Position calculation device

Abstract

To obtain a position calculation device that can accurately indicate the position of a mobile body as probability distribution, including a scene in which it is difficult to detect an object around the mobile body and receive signals from a GNSS.SOLUTION: A current position can be determined by determining a satellite positioning position determined by receiving signals from a satellite, a map reference position determined by grasping the surrounding environment and referring to map information, and a probability distribution obtained by composing probability distributions generated on the basis of the degrees of variations, and composing a self-contained navigation position determined by self-contained navigation and the probability distribution generated on the basis of the degrees of variations, and the self-contained navigation position can be corrected. Consequently, even when it is difficult to detect an object around the device, when it is difficult to receive signals from a GNSS, or when both the detection and reception are difficult, it is possible to accurately calculate the current position and indicate it as the probability distribution.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present application relates to a position calculation device. [Background technology]

[0002] Conventionally, there has been known a technique for calculating the position of a mobile body using information from a position calculation device mounted on the mobile body. A mobile body is a movable object with a control function, such as an automobile, an aircraft, a ship, a mobile cart, or a robot. The mobile body is also equipped with a sensor for identifying external objects, determining their relative positions, and calculating the current position by comparing them with map information, or for independently determining the position of the mobile body. To identify external objects, there are devices that measure information about external objects, such as visible light cameras and radars, devices that acquire information about objects external to the mobile body from map information, and devices that calculate the position of the mobile body from satellite signals.

[0003] Environmental conditions change when a moving object moves indoors and outdoors. Technology has been disclosed for a position calculation device that can accurately calculate the position of a moving object even when the environmental conditions in which the moving object is placed change in this way (for example, Patent Document 1). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 6962007 Summary of the Invention [Problem to be solved by the invention]

[0005] The technology disclosed in Patent Document 1 describes an autonomous vehicle as a moving object. The technology in Patent Document 1 compares the relative positions of surrounding obstacles with map information and calculates the position as a probability distribution. The position is also calculated as a probability distribution using positioning information from a Global Navigation Satellite System (GNSS). The final self-position can be estimated by combining these probability distributions. By changing the combination coefficient depending on the situation in which the vehicle is placed, the self-position can be seamlessly estimated both indoors and outdoors, enabling autonomous driving.

[0006] The technology described in Patent Document 1 estimates the vehicle's position while detecting surrounding objects indoors where there are referable objects around the autonomous vehicle, and estimates the vehicle's position using GNSS in open areas such as a playground where there are no referable objects around. However, estimating the vehicle's position becomes difficult in situations where it is difficult to detect surrounding objects and receive GNSS signals, such as inside a tunnel or on a snowy field during bad weather.

[0007] The present disclosure aims to solve such problems and provide a position calculation device that can accurately represent the position of a moving body as a probability distribution, including in situations where it is difficult to detect surrounding objects and receive GNSS signals. [Means for solving the problem]

[0008] A position calculation device according to the present disclosure includes: a satellite positioning unit that calculates a satellite-based position, which is the position of the mobile object calculated by receiving a signal from a satellite using a receiver mounted on the mobile object, and a first degree of variation when the satellite-based position is calculated; a map reference position calculation unit that calculates a map reference position, which is the position of the mobile body calculated by referring to map information from the surrounding environment grasped by an appearance recognition sensor mounted on the mobile body, and a second degree of variation when the map reference position is calculated; a self-contained navigation unit that calculates a self-contained navigation position, which is the position of the moving body calculated based on the traveling speed and traveling direction of the moving body derived from the output of a moving state detection sensor that detects the moving state of the moving body mounted on the moving body, and a third degree of variation when the self-contained navigation position is calculated; and The apparatus includes a correction calculation unit that generates a first probability distribution, which is a probability distribution of satellite-based positioning positions in a position coordinate space, based on the first degree of variation and the satellite-based position, generates a second probability distribution, which is a probability distribution of map-referenced positions in a position coordinate space, based on the second degree of variation and the map-referenced positions, generates a third probability distribution, which is a probability distribution of autonomous navigation positions in a position coordinate space, based on the third degree of variation and the autonomous navigation positions, generates a fourth probability distribution by combining the first probability distribution and the second probability distribution, generates a fifth probability distribution by combining the fourth probability distribution and the third probability distribution, and corrects the autonomous navigation position using a corrected mobile body position, which is the position of the mobile body obtained from the fifth probability distribution. [Effects of the Invention]

[0009] According to a position calculation device disclosed herein, the position of a moving body can be accurately represented as a probability distribution, including in situations where detecting surrounding objects and receiving GNSS signals are difficult. The satellite positioning position obtained by receiving signals from satellites is calculated as a first probability distribution, the map reference position obtained by understanding the surrounding environment and referring to map information is calculated as a second probability distribution, a third probability distribution is calculated by combining these, a self-contained navigation position obtained by self-contained navigation is calculated as a third probability distribution, and the current position can be represented as a probability distribution obtained by combining the fourth probability distribution and the third probability distribution. This makes it possible to accurately calculate the current position and express it as a probability distribution, even in situations where detecting surrounding objects is difficult, where receiving GNSS signals is difficult, or where both are difficult. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a configuration diagram of a position calculation device according to a first embodiment. [Figure 2]1 is a hardware configuration diagram of a position calculation device according to a first embodiment. [Figure 3A] FIG. 2 is a first diagram illustrating a procedure for calculating a position by the position calculation device according to the first embodiment. [Figure 3B] FIG. 10 is a second diagram illustrating the procedure of position calculation by the position calculation device according to the first embodiment. [Figure 3C] FIG. 10 is a third diagram illustrating the procedure of position calculation by the position calculation device according to the first embodiment. [Figure 4] 5 is a flowchart showing the processing of a satellite positioning unit of the position calculation device according to the first embodiment. [Figure 5] 5 is a flowchart showing the processing of a map reference position calculation unit of the position calculation device according to the first embodiment. [Figure 6] 4 is a first flowchart showing the processing of a self-contained navigation unit and a correction calculation unit of the position calculation device according to the first embodiment. [Figure 7] 10 is a second flowchart showing the processing of the self-contained navigation unit and the correction calculation unit of the position calculation device according to the first embodiment. [Figure 8] 10 is a third flowchart showing the processing of the self-contained navigation unit and the correction calculation unit of the position calculation device according to the first embodiment. [Figure 9] 10 is a fourth flowchart showing the processing of the self-contained navigation unit and the correction calculation unit of the position calculation device according to the first embodiment. [Figure 10] 10 is a fifth flowchart showing the processing of the self-contained navigation unit and the correction calculation unit of the position calculation device according to the first embodiment. [Figure 11] FIG. 4 is an explanatory diagram for a case where the degree of variation in the position calculation device according to the first embodiment is increased. [Figure 12] FIG. 10 is an explanatory diagram for explaining a case where the degree of variation in the position calculation device according to the second embodiment is increased. [Figure 13] 10 is a flowchart of a process for increasing the degree of variation in the position calculation device according to the second embodiment. [Figure 14] FIG. 11 is an explanatory diagram for a case where the degree of variation in the position calculation device according to the third embodiment is reduced. [Figure 15]11 is a flowchart of a process for reducing the degree of variation in the position calculation device according to the third embodiment. [Figure 16] FIG. 11 is an explanatory diagram illustrating a case where the first degree of variation of the position calculation device according to the fourth embodiment is increased. [Figure 17] 13 is a flowchart of a process for increasing a first degree of variation in the position calculation device according to the fourth embodiment. [Figure 18] 13 is a flowchart of a process of increasing a first degree of variation in the left-right direction in the position calculation device according to the fourth embodiment. [Figure 19] 13 is a flowchart of a process for increasing the degree of variation in the position calculation device according to the fifth embodiment. [Figure 20] FIG. 20 is an explanatory diagram illustrating a case where the first degree of variation in the position calculation device according to the sixth embodiment is increased. [Figure 21] 13 is a flowchart of a process for increasing a first degree of variation in the position calculation device according to the sixth embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, embodiments will be described in detail with reference to the drawings. Note that the drawings are schematic, and for the sake of convenience, configurations may be omitted or simplified as appropriate. In the following description, similar components are denoted by the same reference numerals, and their names and functions are also the same. Therefore, detailed descriptions thereof may be omitted to avoid duplication.

[0012] 1. First Embodiment <Configuration of the position calculation device> 1 is a configuration diagram of a position calculation device 100 according to embodiment 1. The position calculation device 100 mounted on a moving object 1 includes a satellite positioning unit 10, a map reference position calculation unit 20, a map information storage unit 22, a self-contained navigation unit 30, and a correction calculation unit 40 (the moving object 1 is shown in FIGS. 16 and 20).

[0013] <Satellite Positioning Unit> The satellite positioning unit 10 receives signals from navigation satellites via the GNSS receiver 11, and can receive signals from multiple satellites to obtain information such as the latitude, longitude, altitude, direction, speed, angular velocity, and degree of error of the current position. While an example using GNSS is shown here, the satellites used are not limited to this. The observation update cycle for satellite positioning is often around 100 ms to 200 ms.

[0014] The satellite positioning unit 10 outputs the satellite positioning position PGS and a first degree of variation EE1 as error information. Errors in the satellite positioning position PGS include errors caused by the number of satellites that can be received, the angle at which the satellites that can be received are located, and clock discrepancies in the transmitters installed on each satellite, as well as errors caused by signal reflection when the GNSS receiver 11 receives radio waves from each satellite, errors caused by noise due to weakened signals, and calculation errors until the satellite positioning position PGS is calculated. These errors are combined to produce the first degree of variation EE1.

[0015] The first degree of variation EE1 may be calculated based on the reception status of radio waves from a satellite, or may be calculated based on an actual measurement value that is set or learned depending on the location where the moving object 1 is traveling.

[0016] In areas with low visibility, such as mountainous regions or near high-rise buildings, GNSS signals can be blocked or reflected, resulting in multipath interference and increased detection errors. Heavy rain, thunderstorms, and heavy snow can also prevent satellite signal reception. Furthermore, satellite signal reception can be disrupted inside buildings and tunnels, making reception impossible.

[0017] <Map reference position calculation section> The map reference position calculation unit 20 receives information about the appearance of surrounding objects from an appearance recognition sensor 21. The appearance recognition sensor 21 can be an image sensor, a radio wave sensor, an optical sensor, an ultrasonic sensor, or the like. The image sensor, as typified by a surveillance camera, photographs the object and calculates the distance to the object from the image data photographed within a certain viewing angle range. The size, direction of movement, speed, type, etc. of the object can also be obtained from the image data. The image sensor can be a visible light camera, an infrared camera, or the like.

[0018] The radio wave sensor can be a millimeter wave radar (MMWR) that uses a frequency band of 24 to 79 GHz, etc. The radio wave sensor can detect the position of an object and the speed of the object's movement by the Doppler effect.

[0019] Examples of optical sensors that can be used include laser radar, LiDAR (Light Detection and Ranging), etc. LiDAR emits laser light within a certain field of view and detects point cloud data obtained by the reflection of the laser light from objects, allowing the position and shape of the objects to be determined.

[0020] Information processing may be performed for each sensor, such as an image sensor, radio wave sensor, optical sensor, or ultrasonic sensor, which is the appearance recognition sensor 21. In this way, data acquired by the various sensors can be processed, and only information about the identified object (for example, the shape and relative position of the object) can be transmitted to the map reference position calculation unit 20.

[0021] The map reference position calculation unit 20 transmits information about surrounding objects received from the appearance recognition sensor 21 to the map information generation unit 23. The map information generation unit 23 can execute SLAM (Simultaneous Localization and Mapping) processing to build a new map about surrounding objects in the map information storage unit 22 and update the map.

[0022] Furthermore, the map reference position calculation unit 20 can directly register information about surrounding objects received from the appearance recognition sensor 21 in the map reference position calculation unit 20 to update the dynamic map. For objects that are fixed and defined in advance in the map information among the information about surrounding objects received from the appearance recognition sensor 21, the map reference unit 24 refers to and collates the map data, and calculates the map reference position PMM, which is the current position, and a second degree of variation EE2. The degree of variation is also called variance.

[0023] The second degree of variation EE2 is generated by a combination of errors in distance and angle detection by sensors that grasp various external environments, recognition errors when recognizing the position and shape of an object from sensor signals, errors in describing the object in the map information, and calculation errors until the map-referenced position PMM is determined. The second degree of variation EE2 may be calculated based on errors until the map-referenced position PMM is calculated. Alternatively, the second degree of variation EE2 may be calculated based on actual measurement values ​​that are set or learned according to the location where the mobile object 1 is traveling.

[0024] The map reference position calculation unit 20 may receive the satellite-based position PGS as an initial value or a reference value from the satellite positioning unit 10. The update cycle of the map reference position calculated by the map reference position calculation unit 20 is often about 20 ms to 50 ms.

[0025] Regarding the appearance recognition sensor 21, the detection accuracy and variability of image sensors and optical sensors can change significantly depending on changes in the natural environment, such as dense fog, heavy rain, and heavy snow, making it difficult to detect objects. Furthermore, in places where there are no landmarks, such as on plains or inside tunnels, it is sometimes impossible to detect an object and calculate its position by referring to a map.

[0026] <Autonomous Navigation Department> The autonomous navigation unit 30 receives information from a movement state detection sensor mounted on the mobile object 1 that detects the movement state of the mobile object. Specifically, it receives information such as movement amount, speed, acceleration, rotation angle, angular velocity, and angular acceleration from an IMU (Inertial Measurement Unit). Here, an IMU is described as an example of a movement state detection sensor, but it may also be an AHRS (Attitude and Heading Reference System), an INS (Inertial Navigation System), or a collection of individual speedometers, accelerometers, angular velocity meters, and angular accelerometers. An internal sensor, which means a sensor used in the field of robotics to detect internal conditions, may also be used. The speed or travel distance (movement amount) may be obtained from the vehicle speed signal. The movement direction (orientation) may also be obtained from the vehicle's steering angle or the orientation detected by a geomagnetic sensor.

[0027] Velocity can be determined by differentiating the amount of movement, and acceleration can be determined by differentiating the velocity. Conversely, velocity can be determined by integrating acceleration, and velocity can be determined by integrating the amount of movement. These relationships also apply to the relationships between rotation angle (azimuth), angular velocity, and angular acceleration. Therefore, the self-contained navigation unit 30 can calculate the position of the mobile object 1 based on the traveling speed and traveling direction of the mobile object derived from the outputs of at least one of a traveling distance sensor, a traveling speed sensor, or a traveling acceleration sensor, and at least one of a traveling direction sensor, an angular velocity sensor, or an angular acceleration sensor mounted on the mobile object. The update period of the self-contained navigation position PIM using self-contained navigation is often approximately 10 ms to 20 ms, but it is also possible to integrate the position at a shorter update period.

[0028] Alternatively, the position and attitude of the moving body 1 may be detected visually by visual odometry using the appearance recognition sensor 21 from camera images, etc., without using an IMU. Self-contained navigation may be achieved by visually determining any of the amount of movement, speed, or acceleration in the vertical, horizontal, or vertical directions, and any of the rotation angle, rotational angular velocity, or rotational angular acceleration around the three axes of yawing, rolling, and pitching. Here, self-contained navigation refers to navigation that determines position, speed, direction of travel, etc., without relying on external equipment such as artificial satellites or radio wave transmission facilities, or external information such as high-resolution maps or dynamic maps.

[0029] The autonomous navigation unit 30 can update the autonomous navigation position PIM, which is the current position, at predetermined time intervals (for example, every 10 ms) based on the signal from the IMU 31. At that time, the autonomous navigation unit 30 can calculate a third degree of variation EE3 as the degree of variation. The third degree of variation EE3 is generated by a combination of errors due to drift and offset of each sensor of the IMU, errors due to resolution, gain errors, and calculation errors until the current position is determined from such information.

[0030] The third degree of variation EE3 may be calculated based on errors of the sensors of the IMU, or may be calculated based on actual measurement values ​​that are set or learned depending on the location where the moving object 1 is traveling.

[0031] The autonomous navigation unit 30 can express the current position as the accumulation of relative position changes, but is not given an initial value. Therefore, the deviation from the initial value given by the satellite positioning unit 10 may be accumulated and corrected for use.

[0032] Self-contained navigation can calculate a position very accurately for relatively short distances. However, because it calculates by accumulating changes in position due to movement, the impact of accumulated errors is significant. Therefore, it is useful to update the current position and reset the accumulated error when an accurate position is successfully detected using another method.

[0033] <Correction calculation section> The correction calculation unit 40 has a synthesis unit 41 and a synthesis correction unit 42. A first probability distribution P1 is generated from the satellite-positioned position PGS and the first degree of variation EE1 received from the satellite positioning unit 10. A second probability distribution P2 is generated from the map-referenced position PMM and the second degree of variation EE2 received from the map-referenced position calculation unit 20. A third probability distribution P3 is generated from the self-contained navigation position PIM and the third degree of variation received from the self-contained navigation unit 30.

[0034] A probability distribution represents the overall probability that a random variable takes on each value. Here, the probability distribution of the current position shows the probability of existence for each point in the area where the current position may exist.

[0035] The combining unit 41 combines the first probability distribution P1 calculated for the satellite position PGS and the second probability distribution P2 calculated for the map reference position PMM. By combining these, the current position can be estimated appropriately from the positions calculated by the two different means. When combining, by changing the combining coefficient (weighting function, etc.) according to the degree of variability of each, it is possible to combine the probability distributions with a small degree of variability by assigning a large weight and a small weight to the probability distribution with a large degree of variability. An extended Kalman filter may be used to combine the probability distributions. The combined probability distribution is called the fourth probability distribution P4.

[0036] The self-contained navigation position PIM calculated by the self-contained navigation unit 30 can be used even when the external environment changes, for example, when the surroundings are a snowy field in stormy weather, when traveling through a tunnel, etc. In these environments, the satellite-based position PGS calculated by the satellite positioning unit 10 and the map-referenced position PMM calculated by the map-referenced position calculation unit 20 cannot be trusted.

[0037] Therefore, the position calculation device 100 according to the first embodiment calculates the current position using a third probability distribution P3 generated based on the third probability distribution EE3 and the self-contained navigation position PIM. The synthesis corrector 42 generates a fifth probability distribution P5 by combining the fourth probability distribution P4 synthesized as described above with the third probability distribution P3. The generated fifth probability distribution P5 enables highly reliable position estimation.

[0038] For example, when the degree of variability of the fourth probability distribution P4 is large, such as when the surroundings are a snowy field in stormy weather or when driving through a tunnel, the influence of the fourth probability distribution P4 during synthesis is small. On the other hand, when the degree of variability of the fourth probability distribution P4 is small, the influence of the fourth probability distribution is large, and a position represented by a more reliable probability distribution can be obtained by synthesis.

[0039] Furthermore, by treating the position of the moving object as a probability distribution, it becomes possible to appropriately estimate the position of the moving object in response to various changes in environmental conditions. The most frequent value in this corrected probability distribution is referred to as the corrected moving object position PCMO. By correcting the latest self-contained navigation position PIM calculated by the self-contained navigation unit 30 to the corrected moving object position PCMO, it is possible to reset the accumulated error of the self-contained navigation unit 30. The corrected moving object position PCMO may also be the average or median value calculated from the corrected probability distribution.

[0040] <Hardware configuration of the position calculation device> 2 is a hardware configuration diagram of a position calculation device 100 according to the first embodiment. In this embodiment, the position calculation device 100 is an electronic control device mounted on a mobile body for calculating the position of the mobile body. Each function of the position calculation device 100 is realized by a processing circuit provided in the position calculation device 100. Specifically, the position calculation device 100 includes, as processing circuits, an arithmetic processing device 90 (computer) such as a CPU (Central Processing Unit), a storage device 91 that exchanges data with the arithmetic processing device 90, an input circuit 92 that inputs external signals to the arithmetic processing device 90, and an output circuit 93 that outputs signals from the arithmetic processing device 90 to the outside. Each piece of hardware, such as the arithmetic processing device 90, the storage device 91, the input circuit 92, and the output circuit 93, is connected to one another via a wired network such as a bus or a wireless network.

[0041] The arithmetic processing device 90 may include an application-specific integrated circuit (ASIC), an integrated circuit (IC), a digital signal processor (DSP), a graphics processing unit (GPU), a field programmable gate array (FPGA), various logic circuits, various signal processing circuits, etc. Furthermore, the arithmetic processing device 90 may include a plurality of the same or different types of devices, each performing a different process. The storage device 91 may include a random access memory (RAM) configured to be able to read and write data from the arithmetic processing device 90, and a read-only memory (ROM) configured to be able to read data from the arithmetic processing device 90. Nonvolatile or volatile semiconductor memories such as flash memory, solid-state drives (SSDs), EPROMs, and EEPROMs may also be used as the storage device 91. The input circuit 92 is connected to various sensors, switches, and communication lines, and includes an A / D converter, communication circuits, etc., that input output signals and communication information from these sensors and switches to the arithmetic processing device 90. 2, the input circuit 92 is connected to the GNSS receiver 11, the appearance recognition sensor 21, and the IMU 31. The output circuit 93 includes a drive circuit that outputs control signals from the arithmetic processing device 90, a communication circuit, and the like. The interfaces of the input circuit 92 and the output circuit 93 may be based on specifications such as CAN (Control Area Network) (registered trademark), Ethernet (registered trademark), or USB (Universal Serial Bus) (registered trademark). Furthermore, separate from the input circuit 92 and the output circuit 93, the arithmetic processing device 90 may be directly connected to a communication device for communication.

[0042] Each function of the position calculation device 100 is realized by the arithmetic processing device 90 executing software (programs) stored in a storage device 91 such as a ROM, and cooperating with other hardware of the position calculation device 100 such as the storage device 91, input circuit 92, and output circuit 93. Setting data such as thresholds and judgment values ​​used by the position calculation device 100 is stored in the storage device 91 such as a ROM as part of the software (programs). Each function of the position calculation device 100 may be configured as a software module, or may be configured as a combination of software and hardware.

[0043] <Position calculation procedure> Fig. 3A is a first diagram showing the procedure of position calculation by the position calculation device 100 according to Embodiment 1. Fig. 3B and Fig. 3C are second and third diagrams showing the procedure of position calculation.

[0044] 3A, 3B, and 3C, triangles indicate satellite-based position PGS, and the surrounding dashed lines indicate the range that the first probability distribution P1 can take. The dashed lines can be said to be the range that varies depending on the first degree of variation EE1. For convenience, the dashed lines are shown here as circles with the satellite-based position PGS at their center.

[0045] In reality, due to various factors, the range that the first probability distribution P1 can take may not be circular. The area indicated by the dashed line around the satellite positioning position PGS indicated by the triangle may be within a range of ±3σ by calculating the standard deviation from the degree of variation (variance).

[0046] The square marks indicate the map reference position PMM, and the surrounding dashed lines indicate the range that the second probability distribution P2 can take based on the second degree of variability EE2. The circle marks indicate the self-contained navigation position PIM, and the surrounding dashed lines indicate the range that the third probability distribution P3 can take based on the third degree of variability EE3. GD indicates the traveling direction of the moving object. The range of each dashed line may be within ±3σ, calculated by calculating the standard deviation from the degree of variability (variance), as in the case of the satellite positioning position PGS.

[0047] Figure 3B shows the synthesis of the first probability distribution P1 calculated for the satellite position PGS and the second probability distribution P2 calculated for the map reference position PMM. The square ■ indicates the synthesized position PCB, and the surrounding dashed line indicates the range of possible fourth probability distribution P4. When synthesizing, by changing the synthesis coefficient (weighting function, etc.) according to the degree of variability of each, it is possible to synthesize a more reliable fourth probability distribution P4.

[0048] 3C shows the combination of the fourth probability distribution P4 and the third probability distribution P3. The black circle indicates the corrected moving object position PCMO, which is a representative point of the combined fifth probability distribution P5. The dashed line around the corrected moving object position PCMO indicates the range that the corrected fifth probability distribution P5 can take.

[0049] The corrected moving object position PCMO may be defined as the mode of the fifth probability distribution P5. The corrected moving object position PCMO may be the mean or median value obtained from the fifth probability distribution. By correcting the latest autonomous navigation position PIM calculated by the autonomous navigation unit 30 to the corrected moving object position PCMO, the accumulated error of the autonomous navigation unit 30 can be reset.

[0050] When performing location estimation, combining probability distributions based on the positioning results of different positioning methods enables optimal synthesis depending on the accuracy of the positioning results. Synthesis can be performed by taking into account the circumstances under which the degree of variation increases and the mobility environment in which the degree of variation increases or decreases for each positioning method.

[0051] According to the position calculation device 100 of the first embodiment, the position of a moving body can be accurately represented as a probability distribution, even in situations where it is difficult to detect surrounding objects and receive GNSS signals. A first probability distribution P1 and a second probability distribution P2 are generated based on the satellite-based position PGS obtained by receiving signals from satellites and the map-reference position PMM obtained by understanding the surrounding environment and referring to map information, and the respective degrees of variation EE1 and EE2, and a fourth probability distribution P4 is obtained by combining these.

[0052] A third probability distribution P3 is generated based on the self-contained navigation position PIM calculated by self-contained navigation and the third degree of variation EE3, and a fifth probability distribution P5 is obtained by combining the third probability distribution P3 and the fourth probability distribution P4. By representing the current position using the fifth probability distribution P5, it is possible to calculate the current position with high accuracy even when it is difficult to detect surrounding objects, when it is difficult to receive GNSS signals, or when both are difficult. Furthermore, a corrected mobile station position PCMO can be calculated from the mode of the fifth probability distribution P5, and the self-contained navigation position PIM can be corrected. The corrected mobile station position PCMO may be the average or median calculated from the fifth probability distribution P5.

[0053] <Satellite positioning processing> Fig. 4 is a flowchart showing the processing of the satellite positioning unit 10 of the position calculation device 100 according to the first embodiment. The processing shown in Fig. 4 is executed by the arithmetic processing unit of the position calculation device 100. This processing may be executed at predetermined time intervals (for example, every 1 ms). It may also be executed in response to an event such as the GNSS receiver 11 newly calculating the current position, rather than at predetermined time intervals.

[0054] The process starts, and in step S101, the GNSS signal reception status is ascertained. The status is ascertained to determine whether the GNSS receiver 11 is able to receive signals, including the positions, number, and reception status of satellites that can be received. In step S102, it is confirmed whether new data for positioning calculations has been collected.

[0055] In step S103, it is determined whether or not all the data has been collected. If all the data has been collected (determination is YES), the process proceeds to step S104. If all the data has not been collected (determination is NO), the process ends.

[0056] In step S104, a positioning calculation is performed to obtain a satellite positioning position PGS. Simultaneously with the positioning calculation, a first degree of variation EE1 associated with the positioning calculation is also calculated. Here, the positioning calculation is performed in step S104, but the positioning calculation may be performed within the GNSS receiver 11, and the satellite positioning unit 10 may receive the calculation result when the positioning calculation is completed.

[0057] In step S105, the satellite-based position PGS and the first degree of variation EE1 are stored in the storage device. The current time at which these were calculated is also stored. In step S106, a satellite position update flag is set. This flag indicates that the satellite-based position PGS has been updated, and its initial value is cleared. Then, the process ends.

[0058] <Processing of map reference position calculation unit> 5 is a flowchart showing the processing of the map reference position calculation unit 20 of the position calculation device 100 according to the first embodiment. The processing shown in FIG. 5 is executed by the arithmetic processing unit of the position calculation device 100. This processing may be executed at predetermined time intervals (for example, every 1 ms). Instead of at predetermined time intervals, this processing may be executed in response to an event such as the appearance recognition sensor 21 newly recognizing an object.

[0059] The process starts, and in step S201, the signal from the appearance recognition sensor 21 is checked. In step S202, it is checked whether the presence of a new object has been recognized and whether sufficient data has been collected to confirm its location by referring to a map.

[0060] In step S203, it is determined whether or not all the data has been collected. If all the data has been collected (determination is YES), the process proceeds to step S204. If all the data has not been collected (determination is NO), the process ends.

[0061] In step S204, the positions and appearances of surrounding objects are determined from the data. In step S205, the objects are added to or updated in the map information. This is because if the object is not registered in the map information, it must be newly registered in the map information. If the object is already registered in the map information, the object information is updated.

[0062] In step S206, if the detected object is a stationary object registered on a map (e.g., a building, a sign, a traffic light, etc.), a map reference position is calculated. The relative positions of the object and the moving object are determined, and the object's position is confirmed using map information, thereby enabling the moving object's current position to be confirmed. At the same time as calculating the map reference position PMM, a second degree of variation EE2 is calculated.

[0063] In step S207, the map reference position PMM and the second degree of variation EE2 are stored in the storage device. Furthermore, the current time at which these are calculated is stored. In step S208, a map reference position update flag is set. This flag indicates that the map reference position PMM has been updated, and its initial value is cleared. Thereafter, the process ends.

[0064] <Processing in the autonomous navigation unit and correction calculation unit> 6 to 10 are flowcharts showing the processing of the self-contained navigation unit 30 and the correction calculation unit 40 of the position calculation device 100 according to the first embodiment. FIG. 7 shows a continuation of FIG. 6. FIG. 8 shows a continuation of FIG. 7. FIG. 9 shows a continuation of FIG. 8. FIG. 10 shows a continuation of FIG. 9. These processes may be executed at predetermined time intervals (for example, every 1 ms). Instead of at predetermined time intervals, they may be executed in response to an event such as the GNSS receiver 11 newly calculating the current position or the appearance recognition sensor 21 newly recognizing an object.

[0065] 6, the process starts in step S301, where it is checked whether 10 ms have passed since the self-contained navigation position was calculated last time. If 10 ms have passed since the self-contained navigation position was last updated (determination is YES), the process proceeds to step S305. If 10 ms have not passed since the self-contained navigation position was last calculated (determination is NO), the process proceeds to step S303.

[0066] In step S303, it is determined whether the satellite position update flag is set. If the satellite position update flag is set (determination is YES), the process proceeds to step S305. If the satellite position update flag is not set (determination is NO), the process proceeds to step S304.

[0067] In step S304, it is determined whether the map reference position update flag is set. If the map reference position update flag is set (determination is YES), proceed to step S305. If the map reference position update flag is not set (determination is NO), end the processing. In this case, neither update of the self-contained navigation position nor processing in the correction calculation unit 40 is performed.

[0068] In step S305, data such as acceleration, velocity, angular acceleration, and angular velocity are acquired from the values ​​of each sensor of the IMU, and a self-contained navigation position PIM is calculated using self-contained navigation. That is, not only is the self-contained navigation position PIM calculated every 10 ms, but the self-contained navigation position PIM and the third degree of variation EE3 are also calculated when the satellite positioning position PGS is calculated and when the map reference position PMM is calculated. Then, in step S306, the self-contained navigation position PIM and the third degree of variation EE3 are stored in the storage device. The calculated current time is also stored. Then, the process proceeds to step S311 in FIG. 7.

[0069] <Synchronization process of map reference position calculation timing> In step S311 of Fig. 7, it is determined whether the satellite position update flag is set. If the satellite position update flag is set (determination is YES), the process proceeds to step S311a. If the satellite position update flag is not set (determination is NO), the process proceeds to step S321 of Fig. 8.

[0070] In step S311a, the satellite position update flag is cleared. After that, data corresponding to the latest value of the map reference position PMM is generated and combined with the latest value of the satellite positioning position PGS. In step S312, the previously calculated map reference position PMM is read. Then, the previously calculated second degree of variation EE2 is read. Then, the time when the map reference position PMM was previously calculated is read and set as T2.

[0071] In step S313, the autonomous navigation position PIM at time T2 is read. Then, the autonomous navigation position PIM at the current time T1 is read. In step S314, an updated movement amount DMN, which is the amount of movement of the moving object between time T2 and time T1, is calculated. The updated movement amount DMN is the value obtained by subtracting the autonomous navigation position PIM(T2) from the autonomous navigation position PIM(T1), and can also be expressed as a vector pointing from the autonomous navigation position PIM(T2) to the autonomous navigation position PIM(T1).

[0072] In step S315, a position equivalent to the map-referenced position PMM at the current time T1 is calculated. The map-referenced position PMM at the current time T1 can be expressed as PMM(T1)=PMM(T2)+DMN. At this time, the second degree of variation EE2 at time T1 may be the same as the second degree of variation EE2 at time T2. Alternatively, the second degree of variation EE2 at time T1 may be a value extrapolated from the second degree of variation EE2 at both time T2 and the previous map-referenced position PMM calculation time T3 (time T3 is not shown). Then, the process proceeds to step S331 in FIG. 9.

[0073] <Satellite positioning position calculation timing synchronization processing> 8, it is determined whether the map reference position update flag is set. If the map reference position update flag is set (determination is YES), the process proceeds to step S321a. If the map reference position update flag is not set (determination is NO), the process ends.

[0074] In step S321a, the map reference position update flag is cleared. After that, data corresponding to the latest value of the satellite-based position PGS is generated and combined with the latest value of the map reference position. In step S322, the previously calculated satellite-based position PGS is read. Then, the previously calculated first degree of variation EE1 is read. Then, the time when the previous satellite-based position PGS was calculated is read and set as T2a.

[0075] In step S323, the autonomous navigation position PIM at time T2a is read. Then, the autonomous navigation position PIM at the current time T1 is read. In step S324, an updated movement amount DMN, which is the amount of movement of the moving object between time T2a and time T1, is calculated. The updated movement amount DMN is the value obtained by subtracting the autonomous navigation position PIM(T2a) from the autonomous navigation position PIM(T1), and can also be expressed as a vector pointing from the autonomous navigation position PIM(T2a) to the autonomous navigation position PIM(T1).

[0076] In step S325, a position equivalent to the satellite-based position PGS at the current time T1 is calculated. The satellite-based position PGS at the current time T1 can be expressed as PGS(T1)=PGS(T2a)+DMN.

[0077] At this time, the first degree of variation EE1 at time T2a may be used as the first degree of variation EE1 at time T1. Alternatively, the first degree of variation EE1 at time T1 may be a value extrapolated from the first degree of variation EE1 at both time T2a and the preceding time T3a when the satellite-based position PGS is calculated (time T3a is not shown). Then, the process proceeds to step S331 in FIG. 9.

[0078] <Expansion of the degree of variation> 9, the distance between the self-contained navigation position PIM and the satellite-based position PGS is calculated as the satellite-based positioning distance deviation ΔPGS. Then, in step S332, the distance between the self-contained navigation position PIM and the map-referenced position PMM is calculated as the map-referenced position distance deviation ΔPMM.

[0079] In step S333, it is determined whether the satellite positioning distance deviation ΔPGS is greater than the first distance threshold Dth1. If the satellite positioning distance deviation ΔPGS is greater than the first distance threshold Dth1 (determination is YES), proceed to step S334. If the satellite positioning distance deviation ΔPGS is not greater than the first distance threshold Dth1 (determination is NO), proceed to step S335.

[0080] In step S334, the first degree of variation EE1 is enlarged by multiplying it by (1+α), where α is a positive number. Here, the first degree of variation EE1 is enlarged because the satellite-based position PGS is too far from the reference self-contained navigation position PIM.

[0081] In step S335, it is determined whether the map-referenced position distance deviation ΔPMM is greater than the second distance threshold Dth2. If the map-referenced position distance deviation ΔPMM is greater than the second distance threshold Dth2 (determination is YES), proceed to step S336. If the satellite positioning distance deviation ΔPGS is not greater than the second distance threshold Dth2 (determination is NO), proceed to step S341 in FIG.

[0082] In step S336, the second degree of variation EE2 is multiplied by (1+α) to enlarge it. Here, the second degree of variation EE2 is enlarged because the map reference position PMM is too far from the reference self-contained navigation position PIM. Then, the process proceeds to step S341 in FIG. 10.

[0083] <Probability distribution synthesis processing> 10, a first probability distribution P1 is generated based on the satellite-based position PGS(T1) at time T1 and the first degree of variation EE1. In step S342, a second probability distribution P2 is generated based on the map-reference position PMM(T1) at time T1 and the second degree of variation EE2.

[0084] In step S343, the combining unit 41 combines the first probability distribution P1 calculated for the satellite position PGS and the second probability distribution P2 calculated for the map reference position PMM. By combining these, the current position can be estimated appropriately from the positions calculated by the two different means. When combining, by changing the combining coefficient (weighting function, etc.) according to the degree of variability of each, it is possible to combine the results by assigning a heavy weight to the position estimation means with a small degree of variability and a light weight to the positions with a large degree of variability. An extended Kalman filter may be used to combine the probability distributions. The combined probability distribution is set as the fourth probability distribution P4, and its representative value is set as the corrected mobile unit position PCMO.

[0085] In step S344, a third probability distribution P3 is generated based on the self-contained navigation position PIM(T1) at time T1 and the third degree of variation EE3. In step S345, the third probability distribution P3 calculated for the self-contained navigation position PIM(T1) and the fourth probability distribution P4 calculated for the combined position PCB are combined to calculate a fifth probability distribution P5.

[0086] When combining, the combination coefficient (weighting function, etc.) is changed according to the degree of variation of each. By doing so, it is possible to combine the images by assigning a large weight to positions with small errors and a small weight to positions with large errors.

[0087] In step S346, the mode of the fifth probability distribution P5 is set as the corrected moving object position PCMO. The corrected moving object position PCMO may be the average or median value obtained from the fifth probability distribution P5. At this time, by correcting the latest autonomous navigation position PIM calculated by the autonomous navigation unit 30 to the corrected moving object position PCMO, the accumulated error of the autonomous navigation unit 30 can be reset. Then, the processing ends.

[0088] <Explanation of the operation to increase the degree of variation> 11 is an explanatory diagram of a case where the degree of variation is increased in the position calculation device 100 according to Embodiment 1. The significance of increasing the degree of variation shown in steps S331 to S336 in FIG.

[0089] In Figure 11, triangles indicate satellite-based position PGS, and the surrounding dashed lines indicate the range that the first probability distribution P1 can take, showing the range of variation due to the first degree of variation EE1. Squares indicate map-referenced positions PMM, and the surrounding dashed lines indicate the range that the second probability distribution P2 can take due to the second degree of variation EE2. Circles indicate the range that the third probability distribution P3 can take due to the self-contained navigation position PIM and the third degree of variation EE3. GD indicates the traveling direction of the moving body.

[0090] The distance between the self-contained navigation position PIM and the satellite-based position PGS is shown as the satellite-based positioning distance deviation ΔPGS. Also, the distance between the self-contained navigation position PIM and the map-referenced position PMM is shown as the map-referenced position distance deviation ΔPMM.

[0091] Here, if the satellite positioning distance deviation ΔPGS, which is the distance from the self-contained navigation position PIM to the satellite-based position PGS, is greater than the first distance threshold Dth1, it means that the satellite-based position PGS is significantly deviated, and the first degree of variation EE1 of the first probability distribution P1 represented by the satellite-based position PGS is multiplied by (1 + α). By doing so, it is possible to reduce the weight of the first probability distribution P1 represented by the satellite-based position PGS when combining. In this case, the first distance threshold Dth1 may be the sum of the first degree of variation EE1 and the third degree of variation EE3.

[0092] Furthermore, if the map-referred position distance deviation ΔPMM, which is the distance of the map-referred position PMM with respect to the self-contained navigation position PIM, is greater than the second distance threshold Dth2, this means that the map-referred position PMM is significantly deviated, and the second degree of variation EE2 of the second probability distribution P2 represented by the map-referred position PMM is multiplied by (1 + α). This makes it possible to reduce the weight of the map-referred position PMM during synthesis. In this case, the second distance threshold Dth2 may be set to the sum of the second degree of variation EE2 and the third degree of variation EE3.

[0093] As described above, the position calculated by the satellite positioning unit 10 is represented by a first probability distribution P1 based on the satellite positioning position PGS and the first degree of variation, and the position calculated by the map reference position calculation unit 20 is represented as a second probability distribution P2 based on the map reference position PMM and the second degree of variation EE2, and by combining these, highly reliable position information can be obtained using an appropriate calculation method depending on the situation.

[0094] Furthermore, even if the timing of the position calculation by the satellite positioning unit 10 and the timing of the position calculation by the map reference position calculation unit 20 differ, the position at one calculation timing and the updated movement amount DMN from the previous calculation timing of the other can be calculated using the self-contained navigation position. Since the generation of both probability distributions is synchronized and combined, accurate position calculation at appropriate timing is possible. In addition, since the calculation and update of the self-contained navigation position by the self-contained navigation unit 30 can be performed in short cycles or at any timing, it is possible to calculate the updated movement amount DMN in accordance with the calculation timing of the position using other methods. This also contributes to highly reliable position calculation.

[0095] In the position calculation device 100 according to the first embodiment, the position calculated by the satellite positioning unit 10 is represented as a first probability distribution P1, the position calculated by the map reference position calculation unit 20 is represented as a second probability distribution, and these are combined to represent a fourth probability distribution P4. Then, the position calculated by the autonomous navigation unit 30 is represented as a third probability distribution P3, which is combined with the fourth probability distribution P4 to obtain a fifth probability distribution P5.

[0096] By expressing and combining the results as probability distributions using this procedure, it becomes possible to reduce the weight of the appropriate position calculation method as the degree of variation increases, thereby enabling highly reliable position calculation. Furthermore, by manipulating the degree of variation of each method depending on the situation, it is possible to further improve accuracy.

[0097] 2. Second Embodiment Fig. 12 is an explanatory diagram of a case where the degree of variation is increased in the position calculation device 100 according to the second embodiment. Fig. 13 is a flowchart of a process of increasing the degree of variation in the position calculation device 100 according to the second embodiment.

[0098] The configuration diagram and hardware configuration diagram of the position calculation device 100 according to the second embodiment are the same as those in Figures 1 and 2. The position calculation device 100 according to the second embodiment can be realized by modifying the software of the position calculation device 100 according to the first embodiment.

[0099] In Fig. 12, triangles indicate satellite positioning positions PGS, and the surrounding dashed lines indicate the range that the first probability distribution P1 can take, indicating the range of variation due to the first degree of variation EE1. In Fig. 12, the range that the first probability distribution P1 can take is indicated by a radius R1. The radius R1 may be determined based on 3σ calculated from the degree of first variation.

[0100] The squares indicate the map reference positions PMM, and the surrounding dashed lines indicate the range that the second probability distribution P2 can take based on the second degree of variation EE2. In Figure 12, the range that the second probability distribution P2 can take is indicated by a radius R2. The radius R2 may be determined based on 3σ calculated from the degree of second variation.

[0101] The circle indicates the autonomous navigation position PIM, and the surrounding dashed line indicates the range that the third probability distribution P3 can take based on the third degree of variation EE3. In Figure 12, the range that the third probability distribution P3 can take is indicated by a radius R3. The radius R3 may be determined based on 3σ calculated from the third degree of variation. GD indicates the traveling direction of the moving body.

[0102] The distance between the self-contained navigation position PIM and the satellite positioning position PGS is shown as the satellite positioning distance deviation ΔPGS. Also, the distance between the self-contained navigation position PIM and the map reference position PMM is shown as the map reference position distance deviation ΔPMM. The distance between the satellite positioning position PGS and the map reference position PMM is shown as the satellite positioning-map reference position deviation ΔP12.

[0103] Here, if the satellite positioning-map reference position deviation P12, which is the distance between the satellite positioning position PGS and the map reference position PMM, is larger than the maximum distance threshold Dthmx, it indicates that the satellite positioning position PGS and the map reference position PMM are far apart. In this case, the self-contained navigation position PIM is used as the reference, and the position farther away from the self-contained navigation position PIM is determined to be less reliable, and an operation is performed to increase the degree of variation.

[0104] This allows the weighting of the larger the degree of variation during synthesis to be reduced, thereby reducing the influence of probability distributions that are considered to be out of sync, thereby improving the reliability of position calculation.

[0105] <Expansion of the degree of variation> The flowchart of Fig. 13 according to the second embodiment is used in place of the flowchart of Fig. 9 according to the first embodiment. The first steps S331 and S332 have the same content as in Fig. 9, and therefore are given the same step numbers.

[0106] In step S343, the distance between the satellite-based position PGS and the map-referenced position PMM is calculated as the satellite-based position-to-map-referenced position deviation P12. Then, in step S344, it is determined whether the satellite-based position-to-map-referenced position deviation P12 is greater than the maximum distance threshold Dthmx.

[0107] If the satellite positioning / map reference position deviation P12 is greater than the maximum distance threshold Dthmx (determination is YES), proceed to step S345. If the satellite positioning / map reference position deviation P12 is not greater than the maximum distance threshold Dthmx (determination is NO), no operation on the degree of variation is performed, and proceed to step S341 in Figure 10 to start synthesizing the probability distribution.

[0108] In step S345, it is determined whether the satellite positioning distance deviation ΔPGS is greater than the map reference position distance deviation ΔPMM. If the satellite positioning distance deviation ΔPGS is greater than the map reference position distance deviation ΔPMM (determination is YES), the process proceeds to step S346.

[0109] In step S346, the first degree of variation is multiplied by (1 + α), where α is a positive number. That is, since the satellite-based position PGS is farther away from the self-contained navigation position PIM, the first degree of variation EE1 is enlarged to reduce the influence of the first probability distribution P1 during synthesis. Then, the process proceeds to step S341 in FIG. 10.

[0110] In step S345, if the satellite positioning distance deviation ΔPGS is not greater than the map reference position distance deviation ΔPMM (determination is NO), proceed to step S347. Then, in step S347, the second degree of variation EE2 is multiplied by (1 + α). That is, since the satellite positioning position PGS is not far from the self-contained navigation position PIM, an operation is performed to enlarge the second degree of variation EE2, thereby reducing the influence of the second probability distribution P2 during synthesis. Then, proceed to step S341 in FIG. 10.

[0111] 3. Embodiment 3 Fig. 14 is an explanatory diagram of a case where the degree of variation in the position calculation device 100 according to the third embodiment is reduced. Fig. 15 is a flowchart of a process for reducing the degree of variation in the position calculation device 100 according to the third embodiment.

[0112] The configuration diagram and hardware configuration diagram of the position calculation device 100 according to the third embodiment are the same as those in Fig. 1 and Fig. 2. The position calculation device 100 according to the third embodiment can be realized by modifying the software of the position calculation device 100 according to the first embodiment.

[0113] In Figure 14, triangles indicate satellite positioning positions PGS, and the surrounding dashed lines indicate the range that the first probability distribution P1 can take, with a radius of R1 representing the range of variation due to the first degree of variation EE1. Squares indicate map reference positions PMM, and the surrounding dashed lines indicate the range that the second probability distribution P2 can take with a radius of R2 due to the second degree of variation EE2. Finally, circles indicate the range that the third probability distribution P3 can take with a radius of R3 due to the autonomous navigation position PIM and the third degree of variation EE3.

[0114] The distance between the satellite-based position PGS and the map-referred position PMM is shown as the satellite-based position-to-map-referred position deviation ΔP12. GD indicates the traveling direction of the moving object. As in the second embodiment, the radii R1, R2, and R3 may be calculated based on 3σ calculated from the degree of variation of each.

[0115] Here, consider a case where the first degree of variation EE1 for the satellite-based position PGS and the second degree of variation EE2 for the map-referred position PMM are both smaller than the variation degree threshold EEth. Furthermore, consider a case where the satellite-based position-to-map-referred position deviation ΔP12, which is the distance between the satellite-based position PGS and the map-referred position PMM, is smaller than the minimum distance threshold Dthmn.

[0116] If the above conditions are satisfied, both are deemed more reliable, and the first degree of variation EE1 and the second degree of variation EE2 are both decreased. This increases the weighting of the probability distribution of the more reliable position calculation method in the synthesis, thereby improving reliability.

[0117] <Variation reduction processing> The flowchart shown in Fig. 15 is inserted before step S341 in Fig. 10. In step S343, the distance between the satellite-based position PGS and the map-referenced position PMM is calculated as the satellite-based position-to-map-referenced position deviation P12.

[0118] In step S354, it is determined whether the first degree of variation EE1 is smaller than the variation degree threshold EEth. If the first degree of variation EE1 is smaller than the variation degree threshold EEth (determination is YES), proceed to step S355. If the first degree of variation EE1 is not smaller than the variation degree threshold EEth (determination is NO), proceed to step S341 in FIG. 10 without performing a variation degree reduction operation.

[0119] In step S355, it is determined whether the second degree of variation EE2 is smaller than the variation degree threshold value EEth. If the second degree of variation EE2 is smaller than the variation degree threshold value EEth (determination is YES), proceed to step S356. If the second degree of variation EE2 is not smaller than the variation degree threshold value EEth (determination is NO), proceed to step S341 in FIG. 10 without performing a variation degree reduction operation.

[0120] In step S356, it is determined whether the satellite positioning-map reference position deviation ΔP12 is smaller than the minimum distance threshold Dthmn. If the satellite positioning-map reference position deviation ΔP12 is smaller than the minimum distance threshold Dthmn (determination is YES), proceed to step S357. If the satellite positioning-map reference position deviation ΔP12 is not smaller than the minimum distance threshold Dthmn (determination is NO), proceed to step S341 in FIG. 10.

[0121] In step S357, the first degree of variation is multiplied by (1-β), and in step S357, the second degree of variation EE2 is multiplied by (1-β), where β is a number greater than 0 and less than 1. Then, the process proceeds to step S341 in FIG.

[0122] 4. Embodiment 4 Fig. 16 is an explanatory diagram of a case where the first degree of variation is increased in the position calculation device 100 according to the fourth embodiment. Fig. 17 is a flowchart of a process of increasing the first degree of variation in the position calculation device 100 according to the fourth embodiment.

[0123] The configuration diagram and hardware configuration diagram of the position calculation device 100 according to the fourth embodiment are the same as those in Fig. 1 and Fig. 2. The position calculation device 100 according to the fourth embodiment can be realized by modifying the software of the position calculation device 100 according to the first embodiment.

[0124] FIG. 16 shows a moving object 1 passing through a road with buildings 2 on both sides. GD indicates the traveling direction of the moving object. The position calculation device 100 of the moving object 1 can know the layout of the surrounding buildings 2 from map information. In addition, the appearance recognition sensor 21 can also detect the status of the surrounding buildings 2.

[0125] In Figure 16, the sky is open along the direction of travel, allowing satellites to be seen. However, the range in which satellites can be seen is limited to the side of the direction of travel due to the influence of buildings 2. In such cases, it is possible to increase the degree of lateral variation in the satellite-based position PGS obtained by the satellite positioning unit 10. By increasing the degree of variation only in a limited direction, there is no need to increase the degree of variation in other directions, which can contribute to improving the accuracy of position calculation.

[0126] <First Variation Degree Enlargement Operation Processing> The flowchart in Figure 17 is inserted before step S341 in Figure 10. In step S361, it is determined whether the satellite positions are concentrated in one direction. One direction includes, for example, when satellites are visible only in positions along the direction of travel.

[0127] In step S362, it is determined whether the satellites are concentrated in one direction based on the signal from the GNSS receiver 11. If the satellites are concentrated in one direction (determination is YES), the process proceeds to step S365. If the satellites are not concentrated in one direction (determination is NO), the process proceeds to step S363.

[0128] In step S363, the sky factor Rsky of the current location is obtained from the map information. In step S364, it is determined whether the sky factor Rsky is smaller than a predetermined sky factor threshold Rsky1. If the sky factor Rsky is smaller than the predetermined sky factor threshold Rsky1 (determination is YES), proceed to step S365. If the sky factor Rsky is not smaller than the predetermined sky factor threshold Rsky1 (determination is NO), proceed to step S314 in Figure 10.

[0129] In step S365, the first degree of variation EE1 is multiplied by (1+α), where α is a positive number, and then the process proceeds to step S314 in FIG.

[0130] Fig. 18 is a flowchart of processing for increasing the first degree of variation in the left-right direction by the position calculation device 100 according to Embodiment 4. The flowchart in Fig. 18 is a modified example in which the flowchart in Fig. 17 is partially changed to increase the first degree of variation EE1 in the left-right direction.

[0131] The flowchart in Figure 18 is similar to the flowchart in Figure 17 and is inserted before step S341 in Figure 10. In step S371, the number of receivable satellites in the forward and backward directions is input to Nsat_lg. Then, the number of receivable satellites in the lateral direction is input to Nsat_sd.

[0132] In step S372, it is determined whether the number of receivable satellites in the forward and backward directions, Nsat_lg, is greater than the product of the number of receivable satellites in the lateral directions, Nsat_sd, and a predetermined value, Ks. Ks is a number equal to or greater than 1. If the number of receivable satellites in the forward and backward directions, Nsat_lg, is greater than the product of the number of receivable satellites in the lateral directions, Nsat_sd, and the predetermined value, Ks, proceed to step S375. If the number of receivable satellites in the forward and backward directions, Nsat_lg, is not greater than the product of the number of receivable satellites in the lateral directions, Nsat_sd, and the predetermined value, Ks, proceed to step S373.

[0133] In step S373, the longitudinal sky factor Rsky_lg and the lateral sky factor Rsky_sd are obtained. In step S374, it is determined whether the longitudinal sky factor Rsky_lg is greater than the product of the lateral sky factor Rsky_sd and a predetermined value Kr, where Kr is a number equal to or greater than 1. If the longitudinal sky factor Rsky_lg is greater than the product of the lateral sky factor Rsky_sd and the predetermined value Kr (determination is YES), proceed to step S375. If the longitudinal sky factor Rsky_lg is not greater than the product of the lateral sky factor Rsky_sd and the predetermined value Kr (determination is NO), proceed to step S341 in Figure 10.

[0134] In step S375, the first degree of variation EE1 is increased in the left-right direction, and the process proceeds to step S341 in FIG.

[0135] 5. Embodiment 5 19 is a flowchart of a process for increasing the degree of variation in the position calculation device 100 according to the fifth embodiment. The configuration diagram and hardware configuration diagram of the position calculation device 100 according to the fourth embodiment are the same as those in FIGS. 1 and 2. The position calculation device 100 according to the fourth embodiment can be realized by modifying the software of the position calculation device 100 according to the first embodiment.

[0136] Here, when approaching an area where the degree of variation is predicted to increase, the degree of variation is increased beforehand before entering. By doing so, it is possible to avoid a sudden change in weight when combining the probability distributions of a specific position calculation method, which would have a significant impact on the combined and corrected mobile object position PCMO. This can contribute to improving the reliability of position calculation.

[0137] <Processing that increases the degree of variation> The flowchart in Figure 19 is inserted before step S341 in Figure 10. In step S381, map information is used to check whether the path of the mobile unit includes a tunnel, a passageway inside a building, a road in a forest, or a road surrounded by high-rise buildings. These areas indicate areas where satellite signals are blocked or diffracted, potentially degrading GNSS accuracy.

[0138] In step S382, it is determined whether either of them is present. If either of them is present (determination is YES), the process proceeds to step S383. If neither of them is present (determination is NO), the process proceeds to step S384.

[0139] In step S383, the first degree of variation EE1 is multiplied by (1+α), where α is a positive number, and the process proceeds to step S384.

[0140] In step S384, it is checked from the map information whether there is a tunnel, a passageway inside a building, a road in a forest, a plain, or a desert on the path of the moving object. These are areas where it is difficult for the appearance recognition sensor 21 to grasp information about the appearance of surrounding objects, and indicate areas where the accuracy of position calculation by the appearance recognition sensor 21 may be reduced.

[0141] In step S385, it is determined whether either of them is present. If either of them is present (determination is YES), the process proceeds to step S386. If neither of them is present (determination is NO), the process proceeds to step S341 in FIG.

[0142] In step S383, the second degree of variation EE2 is multiplied by (1+α) to enlarge the error, and the process then proceeds to step S341 in FIG.

[0143] 6. Embodiment 6 Fig. 20 is an explanatory diagram of a case where the first degree of variation is increased in the position calculation device 100 according to the sixth embodiment. Fig. 21 is a flowchart of a process of increasing the first degree of variation in the position calculation device 100 according to the sixth embodiment.

[0144] The configuration diagram and hardware configuration diagram of the position calculation device 100 according to the sixth embodiment are the same as those in Fig. 1 and Fig. 2. The position calculation device 100 according to the sixth embodiment can be realized by modifying the software of the position calculation device 100 according to the first embodiment.

[0145] FIG. 20 shows a moving object 1 passing through a road with buildings 2 on both sides. GD indicates the traveling direction of the moving object. The position calculation device 100 of the moving object 1 can know the layout of the surrounding buildings 2 from map information. In addition, the appearance recognition sensor 21 can also detect the status of the surrounding buildings 2.

[0146] In Figure 20, the sky is open along the direction of travel, allowing the satellite to be seen. However, an obstacle 3 exists to the side of the direction of travel, making it impossible to identify the boundary of the building 2 ahead on the left side, which should have been detected by the appearance recognition sensor 21. In such a case, the obstacle 3 may cause multipath in the GNSS satellite signals, causing a sudden increase in the first degree of variation EE1 in the calculation of the satellite positioning position.

[0147] Therefore, if the second degree of variation EE2 becomes larger than the predetermined second degree of variation threshold EEth2 due to the obstacle 3, a situation may arise in which an operation to enlarge the first degree of variation EE1 as well may be required. The first degree of variation EE1 related to detection by GNSS can be increased in advance. In this way, even if the first degree of variation EE1 actually increases suddenly, it can be prevented from being affected when the probability distribution is synthesized.

[0148] <First Variation Degree Enlargement Operation Processing> The flowchart of Fig. 21 is inserted before step S341 of Fig. 10, similar to the flowchart of Fig. 17. In step S391, it is checked from the map information whether there are any landmarks in the surrounding area that can be detected by the appearance recognition sensor 21.

[0149] In step S392, it is determined whether or not the object is in the vicinity. If the object is in the vicinity (determination is YES), the process proceeds to step S393. If the object is not in the vicinity (determination is NO), the process proceeds to step S341 in FIG. 10 without performing the enlargement operation process for the first degree of variation EE1.

[0150] In step S393, it is determined whether the second degree of variation EE2 is greater than the second degree of variation threshold EEth2. If the second degree of variation EE2 is greater than the second degree of variation threshold EEth2 (determination is YES), the process proceeds to step S394. In this case, it is possible that a landmark cannot be detected due to an obstacle, and the second degree of variation EE2 has increased. In this case, in step S394, the first degree of variation EE1 is increased. Thereafter, the process proceeds to step S341 in FIG. 10.

[0151] In step S393, if the second degree of variation EE2 is not greater than the second degree of variation threshold EEth2 (determination is NO), the process proceeds to step S341 in FIG.

[0152] The coefficients and constants such as α, β, Rsky1, Ks, and Kr mentioned in the first to sixth embodiments may be appropriately determined by experiment or simulation, or may be variable parameters depending on the turning angle, velocity, turning angular velocity, acceleration, turning angular acceleration, etc. of the moving object.

[0153] Although various exemplary embodiments and examples are described in this application, the various features, aspects, and functions described in one or more embodiments are not limited to the application of a particular embodiment, but may be applied to the embodiments alone or in various combinations. Therefore, countless variations not illustrated are contemplated within the scope of the technology disclosed in this specification. For example, this includes cases where at least one component is modified, added, or omitted, or where at least one component is extracted and combined with components of another embodiment.

[0154] Various aspects of the present disclosure are summarized below as appendices.

[0155] (Appendix 1) a satellite positioning unit that calculates a satellite-based position, which is the position of the moving body calculated by receiving a signal from a satellite using a receiver mounted on the moving body, and a first degree of variation when the satellite-based position is calculated; a map reference position calculation unit that calculates a map reference position, which is the position of the moving body calculated by referring to map information from the surrounding environment grasped by an appearance recognition sensor mounted on the moving body, and a second degree of variation when the map reference position is calculated; a self-contained navigation unit that calculates a self-contained navigation position, which is the position of the moving body calculated based on the traveling speed and traveling direction of the moving body derived from the output of a moving state detection sensor mounted on the moving body that detects the moving state of the moving body, and a third degree of variation when the self-contained navigation position is calculated; and a correction calculation unit that generates a first probability distribution that is a probability distribution of the satellite-based position in a position coordinate space based on the first degree of variation and the satellite-based position; generates a second probability distribution that is a probability distribution of the map-referenced position in a position coordinate space based on the second degree of variation and the map-referenced position; generates a third probability distribution that is a probability distribution of the self-contained navigation position in a position coordinate space based on the third degree of variation and the self-contained navigation position; generates a fourth probability distribution by combining the first probability distribution and the second probability distribution; generates a fifth probability distribution by combining the fourth probability distribution and the third probability distribution; and corrects the self-contained navigation position by a corrected moving body position that is the position of the moving body obtained from the fifth probability distribution. (Appendix 2) 2. The position calculation device according to claim 1, wherein the correction calculation unit sets the corrected moving body position to one of the mean, median, or mode calculated from the fifth probability distribution. (Appendix 3) The position calculation device according to claim 1 or 2, wherein the correction calculation unit, at a first time when the latest position of the moving body is calculated by one of the satellite positioning unit or the map reference position calculation unit, calculates an updated movement amount, which is the movement amount of the moving body between a second time when the position of the moving body was previously calculated by the other unit and the first time, based on the position of the moving body at the second time and the first time calculated by the autonomous navigation unit, adds the updated movement amount to the position of the moving body at the second time previously calculated by the other unit, and generates another probability distribution as the latest position of the moving body calculated by the other unit, and generates a fourth probability distribution by combining the first probability distribution and the second probability distribution. (Appendix 4) 4. The position calculation device according to claim 1, wherein, when combining the first probability distribution and the second probability distribution to generate a fourth probability distribution, the correction calculation unit increases the first degree of variation to generate the first probability distribution if the distance between the autonomous navigation position and the satellite positioning position is greater than a predetermined first distance threshold, and increases the second degree of variation to generate the second probability distribution if the distance between the autonomous navigation position and the map reference position is greater than a predetermined second distance threshold. (Appendix 5) 4. The position calculation device according to claim 1, wherein, when generating a fourth probability distribution by combining the first probability distribution and the second probability distribution, if the distance between the satellite positioning position and the map reference position is greater than a predetermined maximum distance threshold, the correction calculation unit compares a satellite positioning distance deviation, which is the distance between the self-contained navigation position and the satellite positioning position, with a map reference position distance deviation, which is the distance between the self-contained navigation position and the map reference position, and generates the first probability distribution by increasing the first degree of variation if the satellite positioning distance deviation is greater than the map reference position distance deviation, and generates the second probability distribution by increasing the second degree of variation if the map reference position distance deviation is equal to or less than the satellite positioning distance deviation. (Appendix 6) 6. The position calculation device according to claim 1, wherein, when combining the first probability distribution and the second probability distribution to generate a fourth probability distribution, if the first degree of variation and the second degree of variation are both smaller than a predetermined variation degree threshold and the distance between the satellite positioning position and the map reference position is smaller than a predetermined minimum distance threshold, the correction calculation unit reduces the first degree of variation to generate the first probability distribution and reduces the second degree of variation to generate the second probability distribution. (Appendix 7) 7. The position calculation device according to claim 6, wherein the correction calculation unit sets the variation degree threshold based on the magnitudes of the first variation degree and the second variation degree. (Appendix 8) 7. A position calculation device according to any one of appendices 1 to 6, wherein the correction calculation unit increases the first degree of variation to generate the first probability distribution based on the positions of the satellites capable of receiving signals determined by the satellite positioning unit or the sky factor determined from the map information by the map reference position calculation unit. (Appendix 9) a correction calculation unit that generates the first probability distribution by increasing the first degree of variation in the left-right direction of the moving body when the number of satellites from which signals can be received by the satellite positioning unit is fewer in the left-right direction of the moving body than in the forward-backward direction, or when the sky factor calculated from the map information by the map reference position calculation unit is fewer in the left-right direction of the moving body than in the forward-backward direction. (Appendix 10) 10. The position calculation device according to claim 1, wherein, when it is determined from the map information of the map reference position calculation unit that the moving body is proceeding into an area where the first degree of variation or the second degree of variation is increasing, the correction calculation unit increases the degree of variation in advance to generate the first probability distribution or the second probability distribution. (Appendix 11) 11. The position calculation device according to claim 10, wherein the correction calculation unit increases a first degree of variation in advance to generate the first probability distribution when the map reference position calculation unit determines from the map information that the moving body is moving into an area where signals from the satellites are blocked or diffracted. (Appendix 12) 11. The position calculation device according to claim 10, wherein the correction calculation unit increases a first degree of variation in advance to generate the first probability distribution when the map reference position calculation unit determines from the map information that the moving object is proceeding into a tunnel, a passage inside a building, a road in a forest, or a road among a group of high-rise buildings. (Appendix 13) 11. The position calculation device according to claim 10, wherein the correction calculation unit increases a second degree of variation in advance to generate the second probability distribution when the map reference position calculation unit determines from the map information that the vehicle is proceeding into an area where it is difficult for the appearance recognition sensor to grasp the surrounding environment. (Appendix 14) 11. The position calculation device according to claim 10, wherein the correction calculation unit increases a second degree of variation in advance to generate the second probability distribution when the map reference position calculation unit determines from the map information that the moving object is proceeding into a tunnel, a passage inside a building, a road in a forest, a plain, or a desert. (Appendix 15) 15. The position calculation device according to any one of claims 1 to 14, wherein the map information of the map reference position calculation unit includes information on the sky exposure factor or building height distribution for each location. (Appendix 16) 16. The position calculation device according to claim 1, wherein the correction calculation unit generates the first probability distribution by increasing the first degree of variation when the second degree of variation is greater than a predetermined second variation degree threshold. (Appendix 17) 17. The position calculation device according to claim 1, wherein the correction calculation unit increases the first degree of variation to generate the first probability distribution when the second degree of variation is greater than a predetermined second variation degree threshold when a landmark that should be recognized by the appearance recognition sensor is present in the surrounding area from the map information of the map reference position calculation unit. [Explanation of symbols]

[0156] 1 moving body, 2 building, 10 satellite positioning unit, 11 GNSS receiver, 20 map reference position calculation unit, 21 appearance recognition sensor, 22 map information storage unit, 30 autonomous navigation unit, 31 IMU, 40 correction calculation unit, 100 position calculation device

Claims

1. a satellite positioning unit that calculates a satellite-based position, which is the position of the moving body calculated by receiving a signal from a satellite using a receiver mounted on the moving body, and a first degree of variation when the satellite-based position is calculated; a map reference position calculation unit that calculates a map reference position, which is the position of the moving body calculated by referring to map information from the surrounding environment grasped by an appearance recognition sensor mounted on the moving body, and a second degree of variation when the map reference position is calculated; a self-contained navigation unit that calculates a self-contained navigation position, which is the position of the moving body calculated based on the traveling speed and traveling direction of the moving body derived from the output of a moving state detection sensor mounted on the moving body that detects the moving state of the moving body, and a third degree of variation when the self-contained navigation position is calculated; and a correction calculation unit that generates a first probability distribution that is a probability distribution of the satellite-based position in a position coordinate space based on the first degree of variation and the satellite-based position; generates a second probability distribution that is a probability distribution of the map-referenced position in a position coordinate space based on the second degree of variation and the map-referenced position; generates a third probability distribution that is a probability distribution of the self-contained navigation position in a position coordinate space based on the third degree of variation and the self-contained navigation position; generates a fourth probability distribution by combining the first probability distribution and the second probability distribution; generates a fifth probability distribution by combining the fourth probability distribution and the third probability distribution; and corrects the self-contained navigation position by a corrected moving body position that is the position of the moving body obtained from the fifth probability distribution.

2. The position calculation device according to claim 1 , wherein the correction calculation unit sets one of a mean value, a median value, and a mode value obtained from the fifth probability distribution as the corrected moving body position.

3. 2. The position calculation device according to claim 1, wherein the correction calculation unit, at a first time when the latest position of the moving body is calculated by one of the satellite positioning unit or the map reference position calculation unit, calculates an updated movement amount, which is the movement amount of the moving body between a second time when the position of the moving body was previously calculated by the other unit and the first time, based on the position of the moving body at the second time and the first time calculated by the autonomous navigation unit, adds the updated movement amount to the position of the moving body at the second time previously calculated by the other unit to generate another probability distribution as the latest position of the moving body calculated by the other unit, and generates a fourth probability distribution by combining the first probability distribution and the second probability distribution.

4. 4. The position calculation device according to claim 1, wherein, when combining the first probability distribution and the second probability distribution to generate a fourth probability distribution, the correction calculation unit increases the first degree of variation to generate the first probability distribution if the distance between the autonomous navigation position and the satellite positioning position is greater than a predetermined first distance threshold, and increases the second degree of variation to generate the second probability distribution if the distance between the autonomous navigation position and the map reference position is greater than a predetermined second distance threshold.

5. 4. The position calculation device according to claim 1, wherein, when generating a fourth probability distribution by combining the first probability distribution and the second probability distribution, if the distance between the satellite positioning position and the map reference position is greater than a predetermined maximum distance threshold, the correction calculation unit compares a satellite positioning distance deviation, which is the distance between the self-contained navigation position and the satellite positioning position, with a map reference position distance deviation, which is the distance between the self-contained navigation position and the map reference position, and generates the first probability distribution by increasing the first degree of variation if the satellite positioning distance deviation is greater than the map reference position distance deviation, and generates the second probability distribution by increasing the second degree of variation if the map reference position distance deviation is equal to or less than the satellite positioning distance deviation.

6. 4. The position calculation device according to claim 1, wherein, when generating a fourth probability distribution by combining the first probability distribution and the second probability distribution, if the first degree of variation and the second degree of variation are both smaller than a predetermined variation degree threshold and the distance between the satellite positioning position and the map reference position is smaller than a predetermined minimum distance threshold, the correction calculation unit reduces the first degree of variation to generate the first probability distribution and reduces the second degree of variation to generate the second probability distribution.

7. The position calculation device according to claim 6 , wherein the correction calculation unit sets the variation degree threshold based on the magnitudes of the first variation degree and the second variation degree.

8. 4. The position calculation device according to claim 1, wherein the correction calculation unit increases the first degree of variation to generate the first probability distribution based on the positions of the satellites capable of receiving signals determined by the satellite positioning unit or the sky factor determined from the map information by the map reference position calculation unit.

9. 4. The position calculation device according to claim 1, wherein the correction calculation unit generates the first probability distribution by increasing the first degree of variation in the left-right direction of the moving body when the number of satellites from which signals can be received by the satellite positioning unit is fewer in the left-right direction of the moving body than in the forward-backward direction, or when the sky factor calculated from the map information by the map reference position calculation unit is fewer in the left-right direction of the moving body than in the forward-backward direction.

10. 4. The position calculation device according to claim 1, wherein, when it is determined from the map information of the map reference position calculation unit that the moving body is proceeding into an area where the first degree of variation or the second degree of variation is increasing, the correction calculation unit increases the degree of variation in advance to generate the first probability distribution or the second probability distribution.

11. 11. The position calculation device according to claim 10, wherein the correction calculation unit increases a first degree of variation in advance to generate the first probability distribution when the map reference position calculation unit determines from the map information that the moving body is moving into an area where signals from the satellites are blocked or diffracted.

12. 11. The position calculation device according to claim 10, wherein, when the map reference position calculation unit determines from the map information that the moving object is proceeding into a tunnel, a passage inside a building, a road in a forest, or a road among a group of high-rise buildings, the correction calculation unit increases a first degree of variation in advance to generate the first probability distribution.

13. 11. The position calculation device according to claim 10, wherein the correction calculation unit increases a second degree of variation in advance to generate the second probability distribution when the map reference position calculation unit determines from the map information that the vehicle is proceeding into an area where it is difficult for the appearance recognition sensor to grasp the surrounding environment.

14. 11. The position calculation device according to claim 10, wherein, when the map reference position calculation unit determines from the map information that the moving object is proceeding into any one of a tunnel, a passage inside a building, a road in a forest, a plain, or a desert, the correction calculation unit increases a second degree of variation in advance to generate the second probability distribution.

15. The position calculation device according to claim 1 , wherein the map information of the map reference position calculation unit includes information on a sky exposure factor or a distribution of building heights for each location.

16. 4. The position calculation device according to claim 1, wherein, when the second degree of variation is greater than a predetermined second degree of variation threshold, the correction calculation unit increases the first degree of variation to generate the first probability distribution.

17. 4. The position calculation device according to claim 1, wherein when the second degree of variation is greater than a predetermined second variation degree threshold when a landmark that should be recognized by the appearance recognition sensor is present in the surrounding area from the map information of the map reference position calculation unit, the correction calculation unit increases the first degree of variation to generate the first probability distribution.

Images