Parking assistance device and parking assistance method

Abstract

To provide a parking assistance device that can offer appropriate parking assistance while taking into account measurement errors that occur when observing distant objects with external sensors. [Solution] A parking assistance device for assisting in the parking of a vehicle, comprising: an other vehicle information acquisition unit that acquires information on other vehicles in the vicinity of the vehicle based on information acquired from an on-board external sensor; a parking vehicle row identification unit that identifies parked vehicles and rows of parked vehicles in the vicinity of the vehicle based on the information on other vehicles; a likelihood calculation unit that calculates the likelihood that there is an empty parking space between adjacent parked vehicles based on the distance between adjacent parked vehicles and the width of the parking space in the row of parked vehicles; and a vehicle driving control unit that controls the driving of the vehicle based on the likelihood, wherein the vehicle driving control unit generates a driving path that passes through the reference point or a driving path that parks at the reference point, according to the result of comparing a first likelihood at a predetermined reference point within the sensing range of the external sensor and a second likelihood at a parking space farther from the reference point.

Description

[Technical Field]

[0001] The present invention relates to a parking assistance device and a parking assistance method for assisting in the parking of a vehicle. [Background technology]

[0002] Patent Document 1 is an example of a document disclosing a parking assistance function. For example, the abstract of this document states that the problem is "to suppress the possibility of mistakenly identifying a no-parking location as a parking space, even when the parking space cannot be recognized when attempting to park one's own vehicle using the parking assistance function," and that the solution is "the parking assistance device 1 detects parked vehicles Pv in the parking lot based on environmental information around the own vehicle M, and even if the width of the space next to this parked vehicle is greater than or equal to the required parking width Ws set in advance for parking the own vehicle M, if no other parked vehicles Pv are detected at the back of this space, it is presumed that parking is prohibited in this space."

[0003] Furthermore, the document discloses external sensors for recognizing the surrounding environment of a vehicle, including monocular cameras and stereo cameras using CCD or CMOS image sensors, ultrasonic sensors, millimeter-wave radar, microwave radar, infrared sensors, laser radar, and LiDAR (Light Detection and Ranging). [Prior art documents] [Patent Documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2023-17462 [Overview of the project] [Problems that the invention aims to solve]

[0005] However, all of the external sensors exemplified in the document have the characteristic that the measurement error is larger at a distance than at a nearby distance. As a result, in the parking assistance device of Patent Document 1, the reliability of the estimation at a distance is lower than that of the estimation at a nearby distance, which can lead to errors in judgment at a distance and prevent the provision of an appropriate parking assistance function.

[0006] Therefore, the present invention aims to provide a parking assistance device and a parking assistance method that can provide appropriate parking assistance while taking into account measurement errors that occur when observing distant objects with external sensors. [Means for solving the problem]

[0007] To solve the above problems, the present invention provides a parking assistance device for assisting in the parking of a vehicle, comprising: an other vehicle information acquisition unit that acquires information on other vehicles in the vicinity of the vehicle based on information acquired from an on-board external sensor; a parking vehicle row identification unit that identifies parked vehicles and rows of parked vehicles in the vicinity of the vehicle based on the information on other vehicles; a likelihood calculation unit that calculates the likelihood that the space between adjacent parked vehicles is an empty parking space based on the distance between adjacent parked vehicles and the width of the parking space in the row of parked vehicles; and a vehicle driving control unit that controls the driving of the vehicle based on the likelihood, wherein the vehicle driving control unit generates a driving route that passes through the reference point or a driving route that parks at the reference point, according to the result of comparing a first likelihood at a predetermined reference point within the sensing range of the external sensor with a second likelihood at a parking space farther from the reference point. [Effects of the Invention]

[0008] According to the parking assistance device and parking assistance method of the present invention, it is possible to provide appropriate parking assistance while taking into account measurement errors that occur when observing distant objects with an external sensor. [Brief explanation of the drawing]

[0009] [Figure 1] Functional block diagram of the vehicle system in Example 1. [Figure 2]Functional block diagram of the processing unit of the parking support device according to the first embodiment. [Figure 3] An example of the parked vehicle column data group according to the first embodiment. [Figure 4] Conceptual diagram showing the measurement error characteristics of the external sensor. [Figure 5] An example of the measurement error data group according to the first embodiment. [Figure 6A] Conceptual diagram showing the distance between parked vehicles. [Figure 6B] Conceptual diagram showing that the position error of the parked vehicle increases as the distance from the host vehicle increases. [Figure 7] An example of the parking lot environment according to the first embodiment. [Figure 8] Processing flowchart of the likelihood calculation unit according to the first embodiment. [Figure 9] Processing flowchart of the vehicle travel control unit according to the first embodiment. [Figure 10] An example of the parking lot environment where the process proceeds from step S16c to step S16e in FIG. 9. [Figure 11] An example of the parking lot environment where the process proceeds from step S16c to step S16d in FIG. 9. [Figure 12] Functional block diagram of the processing unit of the parking support device according to the second embodiment. [Figure 13] An example of the parking lot environment according to the second embodiment. [Figure 14] Functional block diagram of the processing unit of the parking support device according to the third embodiment. [Figure 15] An example of the parking lot environment according to the third embodiment. [Figure 16] Functional block diagram of the processing unit of the parking support device according to the fourth embodiment. [Figure 17] Processing flowchart of the vehicle travel control unit according to the fourth embodiment.

Best Mode for Carrying Out the Invention

[0010] Hereinafter, embodiments of the parking support device of the present invention will be described with reference to the drawings.

Embodiment

[0011] First, the parking assistance device 1 according to Embodiment 1 of the present invention will be described using Figures 1 to 11.

[0012] Figure 1 is a functional block diagram showing the configuration of the vehicle system 100, including the parking assist device 1 of this embodiment. This vehicle system 100 is a system mounted on the vehicle V0, and after confirming the road and obstacle conditions around the vehicle, it provides appropriate driving assistance functions and autonomous driving functions. In order to realize the parking assist function, which is one of the driving assistance functions and autonomous driving functions, the vehicle system 100 of this embodiment includes, in addition to the parking assist device 1, a vehicle sensor group 2, an external sensor group 3, an actuator group 4, an HMI device group 5, and an in-vehicle network N that connects them. Below, the outlines of the vehicle sensor group 2 to the HMI device group 5 will be described sequentially, and then the parking assist device 1 of this embodiment will be described in detail.

[0013] <Vehicle sensor group 2> The vehicle sensor group 2 is a collection of sensors that detect various states of the vehicle V0. Each vehicle sensor detects information such as the position of the vehicle V0, driving speed, steering angle, accelerator pedal operation amount, brake pedal operation amount, etc., and transmits it to the parking assist device 1 via the in-vehicle network N.

[0014] <External sensor group 3> The external sensor group 3 is a collection of on-board sensors that detect the conditions around the vehicle. Examples of external sensors include cameras, millimeter-wave radar, LiDAR, sonar, and microphones. The external sensor group 3 detects environmental elements such as visible obstacles, road markings, signs, and signals within a predetermined range from the vehicle V0, and these detection results are output to the parking assist device 1 via the on-board network N.

[0015] Here, "obvious obstacles" refer to, for example, other vehicles besides the vehicle V0, pedestrians, objects that have fallen onto the road, the roadside, etc. "Road markings" refer to, for example, white lines, pedestrian crossings, stop lines, etc. In addition, the external sensor group 3 outputs information regarding the detection status to the parking assist device 1 via the in-vehicle network N, based on its own sensing range and status.

[0016] <Actuator group 4> The actuator group 4 is a group of devices that control control elements such as steering, brakes, and accelerator that determine the movement of the host vehicle V0. The actuator group 4 controls the movement of control elements such as steering, brakes, and accelerator based on operation information of a steering wheel, a brake pedal, an accelerator pedal, etc. by a driver and a control command value output from the parking support device 1, thereby controlling the behavior of the host vehicle V0 and executing automatic driving.

[0017] <HMI device group 5> The HMI device group 5 is a group of devices called a Human Machine Interface that performs command input from an occupant to the vehicle system 100 and information notification from the vehicle system 100 to the occupant. The HMI device group 5 includes a display, a speaker, a vibrator, a switch, a touch panel, etc.

[0018] <Parking support device 1> The parking support device 1 is specifically an ECU (Electronic Control Unit) mounted on the host vehicle V0, and includes a processing unit 10, a storage unit 20, and a communication unit 30.

[0019] The processing unit 10 is configured to include, for example, a CPU (Central Processing Unit) which is a central processing unit. However, in addition to the CPU, it may be configured to include a GPU (Graphics Processing Unit), an FPGA (Field-Programmable Gate Array), an ASIC (Application Specific Integrated Circuit), etc., or may be configured by any one of them.

[0020] The storage unit 20 is comprised of storage devices such as an HDD (Hard Disk Drive), flash memory, and ROM (Read Only Memory), as well as memory such as RAM (Random Access Memory). The storage unit 20 stores programs to be processed by the processing unit 10, as well as data sets necessary for that processing. It is also used as main memory when the processing unit 10 executes a program, temporarily storing data necessary for program calculations.

[0021] The communication unit 30 has the function of sending and receiving data based on various protocols with other devices connected via the in-vehicle network N. The communication unit 30 is composed of, for example, a network card compliant with communication standards such as IEEE 802.3 or CAN (Controller Area Network).

[0022] Next, the details of the processing unit 10 of the parking assistance device 1 in this embodiment will be explained using the functional block diagram in Figure 2. As shown here, the processing unit 10 includes an information acquisition unit 11, an other vehicle information acquisition unit 12, a parking vehicle row identification unit 13, a parking space width identification unit 14, a likelihood calculation unit 15, a vehicle driving control unit 16, and an information output unit 17. The processing unit 10 is connected to a vehicle sensor group 2 and an external sensor group 3 on the input side via a communication unit 30 (not shown), and to an actuator group 4 and an HMI device group 5 on the output side. Each functional unit within the processing unit 10 is realized through the cooperation of hardware such as the processing unit 10 and the memory unit 20 and various software, but the following explanation will omit well-known technologies in this type of computer technology field.

[0023] <<Information acquisition section 11>> The information acquisition unit 11 is a functional unit that stores information acquired from the outside by the parking assist device 1 in the storage unit 20. For example, the information acquisition unit 11 stores information related to the behavior of the vehicle V0, such as its movement and state, detected by the vehicle sensor group 2, in the storage unit 20 as the vehicle data group D1. In addition, the information acquisition unit 11 stores information about obstacles around the vehicle V0 and the detection area of ​​the external sensor group 3, detected by the external sensor group 3, in the storage unit 20 as the sensor recognition data group D2.

[0024] <<Other Vehicle Information Acquisition Unit 12>> The other vehicle information acquisition unit 12 is a functional unit that identifies other vehicles (moving vehicles, parked vehicles) around the vehicle based on the sensor recognition data group D2 stored in the storage unit 20, and stores information such as their position, attitude, tilt, and speed in the storage unit 20 as the other vehicle data group D3. The position information of other vehicles in the other vehicle data group D3 is expressed in the relative coordinate system of the vehicle V0.

[0025] <<Parking vehicle row identification section 13, parking space width identification section 14>> The parking vehicle row identification unit 13 is a functional unit that identifies parked vehicles around the vehicle based on the other vehicle data group D3 stored in the storage unit 20, as well as identifying parking rows in which multiple adjacent vehicles are parked in the same direction, and the position of each parked vehicle.

[0026] Furthermore, the parking space width identification unit 14 statistically processes the positional relationship of each parked vehicle within the parking row identified by the parking vehicle row identification unit 13, and determines the width d of each parking space. p This is a functional unit that identifies the parking space width d based on the parking space lines recognized based on the output of the external sensor group 3. p You may specify it.

[0027] Figure 3 shows an example of a parking vehicle row data group D4, which combines the outputs of the parking vehicle row identification unit 13 and the parking space width identification unit 14. In the figure, column D4a is a data column showing the parking row ID assigned to each parking row, column D4b is a data column showing the parking vehicle ID assigned to each parked vehicle, column D4c is a data column showing the relative position of each parked vehicle with respect to the current position of the vehicle V0, and column D4d is a data column showing the width of each parking space.

[0028] <<Likelihood Calculation Unit 15>> The likelihood calculation unit 15 is a functional unit that calculates the likelihood L of each available parking space candidate based on the aforementioned parking vehicle row data group D4 and the pre-prepared measurement error data group D5, and stores it in the storage unit 20 as the available parking space candidate data group D6.

[0029] First, Figures 4 and 5 will be used to explain the measurement error data set D5. Figure 4 is a conceptual diagram illustrating the measurement error characteristics of Lidar3a, a type of external sensor. Lidar3a is a sensor that measures the distance to an object in the direction of laser irradiation based on the time difference between the irradiation time of laser light and the reception time of reflected light. There is an unavoidable angular error α between the design light irradiation direction X of Lidar3a and the actual light irradiation direction X'. e Because of this, the position error e detected by Lidar3a tends to increase in proportion to the distance. Therefore, if you compare the position error e1 at a nearby location with the position error e2 at a distant location, the latter will naturally be larger.

[0030] Note that in Figure 4, the angular error α of Lidar3a is shown. eAlthough attention was paid to [the above], in reality, an error in depth due to light diffusion or the like also occurs. Therefore, the actual position error e is also affected by the error in depth. Also, even when using a stereo camera instead of Lidar3a, there is a similar tendency for the position error e to increase in proportion to the distance due to the influence of quantization error, calibration error, etc. Therefore, if an example of the measurement error data group D5 corresponding to a certain external sensor is given, it becomes data such that the position error increases as the distance increases, as shown in FIG. 5. In the same figure, column D5a is a data column indicating the distance from the external sensor, and column D5b is a data column indicating the average value of the position error.

[0031] Next, the accuracy of the distance between parked vehicles, which is calculated based on the parked vehicle column data group D4, will be described using FIGS. 6A and 6B. As shown in FIG. 6A, if the positions of three parked vehicles V1, V2, and V3 are measured with Lidar3a mounted on the host vehicle V0, the distance d between the two vehicles can be calculated from the difference in the positions of parked vehicles V1 and V2, and the distance d between the two vehicles can be calculated from the difference in the positions of parked vehicles V2 and V3. However, due to the mechanism described in FIG. 4, there is a tendency such that e3 > e2 > e1 in the position errors e1, e2, and e3 of parked vehicles V1, V2, and V3, as shown in FIG. 6B. Therefore, the reliability of the nearby distance d when viewed from the host vehicle V0 is relatively high, and the reliability of the distant distance d is relatively low. b1 can be calculated, and the distance d between the two vehicles can be calculated from the difference in the positions of parked vehicles V2 and V3. b2 However, due to the mechanism described in FIG. 4, there is a tendency such that e3 > e2 > e1 in the position errors e1, e2, and e3 of parked vehicles V1, V2, and V3, as shown in FIG. 6B. Therefore, the reliability of the nearby distance d when viewed from the host vehicle V0 is relatively high, and the reliability of the distant distance d is relatively low. b1 is relatively high, and the reliability of the distant distance d b2 is relatively low.

[0032] Therefore, as in an example of the parking lot environment shown in FIG. 7, when there are two available parking sections B1 and B2 within the sensing range R of Lidar3, the accuracy of the distance d between two vehicles sandwiching the parking section B1 close to the host vehicle V0 is relatively high, and the accuracy of the distance d between two vehicles sandwiching the parking section B2 far from the host vehicle V0 is relatively low. b1 is relatively high, and the accuracy of the distance d between two vehicles sandwiching the parking section B2 far from the host vehicle V0 b2 is relatively low.

[0033] Therefore, the likelihood calculation unit 15 of the present embodiment uses the likelihood L, which is information referred to when the vehicle travel control unit 16 determines the parking section where the vehicle actually parks, as the distance d between parked vehicles. bThe calculation is performed while considering the accuracy and measurement error data set D5. Below, the details of the likelihood calculation process by the likelihood calculation unit 15 will be explained according to the processing flowchart in Figure 8.

[0034] First, in step S15a, the likelihood calculation unit 15 obtains the parking vehicle row data group D4 and the measurement error data group D5 from the storage unit 20.

[0035] Next, in step S15b, the likelihood calculation unit 15 refers to column D4a of the parking vehicle row data group D4 and selects the desired parking row ID (for example, the smallest parking row ID).

[0036] In step S15c, the likelihood calculation unit 15 refers to column D4c of the parking vehicle row data group D4 and selects pairs of adjacent parked vehicles that belong to the parking row ID selected in step S15b.

[0037] In step S15d, the likelihood calculation unit 15 refers to column D4d of the parking vehicle row data group D4 and the inter-vehicle distance d of the parking vehicle pair selected in step S15c. b However, parking space width d p Determine if it is greater. If the requirement is met, proceed to step S15e; otherwise, proceed to step S15i.

[0038] In step S15e, the likelihood calculation unit 15 stores the space between the pair of parked vehicles selected in step S15c as a candidate for an available parking space.

[0039] In step S15f, the likelihood calculation unit 15 refers to column D4c of the parking vehicle row data group D4 and calculates the distances d1 and d2 from its own vehicle V0 to each of the selected parking vehicle pairs.

[0040] In step S15g, the likelihood calculation unit 15 refers to the measurement error data group D5 and obtains position errors e1 and e2 corresponding to distances d1 and d2. If the position errors e1 and e2 for distances d1 and d2 are not registered in the pre-prepared measurement error data group D5, the position error e corresponding to an arbitrary distance d can be obtained by linearly interpolating the preceding and succeeding data.

[0041] In step S15h, the likelihood calculation unit 15 calculates a likelihood L for each candidate for an available parking space, which takes a value between 0 and 1. Various methods can be used to calculate the likelihood L, but for example, the following (Equation 1) can be used.

[0042]

number

[0043] In step S15i, the likelihood calculation unit 15 checks whether the processes in steps S15d to S15h have been performed for all parking vehicle pairs within the parking row ID selected in step S15b. If the requirements are met, the unit proceeds to step S15j; otherwise, it returns to step S15c. This ensures that the above processes are performed for all parking vehicle pairs within the same vehicle row ID.

[0044] In step S15j, the likelihood calculation unit 15 checks whether the processes in steps S15c to S15i have been performed for all parking column IDs registered in column D4a of the parking vehicle column data group D4. If the requirements are met, the process shown in Figure 8 is terminated; otherwise, the process returns to step S15b. This ensures that the above-described process is performed for all vehicle column IDs.

[0045] Through the above process, the likelihood calculation unit 15 can calculate the likelihood L for all available parking space candidates that exist within the sensing range R of its own vehicle V0, and can generate a data set D6 of available parking space candidates by summarizing the likelihood L of each available parking space candidate.

[0046] <<Vehicle Driving Control Unit 16>> The vehicle driving control unit 16 is a functional unit that, based on the sensor recognition data group D2, the parked vehicle line data group D4, and the available parking space candidate data group D6, determines whether to pass the nearest available parking space candidate (hereinafter referred to as "reference point") on the side of the vehicle V0's direction of travel, and stores control command values ​​that automatically drive or propose a replanned route according to the determination result in the storage unit 20 as vehicle control data group D7. The details of the vehicle driving control unit 16 will be explained below with reference to Figures 9 to 11.

[0047] Figure 9 is a flowchart of the processes performed by the vehicle driving control unit 16.

[0048] First, in step S16a, the vehicle driving control unit 16 obtains the sensor recognition data group D2, the parked vehicle row data group D4, and the available parking space candidate data group D6 from the storage unit 20.

[0049] Next, in step S16b, the vehicle driving control unit 16 sets a hypothetical likelihood L0 for the reference point, which will be used as a threshold in the next step. For example, the hypothetical likelihood L0 is 1.0.

[0050] In step S16c, the vehicle driving control unit 16 compares the sum of the likelihoods of available parking space candidates other than the reference point on the side of the vehicle V0's direction of travel with the threshold (provisional likelihood L0) defined in step S16b, and determines whether the sum is greater than the threshold. If the sum is greater than the threshold, the unit proceeds to step S16d; otherwise, it proceeds to step S16e.

[0051] Here, we will use Figures 10 and 11 to explain the cases where we proceed to step S16d and the cases where we proceed to step S16e.

[0052] Figure 10 shows an example where the likelihood L1 of parking space B1, a candidate for an available parking space closer to the vehicle V0, is 0.8, and the likelihood L2 of parking space B2, a candidate for an available parking space further away from the vehicle V0, is 0.4. In this situation, a provisional likelihood L0 of 1.0 is set for the reference point, parking space B1, and the sum of the likelihoods of the other candidate available parking spaces is 0.4. Therefore, since the sum (0.4) is less than the threshold (1.0), the process proceeds to step S16e.

[0053] On the other hand, Figure 11 shows an example where the likelihood L1 of parking space B1, a candidate for an available parking space closer to the vehicle V0, is 0.8, and the likelihoods L2, L3, and L4 of parking spaces B2, B3, and B4, candidate for available parking spaces further away from the vehicle V0, are 0.5, 0.4, and 0.2, respectively. In this situation, a provisional likelihood L0 of 1.0 is set for the reference point, parking space B1, and the sum of the likelihoods of the candidate available parking spaces other than the reference point is 1.1. Therefore, since the sum (1.1) is greater than the threshold (1.0), the process proceeds to step S16d.

[0054] In step S16d, the vehicle driving control unit 16 generates a driving path that passes through the current reference point (parking space B1 in the example of Figure 11) and proceeds to the next reference point.

[0055] On the other hand, in step S16e, the vehicle driving control unit 16 generates a driving route to park at the current reference point (parking space B1 in the example of Figure 10). In the example of Figure 11, when the provisional likelihood L0 (1.0) is set for parking space B2, which is the next reference point after parking space B1, the sum of the likelihoods of the available parking space candidates other than the reference point becomes 0.6, and since the threshold > sum, the process in step S16e is executed.

[0056] In step S16f, the vehicle driving control unit 16 calculates a predetermined vehicle control command value to drive along the driving path generated in step S16d or step S16e.

[0057] Through the above processing, the vehicle driving control unit 16 can generate a set of vehicle control data D7 that will park the vehicle in any parking space located within the sensing range R of the vehicle V0.

[0058] <<Information Output Unit 17>> The information output unit 17 is a functional unit that outputs various information to the actuator group 4 and the HMI device group 5 via the in-vehicle network N. As a result, if the vehicle V0 is an autonomous vehicle, the actuator group 4 can be controlled to automatically pass through reference points or automatically park at reference points according to the vehicle control data group D7. Furthermore, if the vehicle V0 is a vehicle equipped with driver assistance functions, the vehicle driving control unit 16 will notify the occupants of the proposed parking space via the HMI device group 5.

[0059] In either case, the location of potential available parking spaces, the likelihood of each available parking space being available, and the decision on whether or not to pass the nearest available parking space may be communicated to the occupants via the HMI device group 5. In this case, the occupants of the vehicle V0 can choose an action based on the recommendations of the parking assistance device 1.

[0060] <Effects of this embodiment> The parking assistance device of this embodiment, as described above, can provide appropriate parking assistance while taking into account measurement errors that occur when observing distant objects with external sensors. In an environment like that shown in Figure 11, the reason why the vehicle V0 does not immediately park in the nearest available parking space candidate (parking space B1) is that other available parking space candidates (parking spaces B1 to B4) are closer to the target facility and are therefore more desirable parking spaces. [Examples]

[0061] Next, Example 2 of the present invention will be described using Figures 12 and 13. Note that repetitive explanations of points common to Example 1 will be omitted.

[0062] In Example 1, only parking spaces that were currently vacant were considered as candidates for vacant parking spaces. However, in this embodiment, parking spaces that are expected to be vacated soon are considered as candidates for vacant parking spaces even if there is a vehicle parked in them at the moment. To accommodate this, the processing unit 10 of this embodiment includes an exit time level estimation unit 18 that provides information about the vehicle exiting to the vehicle driving control unit 16. The details of the parking assistance device 1 of this embodiment will be described below, focusing on the operation of the exit time level estimation unit 18.

[0063] Figure 12 is a functional block diagram of the processing unit 10 of the parking assistance device 1 in this embodiment. As shown in the figure, the exit time level estimation unit 18 receives sensor recognition data group D2 from the information acquisition unit 11 and parking vehicle row data group D4 from the parking vehicle row identification unit 13. Based on both data groups, it identifies vehicles that are likely to exit the parking space around its own vehicle V0 and estimates the time level until those vehicles exit. The time level estimated here is stored in the storage unit 20 as exit vehicle data group D8 and then used by the vehicle driving control unit 16.

[0064] Figure 13 shows an example of a parking environment in this embodiment, where a vehicle parked in parking space B1 is about to leave, and there is no vehicle parked in parking space B2. In this embodiment, both parking space B1 and parking space B2 are potential empty parking spaces, and the nearest potential empty parking space on the side of the vehicle V0's direction of travel is parking space B1, so parking space B1 is initially used as the reference point.

[0065] Here, the method by which the exit time level estimation unit 18 estimates the time level will be explained in detail. For example, if the engine of a parked vehicle is running, the headlights are on, or the turn signals are flashing, the exit time level estimation unit 18 determines that the parked vehicle will be leaving soon and estimates the time level to be T1. On the other hand, if there is a person getting into the parked vehicle, it determines that it will take some time for the parked vehicle to leave and estimates the time level to be T2.

[0066] Therefore, in step S16b, the vehicle driving control unit 16 sets the provisional likelihood L0 corresponding to the time level T estimated by the exit time level estimation unit 18 as a reference point, and then performs the processing from step S16c onward. For example, if the provisional likelihood L0 assigned to the reference point at time level T1 is 0.6, and the provisional likelihood L0 set for the reference point at time level T2 is 0.3, then if there is a person about to get into a parked vehicle in parking space B1 in Figure 13, time level T2 is set for that parked vehicle, and a threshold (provisional likelihood L0) of 0.3 is set for parking space B1. Therefore, if the likelihood L2 for parking space B2 is 0.4, the sum > threshold, so the vehicle driving control unit 16 selects step S16d (passing through parking space B1, which is the reference point).

[0067] According to the parking assistance device of this embodiment described above, it is possible to identify a parking space while also considering the behavior of parked vehicles that are expected to leave the parking space soon. [Examples]

[0068] Next, Embodiment 3 of the present invention will be described using Figures 14 and 15. Note that repetitive explanations of common points with the above embodiments will be omitted.

[0069] In Examples 1 and 2, the driving path of the vehicle V0 was generated considering only parked vehicles within the parking space. However, in this example, non-parked vehicles such as vehicles traveling in front of the vehicle V0 are also considered. A The behavior of the non-parked vehicle V is also taken into consideration when generating the driving path of the own vehicle V0. Therefore, in the processing unit 10 of this embodiment, the behavior of the non-parked vehicle V A A non-parked vehicle identification unit 19 has been added to provide information to the vehicle driving control unit 16. The details of the parking assistance device 1 of this embodiment will be described below, focusing on the operation of the non-parked vehicle identification unit 19.

[0070] Figure 14 is a functional block diagram of the processing unit 10 of the parking assistance device 1 in this embodiment. As shown in the figure, the non-parked vehicle identification unit 19 receives a group of other vehicle data D3 from the other vehicle information acquisition unit 12, and based on this data group, identifies non-parked vehicles V that are driving ahead of the vehicle V0 or that are temporarily stopped outside the parking space.A Identify the non-parked vehicle V identified here. A The information is stored in the storage unit 20 as non-parked vehicle data group D9, and then used by the vehicle driving control unit 16.

[0071] The non-parked vehicle data group D9 generated by the non-parked vehicle identification unit 19 includes vehicle ID, vehicle position, vehicle speed, direction of travel, parking schedule flag, exit flag, etc., for non-parked vehicles V A This is a set of data recorded for each instance. Note that the parking intention flag indicates a non-parked vehicle (V) that flashes its hazard lights near an empty parking space or changes direction towards an empty parking space. A This is a flag assigned to a non-parked vehicle V that has left its parking space. A This is a flag to be assigned to [the object].

[0072] In this embodiment, each step of the processing flowchart in Figure 8 is performed as follows.

[0073] In step S15a, the likelihood calculation unit 15 acquires the parking vehicle row data group D4 and the measurement error data group D5 from the storage unit 20, as well as the non-parked vehicle data group D9.

[0074] In step S15h, the likelihood calculation unit 15 calculates the likelihood in the same manner as in Example 1, and then adjusts the provisionally calculated likelihood appropriately by referring to the non-parked vehicle data group D9. The likelihood adjustment method using the parking schedule flag and the departure flag will be described in detail below.

[0075] <How to use the parking reservation flag> For example, as shown in the example of a parking environment in Figure 15, if there are four candidate empty parking spaces (parking spaces B1 to B4) within the sensing range R of the vehicle V0, and the likelihoods for each are L1=0.8, L2=0.5, L3=0.4, and L4=0.2, then the likelihood calculation unit 15 of this embodiment will determine the likelihood of the unparked vehicle V in front. A When the parking scheduled flag is assigned to a vehicle, non-parked vehicle V AThe likelihood L of the nearest available parking space candidate is reduced by a certain value (e.g., 0.4). As a result, under the circumstances shown in Figure 15, the likelihood L2 of parking space B2 is adjusted from 0.5 to 0.1, and the sum is also reduced from 1.1 to 0.7.

[0076] Therefore, under the conditions shown in Figure 15, the vehicle driving control unit 16 of Embodiment 1 would determine that threshold < sum and generate a driving path that passes through the nearest parking space B1. However, the vehicle driving control unit 16 of this embodiment would determine that threshold > sum and generate a driving path that parks in the nearest parking space B1.

[0077] In the example above, non-parked vehicle V is assigned the parking planned flag. A Only the likelihood L of the nearest available parking space candidate was adjusted, but the non-parked vehicle V A You may also adjust the likelihood L of all the potential vacant parking spaces in the vicinity.

[0078] <How to use the dispatch flag> Meanwhile, non-parked vehicle V exits the parking space. A If the exit flag is assigned to the non-parked vehicle V, the likelihood calculation unit 15 of this embodiment calculates the likelihood of the non-parked vehicle V A The system changes the parking space from which a vehicle has exited to a candidate for an available parking space, and then sets the likelihood L of that candidate for an available parking space to, for example, 1.0. Furthermore, since it is predicted that a vehicle that has just exited will not re-enter the parking space, the likelihood calculation unit 15 calculates the likelihood of that non-parked vehicle V A Even in an environment like Figure 15, where there are potential empty parking spaces in the direction of travel, the likelihood L of the potential empty parking spaces is not adjusted.

[0079] As a result, the vehicle driving control unit 16 in this embodiment controls the non-parked vehicle V A It can generate appropriate driving routes depending on the situation, such as a driving route that parks in a parking space immediately after exiting the parking lot. [Examples]

[0080] Next, Embodiment 4 of the present invention will be described using Figures 16 and 17. Note that repetitive explanations of points common to the above embodiments will be omitted.

[0081] Figure 16 is a diagram illustrating the details of the processing unit 10 of the parking assist device 1 in this embodiment. As is obvious from comparing Figure 2 and Figure 16, the HMI device group 5 is not connected to the input side of the processing unit 10 in Embodiment 1, whereas the HMI device group 5 is connected to the input side of the processing unit 10 in this embodiment. Therefore, the processing unit 10 in this embodiment is equipped with the function of receiving commands input to the HMI device group 5 by the occupants of the vehicle V0.

[0082] Figure 17 is a flowchart of the processes performed by the vehicle driving control unit 16 in this embodiment. As is obvious from comparing Figure 9 and Figure 17, the vehicle driving control unit 16 in this embodiment can perform steps S16b1 to S16b7 in addition to the processes of steps S16a to S16f in Embodiment 1. The details of each process of steps S16b1 to S16b7, which are unique to this embodiment, will be described in order below.

[0083] In step S16b1, the vehicle driving control unit 16 determines whether a flag waiting for occupant response is set in the vehicle control data group D7. If the requirement is met, the unit proceeds to step S16b2; otherwise, it proceeds to step S16b5.

[0084] In step S16b2, the vehicle driving control unit 16 determines whether it has received a response from the occupant via the HMI device group 5. If the requirements are met, it proceeds to step S16b3; otherwise, it terminates the process.

[0085] If the requirements of step S16b2 are met, the vehicle driving control unit 16 deletes the response waiting flag in step S16b3 and generates a driving route that reflects the occupant selection in step S16b4.

[0086] Furthermore, if the requirements of step S16b1 are not met, the vehicle driving control unit 16 determines in step S16b5 whether the difference between the sum of the likelihoods of the available parking space candidates and the threshold is less than or equal to a certain value. If the requirements are met, the process proceeds to step S16b6; otherwise, the process proceeds to step S16c.

[0087] If the requirements of step S16b5 are met, the vehicle driving control unit 16 sets a response waiting flag in step S16b6 and generates a driving path for a temporary stop in step S16b7.

[0088] By adding these processes, the parking assistance device 1 of this embodiment can, when the difference between the likelihood (threshold) of the nearest available parking space candidate (reference point) of the vehicle V0 and the sum of the likelihoods of available parking spaces at a distance is less than or equal to a certain value, inquire with the occupant via an HMI device such as a touch panel display whether to park at the reference point or pass by the reference point, and implement vehicle control according to the occupant's response. Furthermore, it can wait for the occupant's response input while the vehicle V0 is temporarily stopped in a safe location. [Explanation of symbols]

[0089] 100 Vehicle Systems 1. Parking assist system 10 Processing Unit 11 Information acquisition department 12. Other Vehicle Information Acquisition Unit 13. Identifying the parking lines 14. Parking space width specification area 15 Likelihood Calculation Unit 16. Vehicle running control unit 17 Information Output Unit 18. Estimation unit for dispatch time level 19. Identification of Non-Parked Vehicles 20 Memory section 30 Communications Department 2 Vehicle sensor group 3. Group of external sensors 3a Lidar 4 Actuator Groups 5 HMI equipment group 5a Input HMI device 5b HMI device for output D1 Vehicle Data Set D2 Sensor Recognition Data Set D3 Other Vehicle Data Group D4 Parking Vehicle Queue Dataset D5 Measurement Error Data Set D6 Data set of potential vacant parking spaces D7 Vehicle control data set D8 Departure Vehicle Data Group D9 Non-parked vehicle data set

Claims

1. A parking assist device that assists in parking a vehicle, A vehicle information acquisition unit acquires information about other vehicles in the vicinity of the vehicle based on information obtained from an on-board external sensor, A parking vehicle row identification unit identifies parked vehicles and parking vehicle rows around the vehicle based on the information of the aforementioned other vehicles, A likelihood calculation unit calculates the likelihood that the space between adjacent parked vehicles is an empty parking space, based on the distance between adjacent parked vehicles and the width of the parking space within the aforementioned row of parked vehicles. The system comprises a vehicle driving control unit that controls the driving of the vehicle based on the likelihood, The parking assistance device is characterized in that the vehicle driving control unit generates either a driving route that passes through the reference point or a driving route that parks at the reference point, based on a comparison result between the first likelihood at a predetermined reference point within the sensing range of the external sensor and the second likelihood at a parking space further away from the reference point.

2. In the parking assistance device according to claim 1, A parking assistance device further comprising an HMI device that notifies the occupant of the driving route generated by the vehicle driving control unit and allows the occupant to choose whether or not to adopt the notified driving route.

3. In the parking assistance device according to claim 1, The parking assistance device is characterized in that the vehicle driving control unit controls the driving of the vehicle according to the generated driving path.

4. In the parking assistance device according to claim 1, A parking assistance device further comprising a parking space width determination unit that determines the width of the parking space based on the positional relationship of each parked vehicle determined by the parking vehicle row determination unit.

5. In the parking assistance device according to claim 1, The likelihood calculation unit is characterized in that, when the distance between adjacent parked vehicles is greater than the width of the parking space, it calculates the likelihood that the space between adjacent parked vehicles is an empty parking space, taking into account the positional error when the positions of both parked vehicles are detected by the external sensor.

6. In the parking assistance device according to claim 5, The aforementioned vehicle running control unit, If the first likelihood at the aforementioned reference point is greater than the second likelihood, which is the sum of the likelihoods at parking spaces further away from the reference point, a driving route that parks at the aforementioned reference point is generated. A parking assistance device characterized by generating a driving path that passes through the reference point when the first likelihood is smaller than the second likelihood.

7. In the parking assistance device according to claim 1, The parking assistance device is characterized in that the reference point is located in front of the vehicle and is between adjacent parked vehicles at a distance greater than the width of the parking space.

8. In the parking assistance device according to claim 1, The parking assistance device is characterized in that the aforementioned reference point is a parking space where a parked vehicle that is estimated to be about to leave the parking area is parked.

9. In the parking assistance device according to claim 8, A parking assistance device further comprising a vehicle departure time level estimation unit that estimates a parked vehicle that is about to leave the parking space if the engine is started, the headlights are on, or the turn signals are flashing.

10. In the parking assistance device according to claim 1, The vehicle further includes a non-parked vehicle identification unit that identifies non-parked vehicles in the vicinity of the vehicle based on information about the aforementioned other vehicles. If the non-parked vehicle identification unit generates a parking schedule flag indicating that the non-parked vehicle is scheduled to be parked, The likelihood calculation unit is characterized by reducing the likelihood of parking in the nearest parking space to the non-parked vehicle.

11. In the parking assistance device according to claim 1, The vehicle further includes a non-parked vehicle identification unit that identifies non-parked vehicles in the vicinity of the vehicle based on information about the aforementioned other vehicles. If the non-parked vehicle identification unit generates an exit flag indicating that the non-parked vehicle has left the parking lot, The likelihood calculation unit is characterized by increasing the likelihood of the parking space from which the non-parked vehicle exited, while not adjusting the initial likelihood of the parking space in the direction of travel of the non-parked vehicle.

12. A parking assistance method, which is performed by an ECU to assist in parking a vehicle, A step of acquiring information on other vehicles in the vicinity of the vehicle, based on information obtained from an on-board external sensor, A parking vehicle row identification step to identify parked vehicles and parking vehicle rows around the vehicle based on the information of the other vehicles, A likelihood calculation step of calculating the likelihood that the space between adjacent parked vehicles is an empty parking space, based on the distance between adjacent parked vehicles and the width of the parking space within the row of parked vehicles, The vehicle driving control step includes controlling the driving of the vehicle based on the likelihood, The parking assistance method is characterized in that, in the vehicle driving control step, a driving path that passes through the reference point or a driving path that parks at the reference point is generated according to the result of comparing the first likelihood at a predetermined reference point within the sensing range of the external sensor with the second likelihood at a parking space further away from the reference point.

13. In the parking assistance device according to claim 12, The parking assistance method is characterized in that, in the likelihood calculation step, when the distance between adjacent parked vehicles is greater than the width of the parking space, the likelihood that the space between adjacent parked vehicles is an empty parking space is calculated by taking into account the positional error when the positions of both parked vehicles are detected by the external sensor.

Images