROM or nonvolatile RAM, various data used when executing a program are stored in advance. Various data stored in advance in the ROM or nonvolatile RAM are, for example, initial values, look-up tables, maps, and the like. The RAM is configured to temporarily store data such as calculation results when the CPU executes a program. The non-volatile RAM is, for example, a flash ROM.

As the functional configuration realized by the CPU, the parking assistance ECU 10 includes a distance measurement information processing unit 101, a reflection position acquisition unit 102, an end position estimation unit 103, a type determination unit 104, and an end position correction unit. 105, a parking space detection unit 106, and a traveling condition calculation unit 107.

The distance measurement information processing unit 101 is provided to execute predetermined processing on a plurality of detection distances acquired from the distance measurement unit 2 based on the traveling state acquired from the traveling state acquisition unit 3. Specifically, the distance measurement information processing unit 101 is configured to perform curve approximation on a plurality of detection distances acquired from the distance measurement unit 2. Further, the distance measurement information processing unit 101 is configured to acquire the change amount of the detection distance for each of the plurality of detection distances after the curve approximation. Furthermore, the ranging information processing unit 101 is configured to acquire the frequency distribution of the acquired amount of change.

The reflection position acquisition unit 102 is provided to acquire an estimated reflection position using the principle of triangulation based on the detection distance after processing by the distance measurement information processing unit 101. The estimated reflection position is a position estimated as a specific position on the outer surface of the obstacle B, which is a starting point of a reflected wave corresponding to the detection distance acquired by the distance measuring unit 2 at a certain time. The reflection position acquisition unit 102 is configured to sequentially acquire estimated reflection positions corresponding to each of a plurality of detection distances based on the acquired detection distance and the movement state.

The end position estimation unit 103 detects the end of the obstacle B based on the estimated reflection position located at the end in the moving direction of the vehicle V among the plurality of estimated reflection positions acquired by the reflection position acquisition unit 102. It is provided to estimate the position. Hereinafter, the moving direction of the vehicle V on which the parking assistance device 1 is mounted will be referred to as “vehicle traveling direction”.

The type determination unit 104 is provided to determine the type of the obstacle B based on the frequency distribution of the change amount of the detection distance acquired by the distance measurement information processing unit 101. Specifically, the type determination unit 104 is configured to determine whether the obstacle B is a vehicle based on the feature of the frequency distribution. Furthermore, the type determination unit 104 determines whether or not the obstacle B is a vehicle based on the distribution state of a plurality of estimated reflection positions acquired by the reflection position acquisition unit 102 in correspondence with one obstacle B. It is supposed to The details of the type determination by the type determination unit 104 will be described in more detail in the operation outline description to be described later.

The end position correction unit 105 is provided to correct the end position of the obstacle B based on the type of the obstacle B determined by the type determination unit 104. Specifically, when the type determined by the type determination unit 104 is a vehicle, the end position correction unit 105 corrects the end position of the obstacle B based on the vehicle dimensions defined in the vehicle standard. It is supposed to The details of the correction by the end position correction unit 105 will also be described in more detail in the outline of operation described later.

The parking space detection unit 106 is provided to detect the parking space PS based on the end position of the obstacle B after correction by the end position correction unit 105. The traveling condition calculation unit 107 is provided to calculate traveling conditions for guiding the vehicle V to the parking space PS based on the parking space PS detected by the parking space detection unit 106. The travel conditions at least include a travel path from the current position of the vehicle V to the parking space PS.

(Operation summary)
Hereinafter, an operation outline of the parking assistance device 1 of the present embodiment will be described with reference to FIGS. 3A to 10B in addition to FIGS. 1 and 2. In FIG. 3A etc., in order to simplify the description, it is assumed that the vehicle V is traveling in the X direction while searching for the parking space PS. In this case, X indicates the sensor position, that is, the position of the distance measuring sensor 20. Also, Y indicates a detection distance, that is, a distance from the distance measurement sensor 20.

As shown in FIG. 3A, while searching for the parking space PS, the parking assistance ECU 10 acquires a plurality of detection distances as the vehicle V travels. The position of the distance measurement sensor 20 at the time when the distance measurement unit 2 acquires a certain detection distance can be specified based on the traveling state acquired from the traveling state acquisition unit 3. Therefore, a certain detection distance acquired by the distance measurement unit 2 is indicated as one point in a two-dimensional coordinate by the sensor position X and the detection distance Y. This reversal is referred to as a detection position PD. Therefore, as shown in FIG. 3A, the ranging information processing unit 101 acquires a point sequence at a plurality of detection positions PD while searching for the parking space PS.

Further, the distance measurement information processing unit 101 performs curve approximation on the point sequence in the above two-dimensional coordinates. The approximate curve PL in FIG. 3B is a curve approximation of a series of points at a plurality of detection positions PD in FIG. 3A. Each of the point sequences according to the plurality of detection positions PD in FIG. 3A is corrected to be located on the approximate curve PL in FIG. 3B. Thus, the noise component in the detection distance is favorably reduced by correcting each of the plurality of acquired detection positions PD by curve approximation based on the acquisition history of the plurality of detection distances.

FIG. 4 shows acquisition results of point sequences at a plurality of detection positions PD when the obstacle B is a prism. In FIG. 4, each of the plurality of detection positions PD illustrated is after correction by curve approximation.

In addition, the ranging information processing unit 101 acquires the change amount of the detection distance for each acquisition of the detection distance in the ranging unit 2. That is, the distance measurement information processing unit 101 acquires the change amount of the detection distance based on the detection distance acquired this time and the detection distance acquired at a point before this time.

The “amount of change in detection distance” may be indicated by the inclination of a line segment connecting two adjacent points in a point sequence at a plurality of detection positions PD. Therefore, in the present embodiment, the ranging information processing unit 101 acquires, as a variation, the inclination of a line segment connecting two adjacent detection positions PD in a point sequence based on a plurality of detection positions PD after curve approximation. Do. Specifically, assuming that N is an integer of 1 or more and the Nth detection position PD is a detection position PD (N), the change amount T (N) in the detection position PD (N) is the detection position PD (N). It is an inclination of a line segment connecting -1) and the detection position PD (N).

Further, the distance measurement information processing unit 101 acquires a frequency distribution of the change amount of the detection distance. That is, the ranging information processing unit 101 in the present embodiment functions as a frequency distribution acquisition unit in the present disclosure. In the present embodiment, the ranging information processing unit 101 performs curve approximation and updates the frequency distribution each time the detection distance is acquired once. Details of the frequency distribution will be described later.

The reflection position acquisition unit 102 acquires the estimated reflection position PR by applying the principle of triangulation to the acquisition results of the plurality of detection positions PD. The acquisition of the plurality of detection positions PD, the correction of these corrections by curve approximation, and the acquisition of the estimated reflection position PR based on triangulation as described above are already known at the time of filing of the present application. Detailed description will be omitted. As necessary, refer to US Pat. No. 7,843,767, JP-A-2007-71536, JP-A-2012-146024, JP-A-2012-146025, and the like.

The end position estimation unit 103 determines, among the plurality of estimated reflection positions PR acquired by the reflection position acquisition unit 102, positions at both ends in the X direction as the estimated end positions P1 and P2, respectively. The estimated end positions P1 and P2 are positions estimated based on the plurality of estimated reflection positions PR as the end positions of the obstacle B. Specifically, the estimated end position P1 is the one on the near side in the vehicle traveling direction among the plurality of estimated reflection positions PR corresponding to one obstacle B. The estimated end position P2 is on the vehicle traveling direction side among a plurality of estimated reflection positions PR corresponding to one obstacle B.

The type determination unit 104 determines the type of the obstacle B based on the frequency distribution of the change amount of the detection distance acquired by the distance measurement information processing unit 101. FIG. 5 shows, as an example, an example of a frequency distribution in the case where obstacle B is a parked vehicle parked in such a manner that the front face follows the direction of travel of the vehicle as shown in FIG.

In FIG. 5, the horizontal axis indicates the amount of change. “D7”, “D6”, “D1”, “C”, “U1”, “U2”, “U7” on the horizontal axis are classified by dividing the amount of change within a predetermined range It is. “C” is the case where the slope is substantially zero, that is, the amount of change T (N) is in the range of −α to + α. α is a predetermined small positive value. In other words, “C” is a case where the detection distance corresponding to the detection position PD (N−1) and the detection distance corresponding to the detection position PD (N) are substantially equal. “U1”, “U2”,... Are cases where the slope is a positive value, and “D1”, “D2”, etc. are cases where the slope is a negative value. The larger the subscript number, the larger the absolute value of the slope. Specifically, “U1” is a positive value with a small slope, and the amount of change T (N) is in the range of + α to + 3α. “U2” is a positive value with a small slope, and the amount of change T (N) is in the range of + 3α to + 5α. In FIG. 5, the vertical axis indicates the frequency.

6 to 8 show the relationship between the shape of the obstacle B and the shape of the frequency distribution. As shown in FIG. 6, when the obstacle B is a prism, since the surface facing the traveling path of the vehicle V is flat, the frequency concentrates on a specific class near “C”. On the other hand, as shown in FIG. 7, when the obstacle B is a cylinder, the curvature of the surface facing the traveling path of the vehicle V is constant, so the frequency is almost uniform in a wide class. Distributed in Further, as shown in FIG. 8, when the obstacle B is a vehicle, the shape of the surface facing the traveling path of the vehicle V is configured by a portion close to a plane and a curved portion, and the frequency distribution The shape of is a “convex” shape centered on the class near “C”.

As described above, the inventor of the present invention has found that the shape of the frequency distribution differs depending on the type of the obstacle B, that is, the feature of the shape of the obstacle B appears in the shape feature of the frequency distribution. I found it. Therefore, the type determination unit 104 determines whether the obstacle B is a vehicle based on the shape feature of the frequency distribution.

Specifically, in the present embodiment, the type determination unit 104 calculates the absolute value of the first condition based on the first number Q1 which is the sum of the frequencies of the small change region R1 having a small absolute value of the change amount, and the absolute value of the change amount. Based on the second condition based on the second frequency Q2 which is the sum of the frequencies of the large change region R2 having a large value, and the third condition based on the relationship between the first number Q1 and the second frequency Q2, the obstacle B Determines whether the vehicle is a vehicle.

The first condition is whether or not the first number Q1 is equal to or greater than a predetermined number Qth1, that is, whether a predetermined number Qth1 or more of detection positions PD included in the small change region R1 are present. If the first number Q1 is equal to or greater than the predetermined number Qth1, the possibility that the type of the obstacle B is a vehicle is high. The second condition is whether or not the second frequency Q2 is a predetermined number Qth2 or more, that is, whether a predetermined number Qth2 or more of detection positions PD included in the large change region R2 exist. When the second frequency Q2 is equal to or greater than the predetermined number Qth2, the possibility that the type of the obstacle B is a vehicle is high. The third condition is whether the first number Q1 is larger than the second frequency Q2, that is, the detected position PD included in the small change region R1 is larger than the detected position PD included in the large change region R2 It is. When the first number Q1 is larger than the second frequency Q2, the type of the obstacle B is more likely to be a vehicle.

9A and 9B show the case where the vehicle speed, that is, the moving speed of the distance measurement sensor 20 is a low speed of, for example, about 1 km / h. FIGS. 10A and 10B show the case where the vehicle speed, that is, the moving speed of the distance measurement sensor 20 is higher than that of FIGS. 9A and 9B, for example, about 15 km / h. In FIG. 9A and FIG. 10A, the obstacle B is a parked vehicle parked in such a manner that the front surface is along the vehicle traveling direction. In FIGS. 9B and 10B, the obstacle B is a cylinder. As apparent from FIGS. 9A, 9B, 10A, and 10B, when the vehicle speed, that is, the traveling speed of the vehicle V equipped with the parking assistance device 1 becomes high, the feature caused by the shape of the obstacle B in the frequency distribution Is less likely to appear.

However, as shown in FIG. 10A, when the obstacle B is a vehicle, the frequency distribution becomes wide, and the estimated reflection positions PR are widely distributed in the X direction. On the other hand, as shown in FIG. 10B, when the obstacle B is a cylinder, the frequency distribution becomes wide and the distribution of the estimated reflection position PR in the X direction becomes narrow. Therefore, the type determination unit 104 determines whether the obstacle B is a vehicle based on the distribution state of the plurality of estimated reflection positions PR acquired by the reflection position acquisition unit 102 in correspondence with one obstacle B. judge. Specifically, for example, when the frequency distribution is wide and the density of the plurality of estimated reflection positions PR in the X direction is low for one obstacle B, the type of the obstacle B is likely to be a vehicle . This condition is hereinafter referred to as the "fourth condition".

The end position correction unit 105 corrects the end position of the obstacle B based on the type determined by the type determination unit 104. In the present embodiment, when the determination result is a vehicle, the end position correction unit 105 determines the estimated end positions P1 and P2 as the vehicle standard of “ordinary car” in the Japanese Road Transport Vehicle Act. Make corrections based on the size of the vehicle. Specifically, when the determination result is a vehicle, the end position correction unit 105 sets X as a reference dimension, with a vehicle width of 1,800 mm and a total vehicle length of 4,900 mm, which is a standard vehicle size of an ordinary automobile. The estimated end positions P1 and P2 are corrected so that the distance between P1 and P2 in the direction becomes the above reference dimension.

For example, in the case of FIG. 8, for the estimated end positions P1 and P2 before correction shown in the figure, the distance between P1 and P2 in the X direction is less than the reference dimension described above. Therefore, in this case, for example, the estimated end position P1 before correction shown in the figure is a distance obtained by multiplying the difference between the distance between P1 and P2 and the reference dimension by 0.5, in the direction opposite to the X direction Can be moved to Similarly, the estimated end position P2 before correction shown in the figure can be moved in the X direction by a distance obtained by multiplying the difference between the distance between P1 and P2 and the reference dimension by 0.5. The correction amount of each of P1 and P2 may be obtained by a calculation method different from the above example. Also, the distance between P1 and P2 may be determined by XY coordinates. The movement of each of P1 and P2 due to the correction is not limited to the X direction, but may be performed in a specific direction of the XY coordinates, for example, a movement in a linear direction connecting P1-P2. Good.

The parking space detection unit 106 detects the parking space PS based on the end position of the obstacle B after correction by the end position correction unit 105. The traveling condition calculation unit 107 calculates traveling conditions for guiding the vehicle V to the parking space PS based on the parking space PS detected by the parking space detection unit 106. That is, the determination result of the type of the obstacle B by the type determination unit 104 is the detection of the parking space PS between a plurality of obstacles B and the collision of the obstacle V with the vehicle V while avoiding the collision with the vehicle B. It is used to calculate the traveling conditions for reaching. The traveling conditions include, for example, a traveling route and a traveling speed.

(Operation example)
Hereinafter, a specific operation example according to the configuration of the present embodiment will be described using the flowcharts of FIGS. In the following description in the drawings and specification, "step" is simply abbreviated as "S". In the following description of the flowchart, the CPU, the ROM, the RAM, and the non-volatile RAM of the parking assistance ECU 10 will be simply referred to as “CPU”, “ROM”, “RAM”, and “non-volatile RAM”.

The parking support routine shown in FIG. 11 is initially activated by a predetermined operation on the operation unit 7 by the user such as the driver of the vehicle V, and thereafter started at predetermined time intervals until the parking support operation is completed. . The predetermined time may be set to, for example, 30 milliseconds, which is the same as the acquisition interval of the detection distance in the distance measuring unit 2. When the parking support routine is started, the CPU first determines in S1101 whether a flag F indicating whether a parking support operation is in progress has been reset.

Immediately after the parking assistance routine is started, the flag F is reset (ie, F = 0). Therefore, the determination in S1101 is "YES". Therefore, the CPU advances the process to S1102 to S1106. At S1102, the CPU sets a flag F (ie, F = 1).

On the other hand, if the flag F is already set at the determination time of S1101 (ie, S1101 = NO), the CPU skips the processing of S1102 to S1105 and advances the processing to S1106. That is, the flag F is a flag for switching whether to skip the processing of S1103 to S1105.

At S1103, the CPU determines a parking method. That is, the CPU selects the parking method of the vehicle V in the current parking assistance operation from parallel parking and parallel parking. The choice of parking method may be made based on user input, for example. Alternatively, the selection of the parking method may be determined by the CPU based on the output of the distance measuring unit 2 and / or the camera 4 or the like.

At S1104, the CPU detects the parking space PS based on the distance measurement result and the like by the distance measurement unit 2. At S1105, the CPU calculates traveling conditions for parking the vehicle V in the parking space PS, starting from the current position of the vehicle V, based on the distance measurement result by the distance measuring unit 2 and the like. In the traveling condition calculation, for example, when the obstacle B located along the traveling route is a vehicle, the traveling route may be set to be larger than when the obstacle B is a non-vehicle. For example, in the vicinity of the obstacle B which is a vehicle, the vehicle speed may be reduced more than when traveling in the vicinity of a non-vehicle.

At S1106, the CPU determines whether or not the parking assistance operation under the current traveling conditions can be continued. Specifically, for example, when the obstacle B is present at a position where there is a possibility of collision when the vehicle V travels as it is at the determination time of S1106, the determination of S1106 is "NO". Typically, when the parking space PS is detected in S1104, the obstacle B, which is a parked vehicle adjacent to the parking space PS, starts in the parking assistance operation, but the determination in S1106 is "NO". This is an example of

If the parking assistance operation under the current traveling conditions can be continued (that is, S1106 = YES), the CPU advances the process to S1107. At S1107, the CPU permits traveling for a predetermined time under the current traveling conditions. Thus, the vehicle V travels for a predetermined time. The predetermined time may be set to, for example, 30 milliseconds, which is the same as the acquisition interval of the detection distance in the distance measuring unit 2. After that, the CPU advances the process to S1109.

On the other hand, when the parking assistance operation under the current traveling conditions can not be continued (ie, S1106 = NO), the CPU advances the process to S1108. At S1108, the CPU resets flag F (ie, F = 0). At this time, the vehicle V is temporarily stopped. After that, the CPU advances the process to S1109.

At S1109, the CPU determines whether the parking assistance operation has ended. The end of the parking assistance operation is, for example, when the vehicle V has reached a predetermined position in the parking space PS, and the parking of the vehicle V in the parking space PS is completed. Alternatively, the end of the parking assistance operation is, for example, the case where the parking assistance operation is forcibly ended based on a user input or the like before the vehicle V reaches a predetermined position in the parking space PS.

When the parking assistance operation is completed (that is, S1109 = YES), the CPU executes the processing for the termination of the parking assistance operation in S1110, and then completely terminates the processing according to this routine. The end process in S1110 includes the end notification to the user by the image display device 5 and / or the audio output device 6, and the reset of the flag F.

On the other hand, when the parking assistance operation has not ended (that is, S1109 = NO), the CPU skips the processing of S1110 and once ends this routine. In this case, the CPU starts this routine again at the next start timing.

While the parking assistance operation can be continued under the current traveling conditions, it is not necessary to redetect the parking space PS and reset the traveling conditions. Therefore, while the parking assistance operation is being performed, and the determination of S1106 is YES, the flag F is set, so the processing of S1103 to S1105 is skipped.

On the other hand, when the parking assistance operation under the current traveling condition can not be continued for some reason, it is necessary to redetect the parking space PS and reset the traveling condition. Therefore, in this case, the determination in S1106 is NO, and the flag F is reset. As a result, at the next activation timing, the determination in S1101 becomes YES, and the processing in S1103 to S1105 is executed. As a result, redetection of the parking space PS and resetting of the traveling conditions are performed.

FIG. 12 shows an example of an obstacle detection routine for detecting an obstacle B present around the vehicle V. The detection result of the obstacle B by this routine is used when executing the processing of S1103, S1104, S1105, and S1106.

The CPU repeatedly executes the obstacle detection routine shown in FIG. 12 at a predetermined activation timing. The activation interval of this routine may be set to, for example, 30 milliseconds, which is the same as the acquisition interval of the detection distance in the distance measuring unit 2.

When this routine is started, first, in S1201, the CPU acquires a detection distance based on the output of the distance measurement sensor 20. Next, in S1202, the CPU executes curve approximation on the detection distance obtained this time and before this time. That is, the CPU calculates the approximate curve PL shown in FIG. 3B based on the plurality of detected positions PD shown in FIG. 3A. Subsequently, the CPU corrects the detection position PD based on the calculated approximate curve PL to acquire the corrected detection distance. After that, the CPU advances the process to S1204 and S1205.

In S1204, the CPU acquires the amount of change T (N) at the detection position PD (N) by the slope of the line connecting the detection position PD (N-1) and the detection position PD (N). In S1205, the CPU generates a distribution of change amounts T using the change amounts T (N) acquired this time. That is, the CPU updates the distribution of the change amount T using the change amount T (N) acquired this time. Thereafter, the CPU advances the process to S1206 to S1211.

N is an integer greater than or equal to 1 and is the number of times of detection distance detection for one obstacle B. That is, N corresponds to the number of activations of this routine. In other words, the start of this routine is the Nth start. Therefore, in the case of N = 1, the processing of S1204 and S1205 is skipped because the previously acquired detected position PD (N-1) does not exist.

At S1206, the CPU determines a first condition. That is, the CPU determines whether the first number Q1 is equal to or greater than a predetermined number Qth1. At S1207, the CPU determines a second condition. That is, the CPU determines whether the second frequency Q2 is equal to or greater than a predetermined number Qth2. At S1208, the CPU determines the third condition. That is, the CPU determines whether the first number Q1 is larger than the second frequency Q2.

In S1209, the CPU acquires the estimated reflection position PR by applying the principle of triangulation to the plurality of corrected detection positions PD (N). Subsequently, in S1210, the CPU determines the distribution state of the estimated reflection position PR along the X direction. That is, the CPU calculates the density K in the X direction of the plurality of estimated reflection positions PR acquired corresponding to one obstacle B. Specifically, for example, for a plurality of estimated reflection positions PR acquired corresponding to one obstacle B, assuming that the distribution width in the X direction is XD and the number is M, the density K = M / XD. The unit of K is number / meter. The CPU also determines whether the value of K is smaller than a predetermined value Kth.

In S1211, the CPU determines whether or not the obstacle B currently acquiring the detection distance is a vehicle, based on the determination results of the conditions in S1206 to S1208 and S1210. Specifically, the CPU first performs the first condition that the first number Q1 is the predetermined number Qth1 or more, the second condition the second frequency Q2 is the predetermined number Qth2 or more, and the first number Q1 is the second It is determined whether all the third conditions larger than the frequency Q2 are satisfied. Next, the CPU determines whether a fourth condition where the density K is smaller than the predetermined value Kth is satisfied. If all of the first to fourth conditions are satisfied, the CPU makes the determination of S 1211 “YES”.

Specifically, among classes D1 to D7, the one having a frequency that is not 0 and the maximum number of subscripts is DI. I is an integer of 1 or more and 7 or less. Further, among classes U1 to U7, UJ having a frequency that is not 0 and the number of subscripts is largest is UJ. J is an integer of 1 or more and 7 or less. Let W = I + J + 1. At this time, when W is larger than the predetermined value Wth and the value of K is smaller than the predetermined value Kth, the CPU determines that the fourth condition is satisfied.

If the determination in S1211 is "YES", the CPU advances the process to S1212. At S1212, the CPU sets the type determination result of the obstacle B as "vehicle". On the other hand, if the determination in S1211 is "NO", the CPU advances the process to S1213. At S1213, the CPU sets the type determination result of the obstacle B as "non-vehicle". As described above, after the type determination result of the obstacle B is set based on the determination result of S1211, the CPU temporarily ends this routine.

FIG. 13 shows an example of an end position estimation routine for estimating the end position of the obstacle B. The CPU repeatedly executes the end position estimation routine shown in FIG. 13 at a predetermined activation timing. The activation interval of this routine may be set to, for example, 30 milliseconds, which is the same as the acquisition interval of the detection distance in the distance measuring unit 2.

When this routine is activated, the CPU first determines in S1301 whether an obstacle boundary has been detected. The obstacle boundary is an end point of a series of detection positions PD corresponding to one obstacle B. The obstacle boundary can be detected, for example, when the reflected wave intensity is less than the threshold intensity continuously for a predetermined time. Alternatively, the obstacle boundary may be detected based on, for example, the arrangement state of the detection position PD before or after the curve approximation.

If an obstacle boundary is not detected, acquisition of a series of detection positions PD corresponding to one obstacle B has not been completed. Therefore, in this case, the determination in S1301 is NO, and the CPU skips all the processes after S1302 and once ends this routine.

When an obstacle boundary is detected, acquisition of a series of detection positions PD corresponding to one obstacle B is completed, and acquisition of estimated edge positions P1 and P2 can be performed for the obstacle B. Therefore, in this case, the determination in S1301 is YES, and the CPU advances the process to S1302 and subsequent steps.

At S1302, the CPU obtains estimated end positions P1 and P2. Next, in S1303, the CPU determines whether the type determination result of the obstacle B is a vehicle. If the type determination result of the obstacle B is a vehicle (that is, S1303 = YES), the CPU advances the process to S1304 and acquires a correction value for a vehicle. On the other hand, when the type determination result of the obstacle B is a non-vehicle (that is, S1303 = NO), the CPU advances the process to S1305 and acquires a non-vehicle correction value. The non-vehicle correction value is smaller than the vehicle correction value.

After the process of S1304 or S1305 is executed according to the determination result in S1303, the CPU advances the process to S1306 and S1307. In S1306, the CPU corrects the estimated end positions P1 and P2 with the correction value acquired in S1304 or S1305 according to the determination result in S1303. At S1307, the CPU stores the corrected estimated end positions P1 and P2 in the RAM or nonvolatile RAM. After that, the CPU ends this routine once.

(effect)
As described above, in the configuration of the present embodiment, the type of the obstacle B can be favorably determined based on the frequency distribution of the change amount of the detection distance. Therefore, according to such a configuration, the end position of the obstacle B can be estimated better than in the past based on the determined type. That is, for example, the end position of the obstacle B can be corrected well based on the determined type. Further, based on the determined type, detection of the parking space PS and setting of the traveling condition of the vehicle V may be appropriately performed.

(Modification)
The present disclosure is not limited to the above embodiment. Therefore, the above embodiment can be modified as appropriate. Hereinafter, representative modifications will be described. In the following description of the modification, only parts different from the above embodiment will be described. Moreover, in the said embodiment and modification, the same code | symbol is attached | subjected to the part which is mutually identical or equal. Therefore, in the following description of the modification, with regard to components having the same reference numerals as the above embodiment, the description in the above embodiment may be appropriately incorporated unless a technical contradiction or a special additional description.

The present disclosure is not limited to the specific device configuration shown in the above embodiment. For example, the distance measuring unit 2 may include only the distance measuring sensor 20, and the function of calculating the detected distance may be provided on the parking assist ECU 10 side.

The parking assistance ECU 10 may be configured as an ASIC such as a gate array. ASIC is an abbreviation of Application Specific Integrated Circuit.

The present disclosure is not limited to the specific operation aspect and processing aspect shown in the above embodiment. For example, curve approximation and generation of a frequency distribution based thereon may be started after acquisition of a predetermined number of detection distances. That is, until there is acquisition of a predetermined number of detection distances, curve approximation and generation of a frequency distribution based on this may be awaited. In this case, also in the type determination, the process waits until the predetermined number of detection distances are obtained.

The curve approximation and the creation of the frequency distribution based thereon may be performed after the acquisition of the detection distance corresponding to one obstacle B is completed. That is, until the acquisition of the detection distance corresponding to one obstacle B is completed, the curve approximation and the creation of the frequency distribution based on this may be on standby. In this case, the type determination is also on standby until acquisition of the detection distance corresponding to one obstacle B is completed. "D7", "D6" ..., "D1", "C", "U1", "U2" ... "U7" in the histogram of the frequency distribution shown in Fig. 5 etc. The variation range may be changed from the above example.

During detection of the parking space PS, typically, the vehicle speed is substantially constant, and the traveling direction of the vehicle is also constant. In particular, when parking assistance is performed by automatic driving or semi-automatic driving, the vehicle speed during detection of the parking space PS can be controlled uniformly by the torque control ECU 8. When the vehicle speed and the traveling direction of the vehicle are constant, the amount of change in the detection distance, which is the basis of the frequency distribution, is specified by a parameter different from the slope of the line segment connecting two adjacent points in the point sequence by multiple detection positions PD. It can be done. Specifically, the change amount of the detection distance may be a difference of the detection distance. That is, the change amount of the detection distance can also be defined as a change rate of the detection distance per unit time or per unit travel distance. Further, regardless of the vehicle speed, the tangent of the approximate curve at the detection position PD may be used as the amount of change of the detection distance.

When the determination result by the type determination unit 104 is a vehicle, the correction value by the end position correction unit 105 is not limited to the above specific example. That is, for example, when the determination result by the type determination unit 104 is a vehicle, the reference dimension in the end position correction unit 105 is the maximum value of the practical vehicle dimensions in an ordinary car except limousine type It is good also as 1,920 mm x vehicle total length 5,300 mm. In addition, when the determination result by the type determination unit 104 is a vehicle, the correction value by the end position correction unit 105 may be a constant value as in the above specific example, unlike the above specific example. It may be a variable value.

Specifically, for example, when the determination result by the type determination unit 104 indicates a vehicle, the end position correction unit 105 determines the vehicle type based on the distance between P1 and P2 in the X direction before correction, and The reference dimension may be changeable based on the determination result. That is, when the determination result by the type determination unit 104 is a vehicle and the distance between P1 and P2 is less than 1,480 mm, the end position correction unit 105 determines the vehicle width 1, which is the vehicle standard dimension of a mini vehicle The reference dimension may be 480 mm × total vehicle length 3,400 mm. Alternatively, for example, the end position correction unit 105 may have a function of learning the correction value by measuring the distance between the side surface of the parked vehicle adjacent to the vehicle V and the vehicle V when parking is completed. In this case, the learning value may have an upper limit value based on the reference dimension as described above.

In addition, the end position correction unit 105 may determine the correction amount by means or method other than the means or method described above. For example, the end position correction unit 105 may determine the correction amount based on the country, region, time zone, and other information in which the vehicle V is traveling. Specifically, the end position correction unit 105 determines the reference dimension based on the information of the country, the region, and the like acquired by the traveling state acquisition unit 3 or the like, and determines the correction amount based on the determined reference dimension. May be

The initial activation of the parking assistance routine shown in FIG. 11 is not limited to activation by user operation. That is, for example, the CPU may automatically start the parking assistance routine after a predetermined time has elapsed from the start of the engine. In addition, the CPU automatically activates the parking assistance routine after a predetermined time has elapsed from when the parking assistance operation is interrupted for any reason or when the parking assistance operation is forcibly ended based on user input or the like. May be

The contents of the travel condition calculation in S1105 are not limited to the above specific example. Even while the determination in S1106 is YES, the CPU may recalculate the parking space PS and the traveling condition without skipping the process of detecting the parking space PS and calculating the traveling condition. The traveling for a predetermined time in S1107 may be changed to traveling for a predetermined distance.

The determination of the fourth condition may be performed only when the vehicle speed is equal to or higher than a predetermined value or when the density of the detection position PD is low. In the latter case, the travel distance of the vehicle V from the time of acquisition of PD (1) to the time of acquisition of PD (N) is RD (the detection position PD (N) acquired up to this time corresponding to one obstacle B). N). At this time, when the value obtained by dividing RD (N) by N is less than a predetermined value, the fourth condition determination in S1210 is executed, while when it is the predetermined value or more, the fourth condition determination in S1210 is It may be skipped.

The non-vehicle correction value may be absent. That is, the process of S1305 may be omitted.

When updating the distribution of the change amount T (N) in S1205, the CPU may interpolate between the detection position PD (N-1) and PD (N). Specifically, the CPU divides the approximate curve PL between PD (N-1) and PD (N) at equal intervals in the X direction, calculates the amount of change T for each division point, and updates the distribution You may Such interpolation processing may be performed, for example, when N is less than a predetermined value Nth1.

The determination content in S1211 is not limited to the above specific example. That is, for example, when all of the first to third conditions are satisfied, the CPU makes the determination of S 1211 “YES”, while any one of the first to third conditions is satisfied. Even if not, if the above-mentioned fourth condition is satisfied, the determination of S 1211 may be set to “YES”. The CPU may not determine the fourth condition when N is a predetermined value Nth2 or more, but may determine the fourth condition when N is less than the predetermined value Nth2. The predetermined value Nth2 may be the same value as the predetermined value Nth1 or may be a different value. The determination content of the fourth condition is also not limited to the above specific example.

"Acquisition" can be appropriately changed to similar expressions such as "estimate", "detection", "detection", "calculation" and the like. The inequality sign in each determination process may be with an equal sign or without an equal sign. That is, for example, "less than a predetermined value" may be changed to "equal to or less than a predetermined value".

The modified example is also not limited to the above example. Also, multiple variants may be combined with one another. Furthermore, all or part of the above embodiment and all or part of the modification may be combined with each other.

Claims

An obstacle detection device (1) configured to detect an obstacle (B) present outside the mobile body by being mounted on the mobile body (V),
The detection distance corresponding to the distance to the obstacle is transmitted along with the movement of the mobile body by transmitting the search wave toward the outside of the mobile body and receiving the reflected wave of the search wave due to the obstacle. A distance measuring unit (2) provided to repeatedly obtain
A frequency distribution acquisition unit (101) provided to acquire the change amount of the detection distance for each acquisition of the detection distance in the distance measuring unit and to acquire the frequency distribution of the acquired change amount;
A type determination unit (104) provided to determine the type of the obstacle based on the frequency distribution acquired by the frequency distribution acquisition unit;
Obstacle detection device equipped with
The type determination unit
A first condition based on a first number, which is a frequency of a small change area in which the absolute value of the change amount is small;
A second condition based on a second frequency which is the frequency of the large change area where the absolute value of the change amount is large;
A third condition based on a relationship between the first number and the second frequency,
Provided to determine whether the obstacle is a vehicle based on
The obstacle detection device according to claim 1.
A movement state acquisition unit (3) provided to acquire the movement state of the moving body;
An estimated reflection position of the search wave by the obstacle corresponding to each of the plurality of detected distances acquired by the distance measuring unit is provided based on the detected distance and the movement state. , A reflection position acquisition unit (102),
And further
The type determination unit determines whether or not the obstacle is a vehicle based on the distribution state of the plurality of estimated reflection positions acquired by the reflection position acquisition unit corresponding to one of the obstacles. Provided as
The obstacle detection device according to claim 1.
A movement state acquisition unit (3) provided to acquire the movement state of the moving body;
An estimated reflection position of the search wave by the obstacle corresponding to each of the plurality of detected distances acquired by the distance measuring unit is provided based on the detected distance and the movement state. , A reflection position acquisition unit (102),
An edge position estimation provided to estimate an edge position of the obstacle based on the estimated reflection position located at an end of the plurality of estimated reflection positions in the moving direction of the movable body Part (103),
An end position correction unit (105) provided to correct the end position based on the type determined by the type determination unit;
Further equipped with
The obstacle detection device according to any one of claims 1 to 3.
The end position correction unit is provided to correct the end position based on a vehicle dimension defined in a vehicle standard when the type determined by the type determination unit is the vehicle.
The obstacle detection device according to claim 4.
The determination result of the type is used for detection of a parking space (PS) between a plurality of obstacles or calculation of movement conditions for the movement of the moving body while avoiding a collision with the obstacles. It has become,
The obstacle detection device according to any one of claims 1 to 5.