Azimuth angle acquisition device for automotive vehicle, azimuth angle acquisition system, and azimuth angle acquisition method
The azimuth angle acquisition device and system improve azimuth angle accuracy in self-propelled vehicles by integrating external and internal state detection, enhancing navigation and control precision.
Patent Information
- Application Number
- JP2023083792
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-05-22
- Publication Date
- 2024-12-04
- Estimated Expiration
- 2043-05-22
AI Technical Summary
Existing self-propelled vehicles under autonomous or remote control face challenges in accurately acquiring the azimuth angle, which affects their ability to follow desired routes.
An azimuth angle acquisition device and system that utilize both external and internal state detection units, such as cameras and yaw rate sensors, to determine the vehicle's azimuth by adjusting priority based on the vehicle's state and the certainty of external state detection, combining external and internal azimuth angles for improved accuracy.
Enhances the accuracy of azimuth angle acquisition, enabling precise self-propelled vehicle navigation and control, particularly in stopped, turning, and straight-ahead states, reducing errors and improving the vehicle's ability to reach its destination accurately.
Smart Images

Figure 2024167599000001_ABST
Abstract
Description
[Technical field]
[0001] The present disclosure relates to techniques for obtaining the azimuth angle of a motor vehicle. [Background technology]
[0002] A technology has been proposed for running vehicles by autonomous control or remote control within a vehicle manufacturing system (for example, Patent Document 1). [Prior art documents] [Patent documents]
[0003] [Patent Document 1] JP 2017-538619 A
[0004] In a self-propelled vehicle that runs by autonomous control or remote control, the vehicle's running is controlled using the vehicle's position coordinates and azimuth angle. However, there is a problem that the azimuth angle cannot be acquired with the desired acquisition accuracy. If the acquisition accuracy of the azimuth angle is low, the self-propelled vehicle cannot run along the desired route. Summary of the Invention [Problem to be solved by the invention]
[0005] Therefore, there is a demand for improving the accuracy of acquiring the azimuth angle of a mobile vehicle. [Means for solving the problem]
[0006] The present disclosure can be realized in the following aspects.
[0007] A first aspect provides an azimuth angle acquisition device for a self-propelled vehicle. The azimuth angle acquisition device according to the first aspect includes an acquisition unit that acquires an external state, which is a motion state of the vehicle detected from the external world, and an internal state, which is a motion state of the vehicle detected from the internal world, an azimuth angle calculation unit that acquires an external azimuth angle using the acquired external state and calculates an internal azimuth angle using the acquired internal state, and an azimuth angle determination unit that determines a vehicle azimuth angle used for self-propelling of the vehicle using the calculated external azimuth angle and internal azimuth angle.
[0008] According to the azimuth angle acquisition device of the first aspect, it is possible to improve the accuracy of acquiring the azimuth angle of a self-propelled vehicle.
[0009] In the azimuth angle acquisition device according to the first aspect, the azimuth angle determination unit may determine the vehicle azimuth angle by changing the priority of the external azimuth angle and the internal azimuth angle according to the state of the vehicle and the state of the external state detection unit that detects the external state. In this case, the accuracy of acquiring the azimuth angle of the self-driving vehicle can be further improved according to the state of the vehicle and the state of the external state detection unit.
[0010] In the azimuth angle acquisition device according to the first embodiment, the azimuth angle determination unit may determine the vehicle azimuth angle by giving a higher priority to the external azimuth angle when the vehicle is in a stopped state. In this case, the use of the external azimuth angle can further improve the accuracy of acquiring the azimuth angle of the self-propelled vehicle.
[0011] In the azimuth angle acquisition device according to the first aspect, the azimuth angle determination unit may determine the vehicle azimuth angle by giving a higher priority to the inner azimuth angle when the motion state of the vehicle is a turning state. In this case, by using the inner azimuth angle, it is possible to further improve the acquisition accuracy of the azimuth angle of the self-propelled vehicle.
[0012] In the azimuth angle acquisition device according to the first aspect, the azimuth angle determination unit may correct the internal azimuth angle using the external azimuth angle when the motion state of the vehicle is in a straight-line state and the certainty of the external state detection unit is equal to or greater than a predetermined threshold, and may determine the vehicle azimuth angle by giving a higher priority to the internal azimuth angle when the certainty of the external state detection unit is less than a predetermined threshold. In this case, the accuracy of acquiring the azimuth angle of the self-propelled vehicle can be further improved by correcting the internal azimuth angle using the external azimuth angle according to the certainty of the external state detection unit.
[0013] In the azimuth angle acquisition device according to the first aspect, the external state detection unit may be an imaging device disposed in the external environment of the vehicle, and the azimuth angle calculation unit may calculate the external azimuth angle of the vehicle using an image captured by the imaging device. In this case, the external azimuth angle can be directly calculated using the image, so that the accuracy of acquiring the azimuth angle of the self-driving vehicle can be further improved.
[0014] In the azimuth angle acquisition device according to the first aspect, the internal state detection unit may be a yaw rate sensor provided in the vehicle, and the azimuth angle calculation unit may calculate the internal azimuth angle of the vehicle using the yaw rate acquired by the yaw rate sensor. In this case, the behavior of the vehicle can be directly detected to calculate the internal azimuth angle, so that the accuracy of acquiring the azimuth angle of the self-driving vehicle can be further improved.
[0015] A second aspect provides an azimuth angle acquisition system for a self-propelled vehicle. The azimuth angle acquisition system according to the second aspect includes an external state detection unit disposed in the external world of the vehicle to detect the motion state of the vehicle from the external world, an internal state detection unit mounted on the vehicle to detect the motion state of the vehicle from the internal world, an azimuth angle acquisition device comprising: an azimuth angle calculation unit that calculates an external azimuth angle using an external state of the vehicle detected by the external state detection unit, and calculates an internal azimuth angle using an internal state of the vehicle detected by the internal state detection unit; and an azimuth angle determination unit that determines a vehicle azimuth angle used for self-propelling of the vehicle using the calculated external azimuth angle and internal azimuth angle.
[0016] According to the azimuth angle acquisition system of the second aspect, it is possible to improve the accuracy of acquiring the azimuth angle of a self-driving vehicle. Also, the azimuth angle acquisition system of the second aspect can be realized in various aspects in the same manner as the azimuth angle acquisition device of the first aspect.
[0017] A third aspect provides an azimuth angle acquisition method for a self-propelled vehicle. The azimuth angle acquisition method according to the third aspect includes acquiring an external state, which is a motion state of the vehicle detected from the external world, and an internal state, which is a motion state of the vehicle detected from the internal world, acquiring an external azimuth angle using the acquired external state, calculating an internal azimuth angle using the acquired internal state, and determining a vehicle azimuth angle used for self-propelling of the vehicle using the calculated external azimuth angle and internal azimuth angle.
[0018] According to the azimuth angle acquisition method of the third aspect, it is possible to improve the accuracy of acquiring the azimuth angle of a self-driving vehicle. The azimuth angle acquisition method of the third aspect can be realized in various aspects in the same manner as the azimuth angle acquisition device of the first aspect. The azimuth angle acquisition method of the third aspect can also be realized as an azimuth angle acquisition program or a computer-readable recording medium recording the program. The determination of the vehicle azimuth angle in the first to third aspects may be performed using AI, i.e., a learning model. [Brief description of the drawings]
[0019] [Figure 1] 1 is an explanatory diagram showing a schematic configuration of a vehicle azimuth angle acquisition system according to a first embodiment; [Diagram 2]1 is a block diagram showing an example of the internal configuration of a vehicle azimuth angle acquisition device according to a first embodiment. [Diagram 3] 1 is a block diagram showing an example of an internal configuration of a vehicle according to a first embodiment. [Figure 4] 4 is a flowchart showing a vehicle azimuth angle acquisition process flow executed by the vehicle azimuth angle acquisition device according to the first embodiment. [Diagram 5] FIG. 2 is an explanatory diagram showing an example of a running state of a vehicle. [Figure 6] FIG. 4 is an explanatory diagram showing the correspondence relationship between a vehicle azimuth angle and a movement vector. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0020] An azimuth angle acquisition device, an azimuth angle acquisition system, and an azimuth angle acquisition method for a self-propelled vehicle according to the present disclosure will be described below based on several embodiments.
[0021] First embodiment: 1, a vehicle azimuth angle acquisition system 500 according to the first embodiment includes a vehicle azimuth angle acquisition device 10, a self-propelled vehicle 30, an external state detection unit 21, and an internal state detection unit 31. The vehicle azimuth angle acquisition device 10 is a so-called server computer, and determines and acquires a vehicle azimuth angle that can be used as a parameter indicating the direction of the vehicle 30.
[0022] The vehicle azimuth angle acquisition device 10 is connected to an external state detection unit 21 for detecting the external state of the vehicle 30 by wire or wirelessly. The vehicle azimuth angle acquisition device 10 acquires the external state of the vehicle 30 via the external state detection unit 21. The external state of the vehicle is the motion state of the vehicle detected from the outside of the vehicle 30, that is, from outside the vehicle. The motion state of the vehicle means, for example, the driving state of the vehicle, such as the vehicle being stopped, running, accelerating, decelerating, going straight, or turning. The vehicle azimuth angle acquisition device 10 is connected by wire to an access point 22 for transmitting and receiving signals between the vehicle 30 and the vehicle 30 by wireless communication W / C. The wireless communication is realized, for example, by a wireless local network (LAN) conforming to the IEEE802.11 standard or a wireless connection via Bluetooth.
[0023] In this embodiment, the self-propelled vehicle 30 is a vehicle equipped with an internal combustion engine, a vehicle equipped with a motor, or a hybrid vehicle equipped with an internal combustion engine and a motor as a power source. A vehicle equipped with a motor can be equipped with at least one of a secondary battery and a fuel cell as a power source. The vehicle 30 is equipped with an internal state detection unit 31 for detecting the internal state of the vehicle 30. The vehicle azimuth angle acquisition device 10 acquires the internal state detected by the internal state detection unit 31 and transmitted by wireless communication via the access point 22. The internal state of the vehicle is the vehicle's motion state detected in the internal world of the vehicle 30, i.e., inside the vehicle.
[0024] In this embodiment, a plurality of external environment state detection units 21 are arranged on a route along which the vehicle 30 travels (see FIG. 5). The external environment state detection unit 21 is, for example, at least one of a camera and a Lidar. In general, a camera is an imaging device equipped with an imaging element such as a CCD or an imaging element array, and outputs outer shape information or shape information of an object as still image data or moving image data by receiving visible light. A Lidar generally emits infrared laser light, and measures the distance to the object by receiving the reflected light reflected by the target with a light receiving element. The reflected light reflected by the target is detected as a detection point cloud corresponding to the target. The external environment state detection unit 21 outputs still image data or moving image data, or detection point cloud data as a parameter indicating the external environment state. In this embodiment, the external environment state detection unit 21 only needs to detect information for identifying the outer shape of the vehicle 30 and output it as information indicating the external environment state, as described above. As described below, the vehicle azimuth angle acquisition device 10 determines the driving state of the vehicle 30 by using information for identifying the external shape detected by the external state detection unit 21. The determination of the driving state of the vehicle 30 performed by the vehicle azimuth angle acquisition device 10 may be performed by the external state detection unit 21. In this case, the external state detection unit 21 transmits information indicating the determined driving state to the vehicle azimuth angle acquisition device 10. When a camera and Lidar are used, the image data and the detection point cloud data may be fused by fusion processing and used.
[0025] In this embodiment, the internal state detection unit 31 is used to detect the azimuth angle of the vehicle 30 in the internal world of the vehicle. For example, a yaw rate sensor is used as the internal state detection unit 31. In the following, the internal state detection unit 31 will be mainly described using a yaw rate sensor. The yaw rate sensor outputs the angular velocity around the vehicle yaw axis when the vehicle turns, that is, the yaw rate as a parameter indicating the internal state. The internal state detection unit 31 is mounted on the vehicle 30 and can directly detect the behavior of the vehicle 30. A gyro sensor or a multi-axis acceleration sensor can be used as long as it is a sensor that can detect at least the turning state.
[0026] As shown in FIG. 2, the vehicle azimuth angle acquisition device 10 includes a central processing unit (CPU) 101, a memory 102 as a storage unit, an input / output interface 103 as an acquisition unit, and a clock generator (not shown). The CPU 101, the memory 102, the input / output interface 103, and the clock generator are connected to each other via an internal bus 104 so as to be able to communicate in both directions. The memory 102 includes a memory, such as a ROM, that stores the vehicle azimuth angle acquisition program Pr1 in a non-volatile and read-only manner, and a memory, such as a RAM, that can be read and written by the CPU 101. The vehicle azimuth angle acquisition program Pr1 is executed by the CPU 101 to calculate an external azimuth angle using the acquired external state, calculate an internal azimuth angle using the acquired internal state, and determine a vehicle azimuth angle used for self-propelling of the vehicle 30 using the external azimuth angle and the internal azimuth angle. The non-volatile and readable / writable area of the memory 102 includes a certainty storage area 102a that stores the certainty of the external state detection unit 21. The CPU 101, i.e., the vehicle azimuth angle acquisition device 10, functions as an azimuth angle acquisition unit and an azimuth angle determination unit by expanding the vehicle azimuth angle acquisition program Pr1 stored in the memory 102 into a readable / writable memory and executing it. The CPU 101 may be a single CPU, or may be a plurality of CPUs that execute each program, or may be a multitask or multithread type CPU that can execute a plurality of programs simultaneously. The vehicle azimuth angle acquisition device 10 can function as a remote control device that remotely controls the vehicle 30 by being provided with a remote control program.
[0027] The input / output interface 103 is connected to the camera 21 and the access point 22 as an external state detection unit via signal lines. An imaging control signal instructing imaging processing for object detection is transmitted from the input / output interface 103 to the camera 21, and an imaging signal indicating an image captured as a detection result is transmitted from the camera 21 to the input / output interface 103. When the vehicle azimuth angle acquisition device 10 also functions as a remote control device, a driving control signal for self-driving the vehicle 30 is transmitted from the input / output interface 103 to the access point 22. The driving control signal is, for example, a control signal transmitted to the driving control device 300 provided in the vehicle 30, and is a control signal for instructing a required accelerator opening or a required torque value, a required braking amount, and a required steering angle. The driving control device 300 controls each actuator according to the driving control signal, thereby realizing driving control including acceleration / deceleration and cruising in the vehicle 30, stopping control of the vehicle 30, and turning control including right and left turns. Vehicle information indicating the state of the vehicle is transmitted from the access point 22 to the input / output interface 103. The vehicle information includes, for example, driving information indicating driving conditions detected in the vehicle 30, such as yaw rate, steering angle, speed, and acceleration / deceleration, and information such as whether the wipers are operating and whether the headlights are on. In addition, environmental information indicating environmental conditions outside the vehicle, such as weather, temperature, and brightness, may be input to the input / output interface 103. The environmental information may be acquired by various detection devices arranged on the vehicle's route, or may be acquired via a network as information on an area including the vehicle's route.
[0028] As shown in Fig. 3, the vehicle 30 includes a driving control device 300 including a central processing unit (CPU) 301, a memory 302 as a storage unit, an input / output interface 303, and a clock generator (not shown). The CPU 301, the memory 302, the input / output interface 303, and the clock generator are connected to each other via an internal bus 304 so as to be able to communicate bidirectionally. The memory 302 includes a memory, such as a ROM, that stores a remote control compatible program Pr2 in a non-volatile and read-only manner, and a memory, such as a RAM, that can be read and written by the CPU 301. The remote control compatible program Pr2 controls the driving state of the vehicle 30 in response to a driving control signal received from the remote control device.
[0029] The input / output interface 303 is connected to the yaw rate sensor 31, the wireless transceiver 32, and the vehicle driving control actuator 33 via signal lines. The yaw rate sensor 31 transmits a yaw rate as a detection result to the input / output interface 303. The input / output interface 303 transmits driving information including the yaw rate received from the yaw rate sensor 31 and vehicle information to the wireless transceiver 32. The wireless transceiver 32 transmits a driving control signal to the input / output interface 303. The input / output interface 303 transmits a control signal for indicating a required accelerator opening or a required torque value, a required braking amount, and a required steering angle to the vehicle driving control actuator 33. The driving control device 300 includes a fuel injection device in an internal combustion engine according to the required accelerator opening, a motor control device in a motor according to the required torque value, a brake pad actuator in a braking device according to the required braking amount, and a steering actuator in a steering device according to the required steering angle that controls the steering angle of the steering wheel. The brake actuator controls hydraulic pressure by a motor to adjust the operation amount of the brake pad, i.e., the braking amount / braking force. The steering actuator adjusts the steering angle of the steering wheel by controlling a pinion gear engaged with a rack gear connected to the steering wheel by a motor. The wireless transceiver 32 may be detachable. The detachable wireless transceiver 32 only needs to have a configuration capable of transmitting a driving control signal to the vehicle driving control actuator 33 via an input / output interface 303 provided in the vehicle 30 when attached to an external communication connection port of the vehicle 30. In other words, when the wireless transceiver 32 is detached, the vehicle 30 becomes a normal vehicle that is not controlled by remote control and drives by operation at the vehicle 30. Furthermore, instead of the memory 302, the wireless transceiver 32 may have a remote control compatible program Pr2.
[0030] With reference to Fig. 4 and Fig. 5, the vehicle azimuth angle acquisition process executed by the vehicle azimuth angle acquisition device 10 according to this embodiment will be described. The processing routine shown in Fig. 4 is repeatedly executed at predetermined time intervals, for example, every few milliseconds to every few seconds. The CPU 101 acquires the vehicle azimuth angle as an azimuth angle calculation unit and an azimuth angle determination unit by executing the vehicle azimuth angle acquisition program Pr1. Fig. 5 shows the positional relationship between the vehicle 30 and the external state detection unit 21, and also shows the transition of the vehicle's running state from a stationary state to a straight running state Po1 in which the vehicle 30 is in a state including starting, to a turning state Po2 in which the vehicle 30 is in a straight running state Po3.
[0031] The CPU 101 acquires imaging information from the external state detection unit 21, i.e., the camera 21, via the input / output interface 103 (step S100). The imaging information acquired from the camera 21 is preferably a moving image composed of a plurality of frame images captured at a frame rate in the range of, for example, 30 fps to 60 fps. Alternatively, it may be a still image captured every 1 s. When the displacement of the vehicle 30, for example, the moving speed or turning speed, is low, it may be a still image. The CPU 101 calculates the external azimuth angle and the likelihood of the external state detection unit 21 using the imaging information (step S102). Specifically, the CPU 101 calculates a movement vector related to a predetermined feature point (point of interest) of the vehicle 30 using the plurality of frame images, and calculates the azimuth angle indicated by the calculated movement vector as the external azimuth angle. The predetermined feature of the vehicle 30 may be, for example, a group of multiple feature points defining an area representing the vehicle 30 or a center of gravity of an area extracted as an area representing the vehicle 30, or may be a portion of the vehicle 30 where the luminance value changes greatly, for example, an edge portion on the rear side of the vehicle. The image processing technology for calculating / extracting a movement vector using multiple frame images is a known technology, so a description thereof will be omitted. When calculating the external azimuth angle, a rectangular area representing the vehicle 30 may be extracted from each frame image by pattern matching, the azimuth angle of the main direction, which is the longitudinal direction of the extracted rectangular area, may be specified, and the azimuth angle of the specified main direction may be acquired as the external azimuth angle. Note that, when calculating the external azimuth angle, the position coordinates of the vehicle 30 may also be specified.
[0032] The external state detection unit 21, that is, the likelihood of the external azimuth angle, can be calculated, for example, in the following manner. Pattern matching processing is performed using AI for some or all of the frame images acquired from the camera 21, and the similarity (degree of matching) between the model of the vehicle 30 and the vehicle included in the frame images is scored. If the obtained similarity is high, the likelihood is high, and if the obtained similarity is low, the likelihood is low. The pattern matching processing using AI is performed using, for example, supervised or unsupervised learning that extracts vehicles from images by pattern matching, or a machine learning model that has been learned by reinforcement learning. For example, a neural network or a support vector machine (SVM) may be used as the machine learning model. Deep learning may be used when learning the reinforcement learning model. Image processing is performed on some or all of the frame images acquired from the camera 21 to calculate image quality parameters such as brightness and sharpness. Brightness can be calculated, for example, as an average pixel value obtained by dividing a pixel value (luminance value) for each pixel forming the frame image by the total number of pixels. When brightness is high (bright), the likelihood is high, and when brightness is low (dark), the likelihood is low. When brightness is high, the range of pixel values becomes wide, so the brightness difference in the outline showing the vehicle body becomes large, and the vehicle body can be accurately identified / extracted. Sharpness can be calculated, for example, as the proportion of high-frequency components in the spectral distribution obtained by performing a fast Fourier transform on the frame image. When sharpness is high, the likelihood may be high, and when sharpness is low, the likelihood may be low. When sharpness is high, the outline showing the vehicle body in the frame image is clear, and the vehicle body can be accurately identified / extracted. When the above image quality parameters are used, a value obtained by normalizing the value of the image quality parameter within the range width that the image quality parameter can take can be used as a value indicating the likelihood. For example, the time and weather may be used as the external environmental conditions of the vehicle 30. The time can be used as an index of the brightness of the external environment, and the probability is set high for times between morning and evening when sufficient brightness is expected, and low for other times. The weather obtained from the meteorological information is used as an index of the brightness of the external environment, and can also be used as an index of the clarity of a frame image that is affected by yellow sand and fog. If yellow sand and fog are predicted by the meteorological information, the probability is set low. As an external environmental condition of the vehicle 30, for example, as shown in FIG. 5, the positional relationship between the external environment detection unit 21 and the vehicle 30 can be used. When the vehicle 30 is close to the installation position of the external environment detection unit 21, the vehicle 30 is expected to be captured with an appropriate angle of view, and since more pixels in the frame image indicate the vehicle area, the vehicle 30 can be displayed with higher resolution. The closer the vehicle 30 is to the installation position of the external environment detection unit 21, the higher the likelihood is set. The distance or normalized distance between the vehicle 30 and the external environment detection unit 21 is obtained as a value indicating the likelihood. As a result, the vehicle body can be accurately determined / extracted from the frame image. The installation position of the external environment detection unit 21 is known, and the position information of the vehicle 30 relative to the external environment detection unit 21 is obtained as the distance of the vehicle 30 relative to the external environment detection unit 21 measured using Lidar or radar. Alternatively, the positional relationship between the external environment condition detection unit 21 and the vehicle 30 can be obtained by using the position coordinates of the vehicle 30 measured by a GNSS (Global Navigation Satellite System) equipped in the vehicle 30 and the position coordinates of the external environment condition detection unit 21. The above various kinds of certainties may be used alone or in any combination. When combining, the certainties in each mode may be simply added, or may be added by applying a predetermined weight / coefficient, or a normalization process may be performed in which the maximum value is 1 and the minimum value is 0. The calculated certainties may be stored in the certainty storage area 102a in the memory 102 so as to be used repeatedly for multiple vehicles 30.
[0033] The CPU 101 acquires the yaw rate from the yaw rate sensor 31 (step S104). The CPU 101 calculates the internal azimuth angle using the acquired yaw rate (step S106). As is well known, the calculation of the vehicle's azimuth angle using the yaw rate is performed by an integration operation using the yaw rate acquired earlier than this time. The CPU 101 determines whether the vehicle 30 is in a stopped state (step S108, Po1 in FIG. 5). Specifically, the vehicle speed included in the received driving information transmitted from the vehicle 30 is used. When the CPU 101 determines that the vehicle 30 is in a stopped state, that is, when it determines that the vehicle speed = 0 km / h (step S108: Yes), it sets the priority / priority level of the external azimuth angle high, calculates the vehicle azimuth angle of the vehicle 30 (step S110), and ends this processing routine. The vehicle azimuth angle is the azimuth angle used for the autonomous driving control of the vehicle 30 and is calculated by weighted average, for example, vehicle azimuth angle = (wi × internal azimuth angle + wo × external azimuth angle) / 2. As a mode of setting the priority of the external azimuth angle high, in addition to the mode of using only the external azimuth angle (wi = 0, Wo = 1: total weight = 1), the weighting for the internal azimuth angle can be set to 0 < wi < 0.2, and the weighting for the external azimuth angle can be set to wo = 1 - wi. Generally, the internal azimuth angle requires the previous value / initial value for calculation by integration operation. On the other hand, for example, when the vehicle 30 stops after straight driving during the running of the vehicle 30, the previous internal azimuth angle acquired immediately before the vehicle 30 stops can be used.
[0034] When the CPU 101 determines that the vehicle 30 is not in a stopped state, that is, when the CPU 101 determines that the vehicle speed is not equal to 0 km / h (step S108: No), the CPU 101 determines whether the vehicle 30 is in a turning state (step S112, Po2 in FIG. 5). For example, the steering angle included in the travel information transmitted and received from the vehicle 30 is used to determine whether the vehicle 30 is in a turning state. In this embodiment, the steering angle is the operation angle of the steering wheel by the steering wheel operation of the vehicle 30, and the actual turning angle of the steering wheel is the steering angle. If the vehicle 30 is not provided with a sensor for detecting the steering angle, the steering angle detected and transmitted by the vehicle 30 may be used. When the acquired steering angle or steering angle is equal to or greater than the judged steering angle for a predetermined time, the CPU 101 determines that the vehicle 30 is in a turning state. The predetermined time is, for example, 2 to 5 seconds in order to eliminate temporary steering operations that do not involve turning. The judgment turning angle or judgment steering angle is, for example, 2 to 5 degrees in order to eliminate steering angle adjustment for maintaining straight ahead. In addition, whether the vehicle 30 is turning may be judged by using the output of the yaw rate sensor 31 and checking whether a rotational moment about the yaw axis of the vehicle 30 is generated for a predetermined time period equal to or greater than a predetermined judgment value.
[0035] When the CPU 101 determines that the vehicle 30 is in a turning state (step S112: Yes), it sets a high priority for the inner azimuth angle, calculates the vehicle azimuth angle of the vehicle 30 (step S114), and ends this processing routine. As a mode of setting a high priority for the inner azimuth angle, in addition to a mode using only the inner azimuth angle (wi = 1, Wo = 0: total weight = 1), it is also possible to set the weighting for the outer azimuth angle to 0 < wo < 0.2 and set the weighting for the inner azimuth angle to wi = 1 - wo. In the present embodiment, a movement vector is used for calculating the outer azimuth angle. As shown in FIG. 6, the movement vector Oaz indicates the movement of the feature point of the vehicle 30 and is a physical quantity different from the azimuth angle Iaz of the vehicle 30. In particular, when the vehicle 30 is in a turning state, the movement vector Oaz and the azimuth angle Iaz tend to deviate. Also, generally, the signal-to-noise ratio of the signal indicating the movement vector Oaz is low, and a smoothing filter such as a low-pass filter is applied to improve the signal-to-noise ratio. In a turning state where the displacement of the azimuth angle is large, a deviation from the true value may occur more significantly due to phase delay. On the other hand, the movement vector Oaz is useful as a parameter for estimating whether the vehicle 30 is in a straight-ahead state or a turning state. Further, the calculated inner azimuth angle has the characteristic that an integration error occurs due to the accumulation of sensor tolerances. Therefore, by taking the outer azimuth angle into account, it is possible to obtain a vehicle azimuth angle with reduced integration error. Furthermore, when specifying the outer azimuth angle of the vehicle 30 by pattern matching processing for each frame image, since the outer azimuth angle becomes the same physical quantity as the inner azimuth angle, by taking the outer azimuth angle into account, it is possible to obtain a vehicle azimuth angle with further improved accuracy.
[0036] When the CPU 101 determines that the vehicle 30 is not in a turning state (step S112: No: Po1, Po3 in FIG. 5), it determines whether the likelihood of the external azimuth angle is equal to or greater than a predetermined threshold value (step S116). The likelihood can take on various forms of values as described above. As the predetermined threshold value, a threshold value that can determine that the external azimuth angle is suitable for correcting the internal azimuth angle in the next step S118 is used. For example, regardless of the form of likelihood, a threshold value that can obtain a likelihood of more than 70%, preferably more than 80%, and more preferably more than 90% can be used.
[0037] When the CPU 101 determines that the reliability of the external azimuth angle is equal to or greater than a predetermined threshold value (step S116: Yes), it corrects the internal azimuth angle using the external azimuth angle to calculate the vehicle azimuth angle (step S118), and ends this processing routine. That is, when the traveling state of the vehicle 30 is a straight traveling state in which the external azimuth angle can indicate the vehicle azimuth angle, and the reliability of the external azimuth angle is equal to or greater than a predetermined threshold value, the internal azimuth angle is corrected. The correction of the internal azimuth angle using the external azimuth angle is performed at a timing when the distance between the vehicle 30 and the external state detection unit 21 is short, that is, at a timing when the external state detection unit 21 that detects the vehicle 30 is switched. This is because the reliability of the external azimuth angle is high at this timing, as described above. Examples of the correction mode include a mode in which the internal azimuth angle is replaced by the external azimuth angle at one timing, and a mode in which the internal azimuth angle is gradually replaced by applying a moving average filter process to a number of sampling values of the external azimuth angle, for example, 10 calculation values. Although a smoother correction can be realized by increasing the number of calculation values to which the moving average filter is applied, an appropriate smooth correction can be realized with a sampling number of about 10 to 20 times. Furthermore, as a specific method of replacement, the internal azimuth angle may be replaced by a weighted average value of the internal azimuth angle and the external azimuth angle. As a weighting value, for example, 70 to 100% can be used for the external azimuth angle, and a value obtained by subtracting a weighting value for the external azimuth angle from 100% can be used for the internal azimuth angle.
[0038] When the CPU 101 determines that the certainty of the external azimuth angle is not equal to or greater than a predetermined threshold value (step S116: No), the CPU 101 does not correct the internal azimuth angle using the external azimuth angle, and proceeds to step S114 to end this processing routine. This is because when the certainty of the external azimuth angle is low, correcting the internal azimuth angle using the external azimuth angle may result in an erroneous correction of the internal azimuth angle.
[0039] According to the vehicle azimuth angle acquisition device 10 according to the first embodiment described above, an external azimuth angle is calculated using an external state, which is a motion state of the vehicle 30 detected from the outside, an internal azimuth angle is calculated using an internal state, which is a motion state of the vehicle 30 detected from the inside, and a vehicle azimuth angle used for self-propelling of the vehicle 30 is determined using the calculated external azimuth angle and internal azimuth angle. It is possible to improve the accuracy of acquiring the azimuth angle of the self-propelled vehicle.
[0040] In general, the external azimuth angle, which is an azimuth angle obtained by image processing of a frame image, can be directly obtained as an absolute value. On the other hand, the signal indicating the azimuth angle or the movement vector Oaz has a low S / N ratio, and the signal is smoothed by applying a low-pass filter or the like. Therefore, in a straight-line state in which the displacement of the azimuth angle of the vehicle 30 is small, the error due to the phase delay associated with the signal processing is small, and the external azimuth angle can be used as the vehicle azimuth angle. On the other hand, in a turning state in which the displacement of the azimuth angle of the vehicle 30 is large, the error between the external azimuth angle and the vehicle azimuth angle at the current position of the vehicle 30 becomes large due to the phase delay, and the vehicle azimuth angle of the vehicle 30 cannot be accurately indicated. In addition, when the movement vector Oaz is used as the external azimuth angle, it differs from the vehicle azimuth angle as a physical quantity.
[0041] In contrast, the internal azimuth angle, which is the azimuth angle acquired by the yaw rate sensor 31 mounted on the vehicle 30, has a characteristic of having a high S / N ratio. On the other hand, when the internal state detection unit 31 such as a yaw rate sensor is used, the azimuth angle is calculated by accumulating the detection values, and therefore the tolerance is also accumulated, and the azimuth angle has a characteristic of deviating from the true value.
[0042] The vehicle azimuth angle acquisition device 10 according to the first embodiment calculates the vehicle azimuth angle according to the characteristics of the external azimuth angle and the internal azimuth angle, and therefore, compared with the case where the external azimuth angle and the internal azimuth angle, or the combination of the external azimuth angle and the internal azimuth angle, can reduce the deviation from the true value and can acquire the vehicle azimuth angle with a good S / N ratio. That is, when the vehicle 30 is in a stopped state, the priority of the external azimuth angle is set higher than the priority of the internal azimuth angle to acquire the vehicle azimuth angle. When the vehicle 30 is in a turning state, or when the vehicle 30 is in a straight-moving state and the certainty of the external azimuth angle is low, the priority of the internal azimuth angle is set higher than the priority of the external azimuth angle to acquire the vehicle azimuth angle. Furthermore, when the vehicle 30 is in a straight-moving state and the certainty of the external azimuth angle is high, a correction is performed using the external azimuth angle to calibrate the accumulated error of the internal azimuth angle. In addition, when combining the external azimuth angle and the internal azimuth angle, AI, i.e., a learning model, may be used. By using the learning model and continuing the learning of the learning model, it is possible to realize a more appropriate combination priority of the external azimuth angle and the internal azimuth angle according to each running state of the vehicle.
[0043] According to the vehicle azimuth angle acquisition device 10 of the first embodiment, since it is possible to acquire a highly accurate vehicle azimuth angle, it is possible to control the position of the vehicle 30 with high accuracy in the self-propelled control of the vehicle 30 executed using the vehicle azimuth angle. As a result, it is possible to improve the accuracy of the stopping position of the vehicle 30 at the final destination of the vehicle 30. The various advantages described above can also be obtained in the vehicle azimuth angle acquisition system 100 and the vehicle azimuth angle acquisition method according to this embodiment.
[0044] In other embodiments: (1) The vehicle azimuth angle acquisition device 10 according to the first embodiment changes the priority of the internal azimuth angle and the external azimuth angle when calculating the vehicle azimuth angle according to the traveling state of the vehicle. In contrast, the vehicle azimuth angle acquisition device 10 may change the priority of the internal azimuth angle and the external azimuth angle according to the state of the external state detection unit 21 arranged on the route Rt on which the vehicle 30 travels. For example, in FIG. 5, the arrangement density / arrangement frequency of the external state detection unit 21 around the vehicle in the straight traveling state Po3 is higher than the arrangement density / arrangement frequency of the external state detection unit 21 around the vehicle in the straight traveling state P0. In this case, since it is possible to use an external azimuth angle with a high degree of certainty, the priority of the external azimuth angle may be set higher than the priority of the internal azimuth angle. In particular, when the vehicle 30 is traveling straight, the movement vector Oaz and the vehicle azimuth angle can be treated as substantially the same physical quantity, and by using the exterior azimuth angle as an absolute value with high certainty, it is possible to obtain a highly accurate vehicle azimuth angle without using an interior azimuth angle.
[0045] (2) In the first embodiment, the vehicle azimuth angle acquisition device 10 is described as a server computer having a function of acquiring a vehicle azimuth angle. The server computer can also function as a remote control device that remotely controls the vehicle 30 by being provided with a remote control program for self-driving the vehicle 30. In order to realize the remote control, the server computer acquires vehicle information indicating the driving state from the vehicle 30, and generates a vehicle driving control signal for controlling the self-driving of the vehicle 30 using the acquired vehicle azimuth angle, position coordinates, and vehicle information. The server computer transmits the generated vehicle driving control signal to the vehicle 30. The vehicle 30 that receives the vehicle driving control signal adjusts the output of the drive source, adjusts the braking amount of the braking device, and adjusts the steering angle of the steering device in accordance with the vehicle driving control signal to realize the self-driving of the vehicle 30. The capture of the vehicle 30 during remote control can be performed, for example, by capturing a predetermined end point in a frame image acquired by the external state detection unit 21 as a point of interest. Therefore, the vehicle 30 is captured at the timing when the external azimuth and position coordinates of the target vehicle 30 are acquired by the vehicle azimuth angle acquisition device 10. That is, acquisition of the vehicle azimuth angle in the first embodiment can be executed as part of a processing routine for remote control of the vehicle 30. By using the vehicle azimuth angle acquired by the vehicle azimuth angle acquisition device 10 according to the first embodiment, the accuracy of remote control of the vehicle 30 is improved. The accuracy of remote control includes a position deviation from the route from the departure point to the target point, and a deviation from the arrival position at the target point.
[0046] (3) In the first embodiment, the control mode of the self-propelled vehicle 30 using the vehicle azimuth acquired by the vehicle azimuth acquisition device 10 is not described. The vehicle azimuth acquisition device 10 according to the first embodiment is applied in a factory where the vehicle 30 is manufactured and in a yard where completed vehicles are stored. More specifically, in a factory, the movement of the vehicle 30 between each manufacturing process, the movement of the vehicle to the yard after the vehicle is completed, and the movement of the vehicle between the yard and a vehicle transport ship or train are included. In each of these movements, the vehicle 30 travels under self-propelled control. By improving the acquisition accuracy of the vehicle azimuth when the vehicle 30 travels on its own, the final stop position of the vehicle 30 can be realized with the desired accuracy.
[0047] (4) In the first embodiment, the vehicle 30 may have a configuration capable of moving by remote control, and may be in the form of a platform having the configuration described below, for example. Specifically, the vehicle 30 may have at least a vehicle control device, a drive device, a steering device, a braking device, and a wireless transceiver 32 in order to perform the three functions of "running," "turning," and "stopping" by remote control. That is, the vehicle 30 that can be moved by remote control may not be equipped with at least a part of interior parts such as a driver's seat and a dashboard, may not be equipped with at least a part of exterior parts such as a bumper and a fender, and may not be equipped with a body shell. In this case, the remaining parts such as a body shell may be attached to the vehicle 30 before the vehicle 30 is shipped from the factory, or the remaining parts such as a body shell may be attached to the vehicle 30 after the vehicle 30 is shipped from the factory in a state in which the remaining parts such as the body shell are not attached to the vehicle 30. Note that, for the form of a platform, position determination may be performed to capture a predetermined end point as a point of interest in the same manner as for the vehicle 30.
[0048] (5) In the first embodiment, the vehicle azimuth angle acquisition device 10 is realized by the CPU 101 executing the vehicle azimuth angle acquisition program Pr1, which determines the vehicle azimuth angle used for self-driving of the vehicle 30 using the external azimuth angle and the internal azimuth angle. However, the vehicle azimuth angle acquisition device 10 may be realized in hardware by a pre-programmed integrated circuit or a discrete circuit. That is, the control unit and the method thereof in each of the above embodiments may be realized by a dedicated computer provided by configuring a processor and a memory programmed to execute one or more functions embodied in a computer program. Alternatively, the control unit and the method thereof described in the present disclosure may be realized by a dedicated computer provided by configuring a processor with one or more dedicated hardware logic circuits. Alternatively, the control unit and the method thereof described in the present disclosure may be realized by one or more dedicated computers configured by a combination of a processor and a memory programmed to execute one or more functions and a processor configured with one or more hardware logic circuits. In addition, the computer program may be stored in a computer-readable non-transient tangible recording medium as instructions executed by a computer.
[0049] The present disclosure has been described above based on the embodiments and modifications, but the above-described embodiments of the present invention are intended to facilitate understanding of the present disclosure and do not limit the present disclosure. The present disclosure may be modified or improved without departing from the spirit and scope of the claims, and the present disclosure includes equivalents thereof. For example, the embodiments and technical features in the modifications corresponding to the technical features in each form described in the Summary of the Invention column can be appropriately replaced or combined in order to solve some or all of the above-mentioned problems or to achieve some or all of the above-mentioned effects. Furthermore, if a technical feature is not described as essential in this specification, it can be appropriately deleted. [Explanation of symbols]
[0050] 10...vehicle azimuth angle acquisition device, 21...external state detection unit, 22...access point, 30...vehicle, 31...internal state detection unit, 101...CPU, 102...memory, 103...input / output interface, 104...bus, 300...driving control device, 500...vehicle azimuth angle acquisition system, Pr1...vehicle azimuth angle acquisition program.
Claims
1. An azimuth angle acquisition device for a self-propelled vehicle, an acquisition unit that acquires an external state, which is a motion state of the vehicle detected from the external world, and an internal state, which is a motion state of the vehicle detected from the internal world; an azimuth angle calculation unit that acquires an external azimuth angle using the acquired external state and calculates an internal azimuth angle using the acquired internal state; and an azimuth angle determination unit that determines a vehicle azimuth angle used for self-propelled driving of the vehicle using the calculated external azimuth angle and internal azimuth angle.
2. 2. The azimuth angle acquisition device according to claim 1, The azimuth angle determination unit determines the vehicle azimuth angle by changing the priority of the external azimuth angle and the internal azimuth angle depending on the state of the vehicle and the state of an external state detection unit that detects the external state.
3. 3. The azimuth angle acquisition device according to claim 2, The azimuth angle determination unit determines the vehicle azimuth angle by giving a higher priority to the external azimuth angle when the motion state of the vehicle is at a stop.
4. 3. The azimuth angle acquisition device according to claim 2, The azimuth angle determination unit determines the vehicle azimuth angle by giving a higher priority to the internal azimuth angle when the motion state of the vehicle is a turning state.
5. 3. The azimuth angle acquisition device according to claim 2, wherein, when the motion state of the vehicle is in a straight-ahead state and the certainty of the external state detection unit is equal to or greater than a predetermined threshold, the azimuth angle determination unit corrects the internal azimuth angle using the external azimuth angle, and when the certainty of the external state detection unit is less than a predetermined threshold, the azimuth angle determination unit determines the vehicle azimuth angle by giving a higher priority to the internal azimuth angle.
6. 3. The azimuth angle acquisition device according to claim 2, the external environment state detection unit is an imaging device disposed in an external environment of the vehicle, The azimuth angle calculation unit calculates an external azimuth angle of the vehicle using an image captured by the imaging device.
7. 3. The azimuth angle acquisition device according to claim 2, the internal state detection unit that detects the internal state is a yaw rate sensor provided in the vehicle, The azimuth angle calculation unit calculates an internal azimuth angle of the vehicle by using a yaw rate acquired by the yaw rate sensor.
8. An azimuth angle acquisition system for a self-propelled vehicle, comprising: an external environment state detection unit disposed in an external environment of the vehicle to detect a motion state of the vehicle from the external environment; an internal state detection unit mounted on the vehicle for detecting a motion state of the vehicle from an internal environment; An azimuth angle acquisition device, an azimuth angle calculation unit that calculates an external azimuth angle using the external state of the vehicle detected by the external state detection unit, and calculates an internal azimuth angle using the internal state of the vehicle detected by the internal state detection unit; and an azimuth angle determination unit that determines a vehicle azimuth angle used for self-propelled driving of the vehicle using the calculated external azimuth angle and internal azimuth angle.
9. A method for acquiring an azimuth angle of a self-propelled vehicle, comprising: Acquire an external state, which is a motion state of the vehicle detected from the external world, and an internal state, which is a motion state of the vehicle detected from the internal world; Acquire an external field azimuth angle using the acquired external field state, and calculate an internal field azimuth angle using the acquired internal field state; determining a vehicle azimuth angle to be used for self-propelled driving of the vehicle using the calculated external azimuth angle and internal azimuth angle.
Citation Information
Patent Citations
Information processor, system, method, and program
JP2019125354A
Vehicle positioning system and vehicle remote operation system
JP2022171172A
Method for operating a vehicle and method for operating a manufacturing system
JP2017538619A
