Position estimation method and position estimation system

Abstract

A position estimation system for estimating the position of a vehicle having a magnetic sensor for measuring magnetism acting from the road surface side of the surface forming a driving path comprises: a map DB storing a map (1) corresponding to a road surface magnetic distribution (M2) as the distribution of magnetic quantities at each point on the road surface; a magnetic distribution generating unit acquiring magnetic measurement values ​​obtained by the magnetic sensor and generating a measured magnetic distribution as a distribution of the magnetic measurement values; and a position estimation unit determining an area corresponding to the measured magnetic distribution in the road surface magnetic distribution (M2) corresponding to the map stored in the map DB, and estimating the position of the vehicle based on the position of the area corresponding to the measured magnetic distribution on the map.

Description

Technical Field

[0001] The present invention relates to a position estimating method and a position estimating system for estimating the position of a vehicle on a map. Background Art

[0002] In recent years, various technologies for supporting vehicle driving, such as lane following driving and automatic driving, have been proposed (see, for example, Patent Document 1), and technology development for realization has been prevalent. In order to realize advanced driving support such as automatic driving, a three-dimensional map with high accuracy that represents the driving environment has been proposed and utilized.

[0003] For example, Patent Document 2 below describes a high-precision three-dimensional map including three-dimensional position information of lane markings, curbs, etc. Such a high-precision three-dimensional map is effective for accurately grasping the three-dimensional structure of the driving environment. Accurately grasping the three-dimensional structure of the driving environment is essential for achieving lane following driving, automatic driving, etc.

[0004] Prior art literature

[0005] Patent Literature

[0006] Patent Document 1: Japanese Patent Application Publication No. 2014-34251

[0007] Patent Document 2: Japanese Patent Application Publication No. 2013-186718 Summary of the invention

[0008] Problems to be solved by the invention

[0009] Even if a high-precision three-dimensional map is available, if the accuracy of the position estimation of the host vehicle in the three-dimensional map is insufficient, the high accuracy of the three-dimensional map cannot be fully utilized, and high-precision driving support may not be achieved.

[0010] The present invention has been made in view of the above-mentioned conventional problems, and is intended to provide a method and system for estimating the position of a vehicle on a map.

[0011] Solutions to Solve Problems

[0012] One aspect of the present invention is a position estimation method for estimating the position of a vehicle equipped with a magnetic sensor for measuring magnetism acting from a road surface forming a surface of a travel path, wherein:

[0013] The position estimation method comprises:

[0014] A magnetic measurement process for obtaining a magnetic measurement value obtained by the magnetic sensor;

[0015] a magnetic distribution generation process for generating a measured magnetic distribution indicating a distribution of magnetic data based on the magnetic measurement values ​​obtained by the magnetic measurement process; and

[0016] The position estimation processing determines the area in the road surface magnetic distribution corresponding to the measured magnetic distribution by referring to a map that establishes a correspondence with the road surface magnetic distribution that represents the distribution of magnetic data obtained based on the magnetic quantity of each point on the road surface, and estimates the position of the vehicle based on the position of the area corresponding to the measured magnetic distribution on the map.

[0017] One aspect of the present invention is a position estimation system for estimating the position of a vehicle including a magnetic sensor for measuring magnetism acting from a road surface side forming a surface of a travel path, wherein:

[0018] The position estimation system comprises:

[0019] a storage unit storing a map in which a correspondence is established with a road surface magnetic distribution which is a distribution of magnetic data obtained based on the magnetic quantity of each point on the road surface;

[0020] a magnetic distribution generating unit that acquires the magnetic measurement values ​​obtained by the magnetic sensor and generates a measured magnetic distribution that is a distribution of magnetic data obtained based on the magnetic measurement values; and

[0021] A position estimating unit determines an area corresponding to the measured magnetic distribution in a road surface magnetic distribution associated with a map stored in the storage unit, and estimates the vehicle position based on the position of the area corresponding to the measured magnetic distribution on the map.

[0022] Effects of the Invention

[0023] The present invention is an invention for estimating the position of the vehicle in a map. The map that has established a corresponding relationship with the road surface magnetic distribution can be compared with the measured magnetic distribution obtained on the vehicle side. If the area in the road surface magnetic distribution corresponding to the measured magnetic distribution can be determined, the position of the vehicle can be estimated based on the position on the map of the area corresponding to the measured magnetic distribution. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a diagram for explaining the structure of the map in Example 1.

[0025] Figure 2 This is a front view of the vehicle in Example 1.

[0026] Figure 3 This is a block diagram showing a configuration included in the vehicle in the first embodiment.

[0027] Figure 4 This is a top view of the vehicle in Example 1.

[0028] Figure 5 This is a flowchart showing the steps of the position estimation method in the first embodiment.

[0029] Figure 6 This is an explanatory diagram of the motion estimation process in the first embodiment.

[0030] Figure 7 This is a diagram illustrating the motion estimation result when the vehicle is traveling straight in the first embodiment.

[0031] Figure 8 This is a diagram illustrating the motion estimation result when the vehicle is in a turning state in the first embodiment.

[0032] Fig. 9 This is a diagram illustrating an area where magnetic field distribution is measured when the vehicle is traveling straight in the first embodiment.

[0033] Fig.10 This is a diagram illustrating an area where magnetic field distribution is measured when the vehicle is in a turning state in the first embodiment.

[0034] Fig.11 This is an explanatory diagram of the comparison processing between the transformed magnetic distribution obtained based on the road surface magnetic distribution and the measured magnetic distribution in Example 1.

[0035] Fig.12 This is a diagram illustrating the comparison range of the one-dimensional magnetic distribution in Example 1.

[0036] Fig.13 This is a top view of the road paved with magnetic markers in Example 2.

[0037] Fig.14 This is a flowchart showing the steps of the position estimation method in the second embodiment.

[0038] Fig.15 This is an explanatory diagram illustrating the change in the magnetic meter measurement value in the traveling direction when passing through the magnetic marker in Example 2.

[0039] Fig.16 This is an explanatory diagram illustrating the distribution of magnetic measurement values ​​in the vehicle width direction obtained by the magnetic sensors Cn arranged in the vehicle width direction in the second embodiment.

[0040] Fig.17 This is a diagram illustrating the measured magnetic distribution of the positions where the magnetic markers are plotted in Example 2.

[0041] Fig.18 This is an explanatory diagram of the comparison processing of measuring magnetic distribution in Example 2.

[0042] Fig.19 This is a stereoscopic view of a magnetic marker equipped with an RF-ID tag in Example 3.

[0043] Fig. 20 This is a block diagram showing a configuration included in a vehicle according to the third embodiment.

[0044] Fig.21 This is an explanatory diagram showing the range to which the arrival position of the vehicle after passing the magnetic marker in Example 3 belongs. DETAILED DESCRIPTION

[0045] In the present invention, as magnetic data constituting the road surface magnetic distribution and measuring the magnetic distribution, in addition to the magnetic quantity or the magnetic measurement value itself, the magnetic gradient is preferably used. The magnetic gradient can be obtained, for example, based on the difference in the magnetic quantity or magnetic measurement value at two positions on the road surface. For example, if the magnetism on the road surface is measured by a sensor array formed by arranging magnetic sensors, the magnetic gradient can be obtained, for example, based on the difference in magnetic measurement values ​​obtained by adjacent magnetic sensors. In addition, for example, the temporal magnetic gradient can also be obtained based on the difference in magnetic measurement values ​​at different time points measured by a magnetic sensor. And, for example, it is also considered to obtain such a difference for two magnetic sensors to form a temporal magnetic gradient, and further take a difference between the two magnetic sensors to obtain the magnetic gradient.

[0046] In the magnetic gradient, the magnetic components that act on the magnetic sensor in the same way or nearly in the same way are suppressed. Therefore, in the magnetic gradient, the magnetic components acting from the magnetic sources existing in the farther distance are suppressed, and the magnetic components acting from the magnetic sources existing in the road surface and the like in the nearer distance are relatively emphasized. Therefore, if the magnetic gradient is used as the magnetic data for constituting the road surface magnetic distribution and measuring the magnetic distribution, the influence caused by the magnetic sources such as the surrounding vehicles, guardrails, and signs, and the influence caused by the earth's magnetism can be suppressed, and the magnetic distribution on the road surface can be reflected with high precision.

[0047] The embodiments of the present invention will be specifically described using the following examples.

[0048] (Example 1)

[0049] This example is an example of a position estimation method and a position estimation system 1S for estimating the position of a vehicle on a map with high accuracy. Figures 1 to 12 The content is described.

[0050] The position estimation system 1S is based on the road surface 100S ( Figure 2) to estimate the position of the vehicle (own vehicle position) based on the distribution of magnetic data (magnetic distribution) obtained from the magnetic quantity of each point on the road surface 100S. The road surface 100S forming the surface of the road (an example of a driving road) 100 is formed by paving materials. It is inevitable to mix magnetic materials such as metal powder into the paving materials. The manholes (manholes) and bridge joints (expansion devices) exposed on the road surface 100S are mostly made of metal and may become magnetism generation sources. Therefore, the size of the magnetism at each point on the road surface 100S is not constant, but varies in size.

[0051] The distribution form of the magnetic distribution of the road surface 100S is similar to, for example, the deviation of the brightness of each point in the image obtained by photographing the road surface 100S from directly above with a camera, that is, the distribution form of the brightness distribution. For the magnetic distribution of the road surface 100S, the physical quantity as the object is different from the brightness distribution, but it is the same as the brightness pattern of the road surface 100S represented by the brightness distribution, which represents the magnetic pattern of the road surface 100S. The position estimation system 1S of this example uses the distribution of magnetic data obtained based on the magnetic quantity of each point on the road surface 100S, that is, the road surface magnetic distribution, to estimate the position of the vehicle. It should be noted that in this example, the magnetic quantity itself is used as the magnetic data. Therefore, the road surface magnetic distribution of this example is the distribution of magnetic data as the magnetic quantity of each point on the road surface 100S.

[0052] In the location estimation system 1S, Figure 1 ( Figure 1 ), the road surface magnetic distribution M2 is associated with the structural map M1 indicating the road structure. The structural map M1 and the road surface magnetic distribution M2 are associated with each other through position data indicating absolute positions. The position in the road surface magnetic distribution M2 is uniquely associated with the position on the structural map M1. If an area with the same distribution pattern as the magnetic distribution of the road surface 100S measured by the vehicle can be identified in the road surface magnetic distribution M2, the ground can be estimated based on the position of the area. Figure 1 The position of the vehicle on.

[0053] The position estimation system 1S can be combined with an automatic driving system (not shown) for realizing automatic driving of a vehicle, for example. For example, the automatic driving system uses the position of the vehicle estimated by the position estimation system 1S to grasp the structure of the driving environment ahead. For example, if a three-dimensional map is used, the three-dimensional structure of the driving environment ahead can be grasped, and high-precision automatic driving can be realized.

[0054] The road surface magnetic distribution M2, which represents the distribution of magnetic quantities, i.e., magnetic data, at each point on the road surface 100S, can be generated, for example, using a measuring vehicle (not shown), which has a high-precision positioning system including an RTK-GPS unit, an IMU, etc., and a magnetic measuring unit in which magnetic sensors are arranged along the vehicle width direction. RTK-GPS (Real Time Kinematic Global Positioning System) is a well-known system that uses a reference station to improve the positioning accuracy to em units. IMU (Inertial Measurement Unit) is a well-known unit that uses inertial navigation to estimate the relative position of a vehicle. IMU has an electronic compass, an acceleration sensor, a gyroscope sensor, etc. for measuring the direction, and calculates the relative position relative to the reference position. If the relative position estimated by the IMU is used, high-precision positioning can be achieved even between buildings, tunnels, etc. where the GPS radio waves are unstable.

[0055] The magnetic measuring unit is preferably installed on the vehicle by means of a mechanism that automatically adjusts the height from the road surface 100S to a certain level. If the height of the magnetic measuring unit from the road surface 100S is kept constant, the magnetic distribution of each point on the road surface 100S can be measured with high precision. If it is a measuring vehicle (not shown) with a high-precision positioning system such as an RTK-GPS unit or an IMU, it is possible to generate a road surface magnetic distribution M2 in which high-precision position data and magnetic data of the magnetic quantity of each point on the road surface 100S measured by the magnetic measuring unit are in a corresponding relationship.

[0056] like Figure 2 as well as Figure 3 As shown, the vehicle 5 constituting the position estimation system 1S includes a sensor array 2 in which a plurality of magnetic sensors Cn are arranged in the vehicle width direction, a storage location 2, and a storage location 2. Figure 1 ( Figure 1 ) and a control unit 32 for performing various operations. In addition, the control unit 32 is connected to a GPS unit 351 for performing positioning operations using GPS satellites, a steering angle sensor 353 for detecting the steering direction, a vehicle speed sensor 355, and an altitude sensor 357 for measuring the altitude of the sensor array 2. The map DB 40 is a map database that has established a corresponding relationship with the road surface magnetic distribution M2. Figure 1 An example of a storage unit that performs storage.

[0057] The height sensor 357 is an ultrasonic distance measuring sensor. The height sensor 357 is installed near the sensor array 2 so as to be able to measure the height of the sensor array 2 from the road surface 100S. The height measured by the height sensor 357 is processed as the installation height of the magnetic sensor Cn. It should be noted that the height sensor can also be arranged at two or more positions in the vehicle width direction so as to be able to cope with the inclination of the vehicle body caused by rolling. In this case, the installation heights of multiple magnetic sensors Cn at different positions in the vehicle width direction can be grasped respectively, thereby determining the inclination of the vehicle body. It should be noted that the height sensor 357 can also be a laser distance measuring sensor.

[0058] The sensor array 2 is a thin and long rod-shaped unit having 15 magnetic sensors Cn (n is an integer of 1 to 15) and a detection processing circuit 20 having a CPU (not shown) built in. In the sensor array 2, the 15 magnetic sensors Cn are arranged in a straight line at a certain interval (0.1 m). The sensor array 2 is installed, for example, on the inner side of the front bumper of the vehicle 5 in a state facing the road surface 100S in a manner along the vehicle width direction ( Figure 2 , Figure 4 In the case of the vehicle 5 of this example, the installation height of the sensor array 2 based on the road surface 100S is 200 mm. The magnetic sensor Cn measures the magnetism acting from the road surface 100S side.

[0059] The magnetic sensor Cn is a sensor that detects magnetism by utilizing the well-known MI effect (Magneto Impedance Effect), in which the impedance of a magnetically sensitive body such as an amorphous wire changes sensitively according to an external magnetic field. The magnetic sensor Cn in this example has a highly sensitive detection performance, with a magnetic flux density measurement range of ±0.6 mT and a magnetic flux resolution within the measurement range of 0.02 μT.

[0060] In the magnetic sensor Cn, a magnetically sensitive body is arranged along two orthogonal axial directions, so that the magnetism acting in the two orthogonal axial directions can be detected. In this example, the magnetic sensor Cn is assembled in the sensor array 2 so that the sensor array 2 installed in the measuring vehicle 11 along the vehicle width direction can detect the magnetic components in the travel direction and the vehicle width direction.

[0061] It should be noted that the detection direction of the magnetic component of the magnetic sensor Cn may also be only one direction. In this case, the magnetic sensor may be assembled into the sensor array in a manner to detect the magnetism acting in the vertical direction. A magnetic sensor that detects magnetism in three mutually orthogonal axial directions may also be used.

[0062] Detection processing circuit 20 ( Figure 3) is a circuit that controls the magnetic sensor Cn to perform processing such as detecting magnetism and measuring magnetic quantity. The detection processing circuit 20 is composed of memory elements such as ROM (read only memory) and RAM (random access memory) in addition to the CPU (central processing unit) that performs various operations. The detection processing circuit 20 controls the magnetic sensor Cn to perform magnetic measurement processing at a frequency of 3kHz. The detection processing circuit 20 obtains the magnetic components in the travel direction and the vehicle width direction from each magnetic sensor Cn through magnetic measurement processing.

[0063] The detection processing circuit 20 synthesizes the magnetic component of the traveling direction obtained by each magnetic sensor Cn with the magnetic component of the vehicle width direction. In addition, the detection processing circuit 20 obtains the magnitude of the magnetism acting along the horizontal plane defined by the traveling direction and the vehicle width direction for each magnetic sensor Cn, and outputs the magnetic measurement value of each magnetic sensor Cn, i.e., 15 magnetic data, to the outside. Here, in this example, the magnetic measurement value of the magnetic sensor Cn itself is used as the magnetic data.

[0064] As described above, the sensor array 2 is installed on the vehicle 5 in a manner that 15 magnetic sensors Cn are arranged at intervals of 0.1 m along the vehicle width direction. The magnetic data output by the sensor array 2 as 15 magnetic measurement values ​​of the magnetic sensors Cn represent the magnetic distribution (hereinafter referred to as one-dimensional magnetic distribution) of a one-dimensional range of 1.5 m (0.1×15) width in the vehicle width direction.

[0065] The control unit 32 is used to execute Figure 1 (Refer to Figure 1 The control unit 32 includes a unit for calculating the position of the vehicle on the surface of the vehicle. The control unit 32 includes an electronic substrate (not shown) on which memory elements such as ROM and RAM are mounted, in addition to a CPU for performing various calculations. The control unit 32 includes functions as a motion estimation unit 321, a magnetic distribution generation unit 323, a conversion processing unit 325, and a position estimation unit 327.

[0066] The motion estimation unit 321 performs motion estimation processing to estimate the motion of the vehicle 5. The motion estimation unit 321 estimates the motion of the vehicle 5 by decomposing the motion of the vehicle 5 into a translational motion component that is not accompanied by a change in orientation and a rotational motion component that is a change in orientation. For details, refer to Figure 6 To be described later.

[0067] The magnetic distribution generation unit 323 performs a magnetic distribution generation process for generating a two-dimensional measured magnetic distribution based on the one-dimensional magnetic distribution obtained from the sensor array 2. The magnetic distribution generation unit 323 generates a measured magnetic distribution using the motion of the vehicle 5 estimated by the motion estimation unit 321. The measured magnetic distribution is a two-dimensional distribution of magnetic measurement values ​​obtained by the magnetic measurement process.

[0068] The conversion processing unit 325 performs the operation for converting Figure 1 The corresponding road surface magnetic distribution M2 is converted into a magnetic distribution suitable for comparison with the measured magnetic distribution. The conversion processing unit 325 converts the road surface magnetic distribution M2 into a magnetic distribution at the installation height of the magnetic sensor Cn (appropriately referred to as a converted magnetic distribution).

[0069] The position estimation unit 327 performs local Figure 1 The position estimation unit 327 refers to the map. Figure 1 The area corresponding to the measured magnetic distribution is determined, and the corresponding area is based on the Figure 1 Specifically, the position estimation unit 327 determines the position of the vehicle. Figure 1 The area in the road surface magnetic distribution M2 corresponding to the measured magnetic distribution.

[0070] The vehicle 5 repeatedly executes Figure 5 The processing flow is used to estimate the position of the vehicle. During the travel of the vehicle 5, the sensor array 2 performs magnetic measurement processing to obtain a one-dimensional magnetic distribution, and inputs the one-dimensional magnetic distribution to the control unit 32 (S101). The control unit 32 estimates the motion of the vehicle 5 based on the measured vehicle speed and the measured steering angle (S103, motion estimation processing), and estimates the displacement of the position of the one-dimensional area on the road surface facing the sensor array 2.

[0071] The control unit 32 accumulates the one-dimensional magnetic distribution obtained by the sensor array 2 according to the displacement of the position of the one-dimensional area on the road surface facing the sensor array 2 to generate a two-dimensional measured magnetic distribution (S105, magnetic distribution generation processing). Figure 1 . ) The road surface magnetic distribution M2 with a corresponding relationship is converted into the magnetic distribution of the installation height of the magnetic sensor Cn, that is, the converted magnetic distribution (S107, conversion processing). Then, the control unit 32 determines the corresponding area of ​​the measured magnetic distribution in the road surface magnetic distribution M2, thereby estimating the position of the vehicle (S109, position estimation processing).

[0072] Next, the composition is described. Figure 5 The processing flow includes (A) magnetic measurement processing, (B) motion estimation processing, (C) magnetic distribution generation processing, (D) transformation processing, and (E) position estimation processing.

[0073] (A) Magnetic measurement processing

[0074] Magnetic measurement processing is processing performed by the sensor array 2. The sensor array 2 performs magnetic measurement processing performed by 15 magnetic sensors Cn at a frequency of 3kHz. The sensor array 2 obtains 15 magnetic measurement values ​​obtained by the magnetic sensor Cn at a frequency of 3kHz, and inputs the 15 magnetic measurement values, i.e., magnetic data, to the control unit 32 at the same frequency. As described above, the 15 magnetic measurement values, i.e., magnetic data, input by the sensor array 2 to the control unit 32 are a one-dimensional distribution of discrete values ​​along the vehicle width direction. The width of the one-dimensional magnetic distribution as the one-dimensional discrete value distribution is 1.5m, which is equal to the range in which the magnetic sensors Cn are configured in the sensor array 2. It should be noted that the sensor array 2 associates identification information with the one-dimensional magnetic distribution and inputs it to the control unit 32. The identification information associated with the one-dimensional magnetic distribution can be used to identify the one-dimensional magnetic distribution that constitutes the measured magnetic distribution.

[0075] (B) Motion estimation processing

[0076] The motion estimation process is a process executed by the control unit 32 (motion estimation unit 321) to estimate the motion of the vehicle 5. The control unit 32 estimates the motion of the vehicle 5 using the measured vehicle speed obtained by the vehicle speed sensor 355 and the measured steering angle obtained by the steering angle sensor 353. As described later, the control unit 32 of this example estimates the motion of the vehicle 5 by decomposing the vehicle 5 into a translational motion component and a rotational motion component.

[0077] The control unit 32 estimates the movement of the vehicle 5 during the interval (1 / 3000 second) of the magnetic measurement processing performed at a frequency of 3kHz by measuring the vehicle speed and the steering angle. The movement of the vehicle 5 is the movement of moving forward at the speed of the measured vehicle speed during a period of 1 / 3000 second along the steering direction determined by the measured steering angle. It should be noted that the control unit 32 of this example sets the center of the sensor array 2 (the position of the magnetic sensor C8) as the representative point of the vehicle 5, and estimates the movement of the center of the sensor array 2.

[0078] When the vehicle 5 is traveling on a curve, the movement of the center of the sensor array 2 is, for example, Figure 6 The motion Mo of the vehicle 5 is shown by the arc-shaped arrow in FIG. The motion Mo can be decomposed into the vector V in the figure representing the component of translational motion and the turning angle R representing the component of rotational motion. In the motion estimation process, the motion Mo of the vehicle 5 is estimated by decomposing it into the vector V and the turning angle R. If the vector V and the turning angle R are determined based on a certain position of the vehicle, the position and posture of the vehicle after displacement are uniquely determined.

[0079] As described above, in this example, the center of the sensor array 2 is set as the representative point of the vehicle 5. Therefore, the displacement of the position of the one-dimensional area on the road surface 100S facing the sensor array 2 and the one-dimensional magnetic distribution as the magnetic distribution of the one-dimensional area is consistent with the motion Mo of the vehicle 5. The displacement of the position of the one-dimensional area on the road surface 100S facing the sensor array 2 and the one-dimensional magnetic distribution can be determined by the vector V estimated by the above-mentioned motion estimation processing and the turning angle R.

[0080] (C) Magnetic distribution generation process

[0081] The magnetic distribution generation process is a process performed by the control unit 32 (magnetic distribution generation unit 323) to generate the magnetic distribution on the road surface 100S measured using the sensor array 2, i.e., the measured magnetic distribution. The control unit 32 takes in the above-mentioned one-dimensional magnetic distribution consisting of 15 magnetic measurement values, i.e., magnetic data, of the sensor array 2 at a frequency of 3 kHz, and generates a two-dimensional measured magnetic distribution.

[0082] The measured magnetic distribution is obtained by scanning the road surface 100S by the sensor array 2 of the vehicle 5 like a "line scanner". The area of ​​the measured magnetic distribution is the area that has been scanned by the sensor array 2. When the control unit 32 obtains a new one-dimensional magnetic distribution and incorporates it into the measured magnetic distribution, it eliminates the one-dimensional magnetic distribution with the earliest acquisition time point in the measured magnetic distribution. According to such steps, the magnetic distribution of a two-dimensional area with a width of 1.5m that starts from the one-dimensional area on the road surface 100S facing the sensor array 2 and extends along the driving track of the vehicle 5 can be generated as a measured magnetic distribution. It should be noted that in this example, the number of inclusions of the one-dimensional magnetic distribution is made variable according to the vehicle speed in such a way that the dimension of the long side direction of the measured magnetic distribution (equivalent to the direction of travel of the vehicle) becomes a specified distance. The dimension of the long side direction of the measured magnetic distribution is preferably about 2m to 10m, for example.

[0083] Regarding the positional relationship between the existing measured magnetic distribution A and the newly incorporated one-dimensional magnetic distribution L2 when the new one-dimensional magnetic distribution is incorporated into the measured magnetic distribution, use Figure 7 as well as Figure 8 The positional relationship between the latest one-dimensional magnetic distribution L1 in the existing measured magnetic distribution A and the newly included one-dimensional magnetic distribution L2 can be expressed by the vector V and the turning angle R estimated by the motion estimation process.

[0084] Figure 7 The positional relationship between the one-dimensional magnetic distributions L1 and L2 is shown for example when the vehicle 5 travels straight. In this case, the position of the one-dimensional magnetic distribution L2 newly included in the measured magnetic distribution A is determined by a vector V orthogonal to the longitudinal direction of the one-dimensional magnetic distribution L1. Figure 8The positional relationship between L1 and L2 is shown when the vehicle 5 is traveling on a curve. In this case, the position and orientation (attitude) of the one-dimensional magnetic distribution L2 are determined by rotating the one-dimensional range Lp determined by the vector V oblique to the one-dimensional magnetic distribution L1 by the amount of the turning angle R. Figure 8 In the figure, dirl represents a direction orthogonal to the one-dimensional magnetic distribution L1 and the one-dimensional range Lp, and dir2 represents a direction orthogonal to the one-dimensional magnetic distribution L2. The turning angle R is the angle between dirl and dir2.

[0085] When the vehicle 5 is traveling straight, Fig. 9 As shown in FIG. 1 , a rectangular measurement magnetic distribution A is formed with the one-dimensional magnetic distribution L2 of the area facing the sensor array 2 as the end and extending along the running track H of the vehicle 5. When the vehicle 5 is traveling on a curve, as shown in FIG. Fig.10 As shown, a measured magnetic distribution A is formed which extends in a curved shape along the travel trajectory H of the vehicle 5 with the one-dimensional magnetic distribution L2 in the region facing the sensor array 2 as an end.

[0086] (D) Transformation Processing

[0087] The transformation process is to transform the Figure 1 (Refer to Figure 1 The control unit 32 performs a simulation to estimate the distribution of magnetic quantities acting from the road surface to each point at a predetermined height. By performing this simulation, the control unit 32 converts the road surface magnetic distribution M2 into the magnetic distribution at the installation height of the magnetic sensor Cn, i.e., the converted magnetic distribution M2D (refer to Fig.11 . ).

[0088] (E) Position estimation processing

[0089] The position estimation process is a process performed by the control unit 32 (position estimation unit 327) to estimate the vehicle position on the map. The control unit 32 compares the transformed magnetic distribution M2D obtained based on the road surface magnetic distribution M2 with the measured magnetic distribution A ( Fig.11 ), thereby determining the area in the road surface magnetic distribution M2D that corresponds to the measured magnetic distribution A. Specifically, the control unit 32 performs correlation operations on the transformed magnetic distribution M2D and the measured magnetic distribution A, and determines the area in the transformed magnetic distribution M2D that has a high degree of consistency with the measured magnetic distribution A. It should be noted that if the correlation operations on the transformed magnetic distribution M2D and the measured magnetic distribution A are to be performed on the entire area of ​​the road surface 100S, the amount of calculation may become huge. If the positioning position obtained by the GPS unit 351 is referenced and the comparison of the measured magnetic distribution A is performed within a range near the positioning position, the amount of calculation required for the above-mentioned correlation operations can be suppressed.

[0090] The position estimation process is a process of calculating the normalized mutual correlation coefficient by the correlation operation of Mathematical Formula 1. Mathematical Formula 1 is a formula for calculating the mutual correlation coefficient, which represents the consistency between the transformed magnetic distribution M2D and the measured magnetic distribution A when the measured magnetic distribution A and the area represented by specific coordinates (u, v) in the transformed magnetic distribution M2D are overlapped. In particular, Mathematical Formula 1 is a calculation formula for the normalized mutual correlation coefficient, and the maximum mutual correlation coefficient 1 is obtained when the transformed magnetic distribution M2D and the measured magnetic distribution A are consistent.

[0091] (Mathematical formula 1)

[0092]

[0093]

[0094]

[0095] Here, f represents the converted magnetic distribution M2D, and t represents the measured magnetic distribution A. The size of the region of the measured magnetic distribution A is Nx×Ny, and the size of the region of the converted magnetic distribution M2D is sufficiently larger than that of the measured magnetic distribution A.

[0096] In equation 1, changing (u, v) means changing the area overlapping with the measured magnetic distribution A in the transformed magnetic distribution M2D. Calculating the correlation coefficient for various coordinates (u, v) is equivalent to, for example, Fig.11 As shown, the consistency is checked while the position of the measured magnetic distribution A is moved relative to the transformed magnetic distribution M2D. The area corresponding to the coordinates (u0, v0) can be determined as the corresponding area of ​​the measured magnetic distribution A. The coordinates (u0, v0) are the coordinates when the mutual correlation coefficient closest to the maximum correlation value, i.e., 1, is obtained among the mutual correlation coefficients calculated for various coordinates (u, v). If the corresponding area of ​​the measured magnetic distribution A in the transformed magnetic distribution M2D obtained based on the road surface magnetic distribution M2 can be determined, the position of the vehicle in the road surface magnetic distribution M2 can be estimated. The road surface magnetic distribution M2 is compared with the structural map M1 (refer to Figure 1 Estimating the vehicle position in the road surface magnetic distribution M2 is synonymous with estimating the vehicle position on the map.

[0097] It is also possible to rotate the measured magnetic distribution A and calculate the mutual correlation coefficient for each rotation angle when performing the correlation operation on each coordinate (u, v). At this time, it is preferable to assume that the driving direction of the vehicle 5 is angularly offset relative to the direction of the road 100, such as a lane change, and limit the range of the rotation angle for performing the mutual correlation. If the measured magnetic distribution A is rotated and the mutual correlation coefficient is calculated, the position of the vehicle can be estimated not only when the vehicle 5 is traveling along the road 100, but also when the lane change is in progress. When the mutual correlation coefficient of each coordinate and each rotation angle is calculated, not only the position of the vehicle 5 can be estimated, but also the orientation (attitude) of the vehicle 5 can be estimated.

[0098] As described above, the position estimation system 1S of this example estimates the vehicle position on the map using the magnetic distribution of each point on the road surface 100S, that is, the road surface magnetic distribution M2. If the vehicle position on the map can be estimated, various driving support controls including lane following control and automatic driving can be implemented.

[0099] In the composition Figure 1 In the road surface magnetic distribution M2, the magnetic amount is large or small due to the magnetic material mixed in the paving material forming the road surface 100S, the metal working opening set on the road surface 100S, the joints of the bridge and other magnetic generating sources. These magnetic generating sources are fixed in position on the road surface 100S, so the position change is small. If the road surface magnetic distribution M2 originating from the magnetic generating source whose position is fixed is used, the position of the vehicle on the map can be estimated with high accuracy.

[0100] It should be noted that, as described above, when the position estimation process is initially performed, it is effective to use, for example, the positioning position obtained by the GPS unit 351. If the positioning position is used, the range of the comparison of the measured magnetic distribution A in the road surface magnetic distribution M2 can be limited, and the amount of calculation required for the correlation calculation between the transformed magnetic distribution M2D and the measured magnetic distribution A can be suppressed. Furthermore, it is preferred that after estimating the position of the vehicle, the position of the vehicle after displacement is predicted by estimating the movement of the vehicle 5. If the position of the vehicle after displacement is predicted, the comparison range of the measured magnetic distribution A can be limited in the road surface magnetic distribution M2 (transformed magnetic distribution M2D), which can improve the efficiency of the correlation calculation.

[0101] Furthermore, it is preferred that after estimating the position of the vehicle, for example Fig.12 As shown in FIG. 1 , a range RS of candidates for the position of the vehicle (in this example, the center of the sensor array 2) is set based on the estimated position P of the vehicle. This range RS is a candidate for the area that the sensor array 2 faces after displacement, so the one-dimensional magnetic distribution obtained by the magnetic measurement processing can be compared within this range RS. Fig.12The vehicle position can be estimated by comparing the one-dimensional magnetic distribution within the range RS. The correlation calculation of the one-dimensional magnetic distribution is less computationally intensive and more efficient than the correlation calculation of the two-dimensional magnetic distribution.

[0102] If the road surface magnetic distribution M2 (transformed magnetic distribution M2D) is compared with the measured magnetic distribution A and the one-dimensional magnetic distribution obtained by the vehicle 5, the position and posture of the measured magnetic distribution A and the one-dimensional magnetic distribution in the road surface magnetic distribution M2 can be determined. For example, if the position of the vehicle can be estimated, the change of the relative position and orientation (the direction and posture of the vehicle) based on the estimated position of the vehicle can be grasped by determining the position and posture of the measured magnetic distribution A and the one-dimensional magnetic distribution in the road surface magnetic distribution M2. In this way, the structure of estimating the position of the vehicle 5 by comparing the road surface magnetic distribution M2 (transformed magnetic distribution M2D) with the measured magnetic distribution A and the one-dimensional magnetic distribution can replace the IMU.

[0103] In this example, a structure using the positioning position obtained by the GPS unit 351 is illustrated in order to suppress the amount of calculation of the position estimation process and implement the position estimation process efficiently. Instead of using GPS, a communication unit such as a radio beacon or an infrared beacon installed on the roadside can be used. If the vehicle 5 is equipped with a receiving unit, the rough position of the vehicle can be grasped based on the reception of the beacon radio wave. Furthermore, the position of the vehicle can be roughly grasped by processing the captured image obtained by the front camera to identify the name of the intersection, the place name, etc. Alternatively, the position of the vehicle can be roughly grasped by comparing it with the foreground image stored in the vehicle-mounted database or a database that can be accessed via the Internet. If the position of the vehicle can be roughly grasped, the amount of calculation required for related calculations in the position estimation process can be suppressed.

[0104] In this example, an automatic driving system is illustrated as a system combined with the position estimation system 1S. Instead of the automatic driving system, a departure warning system for warning of departure from a lane or a lane keeping system for generating a steering assist force for automatically steering the steering wheel along the lane or avoiding departure from the lane may be applied.

[0105] It should be noted that if the vehicle 5 can be connected to a communication line such as the Internet, the server device can also have the function of the map DB 40. Preferably, the vehicle 5 sends the information required for estimating the position of the vehicle to the server device. The server device can also have the function of performing magnetic distribution generation processing and position estimation processing. In this case, each time the magnetic measurement processing is performed, a one-dimensional magnetic distribution is sent from the vehicle 5 to the server device. The server device can use the one-dimensional magnetic distribution received from the vehicle to estimate the vehicle position of the source vehicle and reply to the vehicle position.

[0106] In this example, based on the road surface magnetic distribution M2, the magnetic distribution at the installation height of the magnetic sensor Cn is converted (converted magnetic distribution), and compared with the measured magnetic distribution. Alternatively, the measured magnetic distribution may be converted to the magnetic distribution of the road surface 100S, and compared with the road surface magnetic distribution M2. Alternatively, the road surface magnetic distribution M2 may be directly compared with the measured magnetic distribution without converting it.

[0107] In order to convert the road surface magnetic distribution M2 into the magnetic distribution of the installation height of the magnetic sensor Cn, the installation height of the magnetic sensor Cn is measured using the height sensor 357. Then, the road surface magnetic distribution M2 is converted into the magnetic distribution of the installation height of the magnetic sensor Cn by simulation taking the installation height into consideration, thereby obtaining the converted magnetic distribution. Instead of the height sensor 357, a storage unit for storing the installation height of the magnetic sensor Cn as a set value may be provided.

[0108] It is also possible to configure two or more magnetic sensors with different heights along the vertical direction. In this case, the ratio (attenuation rate, etc.) of the magnetic measurement values ​​obtained by the two magnetic sensors configured along the vertical direction can be determined. In addition, by performing a simulation operation including using the ratio of the magnetic measurement values ​​of two magnetic sensors with a known height difference as a variable, the road surface magnetic distribution M2 can be transformed into the magnetic distribution of the installation height of the magnetic sensor Cn.

[0109] Furthermore, for example, a magnetic marker with a known magnetic quantity may be provided on the road surface 100S. In this case, it is preferred to obtain the attenuation rate of the magnetic quantity measured by the magnetic sensor Cn for the magnetic marker. If the attenuation rate of the magnetic quantity is used, it can be converted into the magnetic distribution of the installation height of the magnetic sensor Cn based on the road surface magnetic distribution.

[0110] It should be noted that the road surface may also be paved with paving materials made of magnetic materials such as magnetic powder mixed with iron oxide. In this case, the magnetic material in the paving material is magnetized, and the fluctuation of the road surface magnetic distribution M2 tends to become larger. If the fluctuation of the magnetic distribution increases, the comparison of the magnetic distribution becomes easier and the accuracy can be improved. It should be noted that the road surface paved with paving materials mixed with magnetic materials may also be magnetized unevenly. In this case, the fluctuation of the magnetic distribution can be further expanded, and the comparison of the magnetic distribution becomes easier. It is also possible to prepare paving materials mixed with magnetic materials and ordinary paving materials that do not contain magnetic materials. In this case, the two paving materials may be unevenly supplied to the road surface instead of being mixed. Since the composition ratio of the magnetic material differs depending on the position of the road surface, the fluctuation of the magnetic distribution can be expanded.

[0111] The road surface may also be magnetized after being paved with paving materials mixed with magnetic materials. For example, the road surface may be magnetized in a prescribed pattern such as a checkered pattern in which rectangular areas with stronger magnetism and rectangular areas with weaker magnetism appear alternately. If a pattern with strong and weak magnetism is formed, it becomes easy to compare the magnetic distributions, and the accuracy of position estimation can be improved. In addition, when the road surface is magnetized in a prescribed pattern, for example, by using a dividing line between a rectangular area with a larger magnetic amount and a rectangular area with a smaller magnetic amount, the comparison of the measured magnetic distribution can be performed efficiently. For example, when the measured magnetic distribution includes a dividing line, the comparison can be performed on the premise of consistency with the dividing line in the road surface magnetic distribution M2. The prescribed pattern may also be a pattern that can read information, such as a one-dimensional or two-dimensional barcode.

[0112] Furthermore, the road surface 100S may be magnetized in such a way that magnetic singularities appear in the road surface 100S paved with the paving material mixed with the magnetic material. In this way, if the magnetic singularities are made of magnetic powder of iron oxide or the like, it becomes easy to compare the magnetic distributions. If the consistency of the singularities is a premise, when comparing the magnetic distributions, combinations with inconsistent singularities can be excluded, and comparison can be performed efficiently.

[0113] In addition, it is also possible to prepare paving materials mixed with magnetic materials and ordinary paving materials that do not contain magnetic materials. In this case, the type of paving material can also be switched for each area. In this case, the density of magnetic materials is formed for each area, so that the fluctuation of magnetic distribution can be expanded, which can make the comparison of magnetic distribution easier.

[0114] In this example, a comparison of the two-dimensional magnetic distribution, i.e., the measured magnetic distribution, which is the accumulation of the one-dimensional magnetic distribution obtained by the magnetic measurement process is performed. It is also possible to compare the temporal distribution of the magnetic measurement values ​​obtained by any one of the magnetic sensors Cn, i.e., the one-dimensional distribution along the driving trajectory of the vehicle. It is also possible to use a sensor array formed by a two-dimensional arrangement of magnetic sensors. In this case, it is possible to compare the two-dimensional magnetic distribution measured by the sensor array. Alternatively, it is also possible to prepare a plurality of sensor arrays of the same specifications as the above-mentioned sensor array 2 and to arrange them at a plurality of positions in the direction of travel of the vehicle. In this case, it is possible to obtain a plurality of one-dimensional magnetic distributions whose relative positional relationship is known. It is also possible to compare the plurality of one-dimensional magnetic distributions. The plurality of sensor arrays can be installed in parallel with each other or in an L-shaped or cross-shaped manner.

[0115] It should be noted that the road surface magnetic distribution can also be set as the distribution of the magnetic gradient as an example of magnetic data obtained based on the magnetic quantity of each point on the road surface. The magnetic gradient can be obtained, for example, as the difference in the magnetic quantities of adjacent points. In addition, the measured magnetic distribution compared with the road surface magnetic distribution of the magnetic gradient is preferably set as the distribution of the magnetic gradient obtained based on the magnetic measurement value. The magnetic gradient can be obtained, for example, as the difference in the magnetic measurement values ​​obtained by the adjacent magnetic sensors in the magnetic sensors Cn constituting the sensor array 2.

[0116] In the magnetic gradient, the magnetic components that act in the same or nearly the same way on the magnetic sensor are suppressed. Therefore, in the magnetic gradient, the magnetic components acting from the magnetic generating sources existing in the farther distance are suppressed, and the magnetic components acting from the magnetic generating sources existing in the closer road surface are relatively emphasized. Therefore, when estimating the position using the magnetic pattern of the road surface facing the magnetic sensor, it is preferable to use the distribution of the magnetic gradient. If the distribution of the magnetic gradient is used, the influence caused by the magnetic generating sources such as the surrounding vehicles, guardrails, signs, etc., and the influence caused by the geomagnetism can be suppressed, and the accuracy of the position estimation can be improved.

[0117] It should be noted that, for example, the difference of the magnetic measurement values ​​obtained at different time points can also be converted into a difference per unit distance determined based on the vehicle speed information given separately. It is also possible to pre-calculate the difference per unit distance for the two magnetic sensors, and further take the difference between the two magnetic sensors to calculate the magnetic gradient. In this case, the following effect can be expected: the above-mentioned magnetic gradient can be calculated without pre-calculating the interval of the magnetic markers arranged on the road. And, for example, when calculating the magnetic gradient obtained by the difference of the magnetic measurement values ​​at different time points based on one magnetic sensor, the magnetic gradient obtained by the difference per unit distance determined by the vehicle speed information given separately can be calculated. In this case, the necessity of setting two vehicle-mounted sensor units in front and behind becomes less. If the magnetic gradient can be calculated using one sensor unit, the following effect can be expected: while being able to reduce the burden on the design for ensuring the mounting position of the sensor unit, the mounting cost of the sensor unit can be reduced.

[0118] (Example 2)

[0119] This example is an application example of the position estimation system 1S to a road 100 paved with magnetic markers 10. Figure 3 , Figure 13 to Figure 18 Provide explanation.

[0120] like Fig.13As shown, on the illustrated road 100, magnetic markers 10 are laid every 10 m, for example. When a vehicle travels on the road 100, the sensor array 2 performs magnetic measurement processing (S101), and the control unit 32 performs motion estimation processing (S103), magnetic distribution generation processing (S105), and position estimation processing (S109) as in the first embodiment. Fig.14 ). The main difference from the first embodiment is that the sensor array 2 performs a marker detection process (S102) in addition to the magnetic measurement process (S101). The sensor array 2 performs the marker detection process using the magnetic sensor Cn at a frequency of 3 kHz.

[0121] As described in the first embodiment, the magnetic sensor Cn is configured to measure the magnetic components in the traveling direction and the vehicle width direction of the vehicle. For example, when the magnetic sensor Cn moves in the traveling direction and passes directly above the magnetic marker 10, the magnetic measurement value in the traveling direction is as follows: Fig.15 As shown, the positive and negative values ​​are reversed in front of and behind the magnetic marker 10 and change in a manner that crosses zero at the position directly above the magnetic marker 10. Therefore, when the magnetic measurement value of the traveling direction detected by any magnetic sensor Cn generates a zero crossing Zc with a positive and negative reversal during the driving of the vehicle, it can be determined that the sensor array 2 is located directly above the magnetic marker 10. When the sensor array 2 is located directly above the magnetic marker 10 and a zero crossing Zc of the magnetic measurement value of the traveling direction is generated, the detection processing circuit 212 determines that the magnetic marker 10 is detected.

[0122] In addition, for example, regarding a magnetic sensor of the same specification as the magnetic sensor Cn, try to imagine the movement along an imaginary line in the vehicle width direction passing directly above the magnetic marker 10. The magnetic measurement value in the vehicle width direction changes in a manner such that the positive and negative are reversed on both sides of the magnetic marker 10 and cross zero at the position directly above the magnetic marker 10. In the case of a sensor array 2 in which 15 magnetic sensors Cn are arranged in the vehicle width direction, the positive and negative signs of the magnetic measurement values ​​in the vehicle width direction detected by the magnetic sensor Cn differ depending on which side the magnetic marker 10 is located on ( Fig.16 ).

[0123] If the magnetic measurement values ​​in the vehicle width direction of each magnetic sensor Cn of the sensor array 2 are Fig.16 The vehicle width direction position of the magnetic marker 10 is the position in the middle of two adjacent magnetic sensors Cn across the zero crossing Zc of the magnetic measurement value in the vehicle width direction, or the position directly below the magnetic sensor Cn where the magnetic measurement value detected in the vehicle width direction is zero and the magnetic measurement values ​​of the two outer magnetic sensors Cn are inverted. The vehicle width direction position indicates the relative position of the vehicle 5 with respect to the magnetic marker 10.

[0124] When the sensor array 2 detects the magnetic marker 10 as described above, it transmits marker detection information including information such as the detected purpose and the position in the vehicle width direction to the control unit ( Figure 3 The position in the vehicle width direction is represented as a one-dimensional position in the one-dimensional magnetic distribution obtained by the magnetic measurement process. The sensor array 2 associates the identification information of the one-dimensional magnetic distribution when the magnetic marker 10 is detected with the marker detection information.

[0125] The control unit that executes the magnetic distribution generation process (S105) plots the position of the magnetic marker 10 in the measured magnetic distribution A as a marker arrangement point 10P ( Fig.17 ). In the measured magnetic distribution A formed by two-dimensionally accumulating the one-dimensional magnetic distribution along the vehicle's driving track, on which one-dimensional magnetic distribution the marker configuration point 10P is located can be determined by the identification information of the one-dimensional magnetic distribution associated with the marker detection information. In addition, the position of the marker configuration point 10P in the measured magnetic distribution A in the vehicle width direction can be determined by the position in the vehicle width direction contained in the marker detection information.

[0126] Here, in the road surface magnetic distribution M2 of the object to be compared with the measured magnetic distribution A in this example, the position of the magnetic marker 10 is drawn as a marker configuration point 10P, similarly to the above-mentioned measured magnetic distribution A. The road surface magnetic distribution M2 can be obtained by a measurement vehicle that can perform the above-mentioned marker detection processing in addition to the function of the measurement vehicle illustrated in Example 1. And, in the magnetic distribution obtained by transforming the road surface magnetic distribution M2, that is, the transformed magnetic distribution M2D, the marker configuration point 10P is drawn at the same position as the marker configuration point in the road surface magnetic distribution M2.

[0127] The control unit that performs the position estimation process (S109) compares the transformed magnetic distribution M2D at which the marker arrangement point 10P is drawn. Fig.17 The measured magnetic distribution A is shown in FIG. Fig.17 As shown, in the measured magnetic distribution A, the marker configuration point 10P indicating the position of the magnetic marker 10 is drawn. When the measured magnetic distribution A is compared with the transformed magnetic distribution M2D, as shown in FIG. Fig.18 As shown, it is preferred to find the mutual correlation coefficient at each position every 10m where the marker configuration point 10P is consistent. In the case where there is a combination of the transformed magnetic distribution M2D drawn at the marker configuration point 10P and the measured magnetic distribution A like this, the correlation operation can be performed only at the position where the marker configuration point P is consistent with each other. That is, for the transformed magnetic distribution M2D, the position of the measured magnetic distribution A can be offset every 10m, which can significantly reduce the amount of calculation used for the correlation operation.

[0128] That is, in this case, it is preferred to determine the area corresponding to the measured magnetic distribution A from the area where the marker arrangement point 10P (the position of the magnetic marker 10) in the road surface magnetic distribution M2 or the transformed magnetic distribution M2D is consistent. If it is limited to the area where the marker arrangement point 10P is consistent, the number of times the measured magnetic distribution A is compared can be significantly reduced, and the area corresponding to the measured magnetic distribution A can be efficiently determined. The marker arrangement point 10P can be used as a magnetic singular point in the magnetic distribution.

[0129] It should be noted that other structures and effects are the same as those in Example 1.

[0130] (Example 3)

[0131] This example is an example of a position estimation system based on the first embodiment, and is an application example to a road on which magnetic markers 10 are laid so that the absolute position can be determined. Figure 19 to Figure 21 Provide explanation.

[0132] In this example, a magnetic marker 10 with an RF-ID tag 15 attached to the surface is laid on the road surface ( Fig.19 In addition to the structure of the first embodiment, the vehicle 5 is provided with a tag reader 34 ( Fig. 20 ). The RF-ID tag 15 sends position information indicating the absolute position of the corresponding magnetic marker 10.

[0133] The detection processing circuit 20 of the sensor array 2 can execute the marker detection processing illustrated in Example 2. As described above, according to the marker detection processing, the magnetic marker 10 can be detected and the position of the vehicle 5 in the vehicle width direction relative to the magnetic marker 10 can be measured.

[0134] When the magnetic marker 10 is detected, the control unit 32 receives position information from the RF-ID tag 15 attached to the detected magnetic marker 10, thereby obtaining the absolute position of the magnetic marker 10. Furthermore, the position obtained by shifting the absolute position of the magnetic marker 10 by the position of the magnetic marker 10 in the vehicle width direction measured by the marker detection processing is determined as the vehicle position (the center position of the sensor array 2).

[0135] On the other hand, when the magnetic marker 10 is not detected, the control unit 32 refers to the map including the road surface magnetic distribution M2, determines the area corresponding to the measured magnetic distribution, and estimates the position of the vehicle, as in the structure described in the first embodiment. At this time, based on the steering angle measured by the steering angle sensor 353 and the vehicle speed measured by the vehicle speed sensor 355, the motion Mo( Fig.21). If the movement Mo of the vehicle 5 is estimated, the arrival position of the vehicle 5 corresponding to the time after passing the magnetic marker 10 can be estimated. If the range RS on the map to which the arrival position belongs is selected and the comparison of the measured magnetic distribution is performed within the range RS, the area corresponding to the measured magnetic distribution A in the road surface magnetic distribution M2 or the transformed magnetic distribution M2D can be determined efficiently.

[0136] It should be noted that other structures and effects are the same as those of the first embodiment.

[0137] The specific examples of the present invention are described in detail above as in the embodiments, but these specific examples are only examples of the technology included in the technical solution. Of course, the technical solution should not be interpreted restrictively by the structure, numerical value, etc. of the specific example. The technical solution includes the technology formed by variously deforming, changing or appropriately combining the specific examples using the known technology, the knowledge of those skilled in the art, etc.

[0138] Description of Reference Numerals

[0139] 1S Position Estimation System

[0140] 1 Map

[0141] 10 Magnetic marker

[0142] 2 Sensor array

[0143] 32 Control Unit

[0144] 321 Sports Estimation Department

[0145] 323 Magnetic distribution generation unit

[0146] 325 Transformation Processing Unit

[0147] 327 Position Estimation Department

[0148] 351 GPS Unit

[0149] 353 Steering angle sensor

[0150] 355 Vehicle speed sensor

[0151] 357 Height Sensor

[0152] 40 map database (storage unit)

[0153] 5 Vehicles

[0154] A. Measuring magnetic distribution

[0155] Cn Magnetic Sensor

[0156] M1 structural map

[0157] M2 Road Magnetic Distribution

[0158] M2D transforms magnetic distribution.

Claims

1. A method for estimating a position of a vehicle provided with a magnetic sensor for measuring magnetism acting from a road surface forming a surface of a travel path, wherein: The position estimation method comprises: A magnetic measurement process for obtaining a magnetic measurement value obtained by the magnetic sensor; a magnetic distribution generation process for generating a measured magnetic distribution indicating a distribution of magnetic data obtained based on the magnetic measurement values ​​obtained by the magnetic measurement process; Position estimation processing, which determines the area corresponding to the measured magnetic distribution in the road surface magnetic distribution by referring to a map corresponding to the road surface magnetic distribution representing the distribution of magnetic data obtained based on the magnetic quantity of each point on the road surface, and estimates the vehicle position based on the position of the area corresponding to the measured magnetic distribution on the map; and a conversion process, which converts the road surface magnetic distribution into a distribution of the installation height of the magnetic sensor, The magnetic data obtained based on the magnetic measurement values ​​is a magnetic gradient obtained from the difference between the magnetic measurement values ​​at two positions on the road surface, and the measured magnetic distribution is the distribution of the magnetic gradient. The magnetic data obtained based on the magnetic quantity of each point on the road surface is a magnetic gradient obtained based on the difference in magnetic quantity between two positions on the road surface, and the map is a map that establishes a corresponding relationship with the road surface magnetic distribution representing the distribution of the magnetic gradient. In the position estimation process, the transformed magnetic distribution obtained by the transformation process is compared with the measured magnetic distribution to determine the area in the transformed magnetic distribution that has a high degree of consistency with the measured magnetic distribution, thereby determining the area in the road surface magnetic distribution that corresponds to the measured magnetic distribution.

2. The position estimation method according to claim 1, wherein: The conversion process is a process of converting the distribution according to the ratio of the magnetometer measurement values ​​obtained by two magnetic sensors installed at different heights.

3. A method for estimating a position of a vehicle equipped with a magnetic sensor for measuring magnetism acting from a road surface forming a surface of a travel path, wherein: The position estimation method comprises: A magnetic measurement process for obtaining a magnetic measurement value obtained by the magnetic sensor; a magnetic distribution generation process for generating a measured magnetic distribution indicating a distribution of magnetic data obtained based on the magnetic measurement values ​​obtained by the magnetic measurement process; Position estimation processing, which determines the area corresponding to the measured magnetic distribution in the road surface magnetic distribution by referring to a map corresponding to the road surface magnetic distribution representing the distribution of magnetic data obtained based on the magnetic quantity of each point on the road surface, and estimates the vehicle position based on the position of the area corresponding to the measured magnetic distribution on the map; and a conversion process that converts the measured magnetic distribution into a distribution of the height of the road surface, The magnetic data obtained based on the magnetic measurement values ​​is a magnetic gradient obtained from the difference between the magnetic measurement values ​​at two positions on the road surface, and the measured magnetic distribution is the distribution of the magnetic gradient. The magnetic data obtained based on the magnetic quantity of each point on the road surface is a magnetic gradient obtained based on the difference in magnetic quantity between two positions on the road surface, and the map is a map that establishes a corresponding relationship with the road surface magnetic distribution representing the distribution of the magnetic gradient. In the position estimation process, the transformed magnetic distribution obtained by the transformation process is compared with the road surface magnetic distribution to determine the area in the road surface magnetic distribution that has a high degree of consistency with the transformed magnetic distribution, thereby determining the area corresponding to the measured magnetic distribution.

4. The position estimation method according to claim 3, wherein: The conversion process is a process of converting the distribution according to the ratio of the magnetometer measurement values ​​obtained by two magnetic sensors installed at different heights.

5. The position estimation method according to any one of claims 1 to 4, wherein: The conversion process is a process of converting the distribution according to the magnitude of the magnetic measurement value obtained by the magnetic sensor for the magnetic marker as a magnetic generation source with known magnetic intensity laid on the road surface.

6. The position estimation method according to any one of claims 1 to 4, wherein: On the travel path, magnetic markers serving as magnetic generation sources are laid at intervals so as to enable absolute positions to be determined. In the position estimation process, when the magnetic marker is detected by the vehicle, the vehicle position is estimated based on the absolute position of the magnetic marker. On the other hand, when the magnetic marker is not detected, the vehicle position is estimated based on the position of the area corresponding to the measured magnetic distribution on the map.

7. The position estimation method according to claim 6, wherein: When the magnetic marker is not detected, the arrival position of the vehicle after any magnetic marker is detected is estimated and the range on the map to which the arrival position belongs is selected, and the area corresponding to the measured magnetic distribution is determined within the range on the map.

8. The position estimation method according to claim 1 or 3, wherein: In the road surface magnetic distribution and the measured magnetic distribution, the position of the magnetic marker laid on the road surface is determined as the marker configuration point. In the position estimation process based on the measured magnetic distribution including the marker configuration points, when determining the area in the road surface magnetic distribution corresponding to the measured magnetic distribution, the area corresponding to the measured magnetic distribution is determined from the area where the position of the marker configuration points in the road surface magnetic distribution is consistent with the marker configuration points included in the measured magnetic distribution.

9. The position estimation method according to claim 7, wherein: In the road surface magnetic distribution and the measured magnetic distribution, the position of the magnetic marker laid on the road surface is determined as the marker configuration point. In the position estimation process based on the measured magnetic distribution including the marker configuration points, when determining the area in the road surface magnetic distribution corresponding to the measured magnetic distribution, the area corresponding to the measured magnetic distribution is determined from the area where the position of the marker configuration points in the road surface magnetic distribution is consistent with the marker configuration points included in the measured magnetic distribution.

10. A position estimation system for estimating the position of a vehicle provided with a magnetic sensor for measuring magnetism acting from a road surface side forming a surface of a travel path, wherein: The position estimation system comprises: a storage unit storing a map in which a correspondence is established with a road surface magnetic distribution which is a distribution of magnetic data obtained based on the magnetic quantity of each point on the road surface; a magnetic distribution generating unit that acquires the magnetic measurement values ​​obtained by the magnetic sensor and generates a measured magnetic distribution that is a distribution of magnetic data obtained based on the magnetic measurement values; a position estimating unit that determines an area corresponding to the measured magnetic distribution in a road surface magnetic distribution that is associated with a map stored in the storage unit, and estimates the vehicle position based on a position on the map of the area corresponding to the measured magnetic distribution; and a conversion processing unit that converts the road surface magnetic distribution into a distribution of installation heights of the magnetic sensor, The magnetic data obtained based on the magnetic measurement values ​​is a magnetic gradient obtained from the difference between the magnetic measurement values ​​at two positions on the road surface, and the measured magnetic distribution generated by the magnetic distribution generating unit is the distribution of the magnetic gradient. The magnetic data obtained based on the magnetic quantity of each point on the road surface is a magnetic gradient obtained based on the difference in magnetic quantity between two positions on the road surface, and the map is a map that establishes a corresponding relationship with the road surface magnetic distribution representing the distribution of the magnetic gradient. The position estimation unit is configured to compare the transformed magnetic distribution obtained by the transformation processing unit with the measured magnetic distribution to determine an area in the transformed magnetic distribution that has a high degree of consistency with the measured magnetic distribution, thereby determining an area in the road surface magnetic distribution that corresponds to the measured magnetic distribution.

11. The position estimation system according to claim 10, wherein: The road surface is a paved surface composed of a paving material in which magnetic powder as a powder of a magnetic material is dispersed, and the magnetic powder dispersed in the pavement is magnetized so that the magnetic distribution of the road surface takes a predetermined pattern.

12. A position estimation system for estimating the position of a vehicle provided with a magnetic sensor for measuring magnetism acting from a road surface side forming a surface of a travel path, wherein: The position estimation system comprises: a storage unit storing a map in which a correspondence is established with a road surface magnetic distribution which is a distribution of magnetic data obtained based on the magnetic quantity of each point on the road surface; a magnetic distribution generating unit that acquires the magnetic measurement values ​​obtained by the magnetic sensor and generates a measured magnetic distribution that is a distribution of magnetic data obtained based on the magnetic measurement values; a position estimating unit that determines an area corresponding to the measured magnetic distribution in a road surface magnetic distribution that is associated with a map stored in the storage unit, and estimates the vehicle position based on a position on the map of the area corresponding to the measured magnetic distribution; and a conversion processing unit that converts the measured magnetic distribution into a distribution of the height of the road surface, The magnetic data obtained based on the magnetic measurement values ​​is a magnetic gradient obtained from the difference between the magnetic measurement values ​​at two positions on the road surface, and the measured magnetic distribution generated by the magnetic distribution generating unit is the distribution of the magnetic gradient. The magnetic data obtained based on the magnetic quantity of each point on the road surface is a magnetic gradient obtained based on the difference in magnetic quantity between two positions on the road surface, and the map is a map that establishes a corresponding relationship with the road surface magnetic distribution representing the distribution of the magnetic gradient. The position estimating unit is configured to compare the transformed magnetic distribution obtained by the transformation processing unit with the road surface magnetic distribution to determine an area in the road surface magnetic distribution that has a high degree of consistency with the transformed magnetic distribution, thereby determining an area corresponding to the measured magnetic distribution.

13. The position estimation system according to claim 12, wherein: The road surface is a paved surface composed of a paving material in which magnetic powder as a powder of a magnetic material is dispersed, and the magnetic powder dispersed in the pavement is magnetized so that the magnetic distribution of the road surface takes a predetermined pattern.

14. The position estimation system according to any one of claims 10 to 13, wherein: Magnetic markers are laid along the travel path.

15. The position estimation system according to claim 14, wherein: The position of the magnetic marker is determined in the road surface magnetic distribution and the measured magnetic distribution. The position estimating unit is configured to identify a region corresponding to the measured magnetic distribution from among regions where the positions of the magnetic markers in the road surface magnetic distribution coincide with each other.

Images