Reliability map creation method, self-position determination method, device, and program

Abstract

A reliability map creation device according to the present disclosure: identifies a tentative self-position, which is a position on a reliability map corresponding to the position of a mobile robot at a time point of interest as estimated on the basis of a travel history of the mobile robot; when reliabilities are registered in association with the tentative self-position, determines a definite self-position on the basis of a self-position calculated using the result of detection at the time point of interest by a selected sensor corresponding to a reliability that meets a prescribed criterion indicating a high reliability among the registered reliabilities; calculates a reliability on the basis of the tentative self-position, the definite self-position determined on the basis of the tentative self-position, and an environment map; and registers the calculated reliability and information identifying the selected sensor, in association with a position on the reliability map corresponding to the definite self-position.

Description

Reliability map creation method, self-location determination method, device, and program

[0001] The present disclosure relates to a reliability map creating method, a reliability map creating device, a reliability map creating program, a self-location determining method, a self-location determining device, and a self-location determining program.

[0002] A control device for an autonomously traveling automated guided vehicle has been proposed, which performs autonomous driving control based on a current self-position calculated by a self-position estimation process. The control device performs an image processing-based self-position estimation process that calculates a first self-position based on landmarks photographed by an imaging device mounted on the autonomously traveling automated guided vehicle, and an odometry-based self-position estimation process that calculates a second self-position based on driving information of the autonomously traveling automated guided vehicle, and estimates the current self-position using the first self-position and the second self-position. The control device also changes the contribution of the first self-position and the second self-position to the estimation of the current self-position based on the calculation accuracy of a calculation accuracy map of the image processing-based self-position estimation process within a predetermined area including the driving route. In addition, the autonomously traveling automatic guided vehicle is equipped with multiple imaging devices with different imaging directions, and the calculation accuracy map includes a calculation accuracy map for each imaging direction based on each landmark group photographed from the same position but with different imaging directions. Using the calculation accuracy map for each imaging direction, the calculation accuracy of the vehicle's own position calculated based on a first landmark group photographed from a first imaging direction is compared with the calculation accuracy of the vehicle's own position calculated based on a second landmark group photographed from a second imaging direction different from the first imaging direction, and the vehicle's own position calculated based on the landmark group photographed from the imaging direction with the higher calculation accuracy and the calculation accuracy map for each imaging direction are selected to estimate the vehicle's current position (Patent Document 1).

[0003] Patent No. 7162584

[0004] Patent Document 1 describes the use of a calculation accuracy map for the image processing method when estimating a current self-location using a first self-location determined by an image processing method and a second self-location determined by an odometry method, but does not disclose how to create the calculation accuracy map.

[0005] Furthermore, even if a plurality of sensors can be used to determine the self-position, the technology described in Patent Document 1 can only take into consideration the calculation accuracy related to the image processing method.

[0006] The present disclosure has been made in consideration of the above points, and aims to provide a method for creating a reliability map in which reliability, which is an index showing the certainty of a self-position determined when multiple sensors are available, is registered in association with each position within a specified area, and a method for determining a self-position using the reliability map.

[0007] In order to achieve the above object, the present disclosure provides a reliability map creation method for creating a reliability map used to determine the self-location of a mobile robot, the method comprising: a plurality of self-location detection means including an environment detection sensor that detects the status of the arrangement of objects around the mobile robot; an environment map that shows the arrangement of objects around a travel path of the mobile robot; and a calculation unit that calculates the self-location of the mobile robot from the detection results of the environment detection sensor and the environment map. The reliability map is a map in which reliability, which is an index indicating the accuracy of the self-location calculated by the self-location detection means, and information identifying the self-location detection means corresponding to the reliability, are registered in association with positions on the map. The reliability map creation method includes the steps of: determining a tentative self-location, which is a position on the reliability map that corresponds to the location of the mobile robot at a time of interest estimated based on the travel history of the mobile robot; and, if the reliability is registered in association with the tentative self-location, selecting a position on the reliability map that satisfies a predetermined criterion indicating a high reliability among the registered reliability values. the self-location detection means corresponding to the reliability set by the selected self-location detection means is selected as the self-location detection means, and a confirmed self-location is determined based on the self-location calculated using the result detected by the environment detection sensor of the selected self-location detection means at the time of interest; a reliability calculation step calculating the reliability based on the tentative self-location, the confirmed self-location determined based on the tentative self-location, and the environment map; and a registration step registering the reliability calculated in the reliability calculation step and information identifying the self-location detection means used to calculate the reliability, in association with a position on the reliability map corresponding to the confirmed self-location.

[0008] The present disclosure also provides a self-localization method based on a reliability map for a mobile robot equipped with multiple self-localization means, including an environment detection sensor that detects the status of the placement of objects around the mobile robot, an environment map that shows the placement of objects around a travel path of the mobile robot, and a calculation unit that calculates the self-localization of the mobile robot from the detection results of the environment detection sensors and the environment map, wherein the reliability map is registered in association with positions on the map, with reliability being an index indicating the accuracy of the self-localization calculated by the self-localization means and information identifying the self-localization means corresponding to the reliability. The self-location determination method includes a tentative self-location determination step of determining a tentative self-location, which is a position on the reliability map corresponding to the location of the mobile robot at a time of interest estimated based on the mobile robot's travel history; and a self-location determination step of, if the reliability is registered in association with the tentative self-location, selecting the self-location detection means corresponding to a reliability that satisfies a predetermined criterion indicating a high reliability among the registered reliability levels as a selected self-location detection means, and determining a definitive self-location based on the self-location calculated using the result detected by the environment detection sensor of the selected self-location detection means at the time of interest.

[0009] The present disclosure also provides a reliability map creation device that creates a reliability map used to determine the self-location of a mobile robot, the reliability map creation device comprising a plurality of self-location detection means including an environment detection sensor that detects the status of the arrangement of objects around the mobile robot, an environment map that shows the arrangement of objects around a travel path of the mobile robot, and a calculation unit that calculates the self-location of the mobile robot from the detection results of the environment detection sensor and the environment map, wherein the reliability map is a map in which reliability, which is an index indicating the accuracy of the self-location calculated by the self-location detection means, and information identifying the self-location detection means corresponding to the reliability, are registered in association with positions on the map, and the reliability map creation device calculates a reliability map used to determine the self-location of the mobile robot based on the estimated position of the mobile robot at a time of interest based on the travel history of the mobile robot. a self-location determination unit that, when the reliability is registered in association with the tentative self-location, selects the self-location detection means corresponding to a reliability that meets a predetermined criterion indicating high reliability among the registered reliabilities as a selected self-location detection means, and determines a confirmed self-location based on the self-location calculated using the result detected by the environment detection sensor of the selected self-location detection means at the time of interest; a reliability calculation unit that calculates the reliability based on the tentative self-location, the confirmed self-location determined based on the tentative self-location, and the environment map; and a registration unit that registers the reliability calculated in the reliability calculation unit and information identifying the self-location detection means used to calculate the reliability, in association with a position on the reliability map that corresponds to the confirmed self-location.

[0010] The present disclosure also provides a self-location determination device that determines its own location based on a reliability map of the mobile robot, the device including a plurality of self-location detection means, the plurality of self-location detection means including an environment detection sensor that detects the situation regarding the placement of objects around the mobile robot, an environment map that shows the placement of objects around a travel path of the mobile robot, and a calculation unit that calculates the self-location of the mobile robot from the detection results of the environment detection sensor and the environment map, and the reliability map is displayed on a map with reliability as an index indicating the accuracy of the self-location calculated by the self-location detection means and information identifying the self-location detection means corresponding to the reliability. The self-location determination device includes a tentative self-location determination unit that determines a tentative self-location, which is a position on the reliability map that corresponds to the location of the mobile robot at a time of interest estimated based on the mobile robot's driving history, and a self-location determination unit that, when the reliability is registered in association with the tentative self-location, selects the self-location detection means that has a reliability that meets a predetermined criterion indicating a high reliability as a selected self-location detection means, and determines a definite self-location based on the self-location calculated using the result detected by the environment detection sensor of the selected self-location detection means at the time of interest.

[0011] The present disclosure also provides a reliability map creation program for creating a reliability map used to determine the self-location of a mobile robot having multiple self-location detection means, including an environment detection sensor that detects the status of the placement of objects around the mobile robot, an environment map that shows the placement of objects around a travel path of the mobile robot, and a calculation unit that calculates the self-location of the mobile robot from the detection results of the environment detection sensor and the environment map. The reliability map is a map in which reliability, which is an index indicating the accuracy of the self-location calculated by the self-location detection means, and information identifying the self-location detection means corresponding to the reliability, are registered in association with positions on the map, and the program is configured to execute a computer-generated program to generate a reliability map corresponding to the location of the mobile robot at a time of interest, which is estimated based on the travel history of the mobile robot. The program is a program for causing the device to function as a tentative self-location identification unit that identifies a tentative self-location, which is a position on a map; a self-location determination unit that, when the reliability is registered in association with the tentative self-location, selects the self-location detection means corresponding to the registered reliability that meets a predetermined standard indicating high reliability as a selected self-location detection means, and determines a confirmed self-location based on the self-location calculated using the result detected by the environment detection sensor of the selected self-location detection means at the time of interest; a reliability calculation unit that calculates the reliability based on the tentative self-location, the confirmed self-location determined based on the tentative self-location, and the environmental map; and a registration unit that registers the reliability calculated in the reliability calculation unit and information identifying the self-location detection means used to calculate the reliability in association with a position on the reliability map that corresponds to the confirmed self-location.

[0012] The present disclosure also provides a self-location determination program for a mobile robot that determines its own location based on a reliability map of the mobile robot, the reliability map including a plurality of self-location detection means including an environment detection sensor that detects the status of the placement of objects around the mobile robot, an environment map showing the placement of objects around a travel path of the mobile robot, and a calculation unit that calculates the self-location of the mobile robot from the detection results of the environment detection sensor and the environment map, the reliability map being a map in which reliability, which is an index indicating the accuracy of the self-location calculated by the self-location detection means, and information identifying the self-location detection means corresponding to the reliability, are associated with positions on the map. The program causes a computer to function as a tentative self-location determination unit that determines a tentative self-location, which is a position on the reliability map that corresponds to the location of the mobile robot at a time of interest estimated based on the mobile robot's driving history, and a self-location determination unit that, when the reliability is registered in association with the tentative self-location, selects the self-location detection means that corresponds to a reliability that meets a predetermined criterion indicating a high reliability among the registered reliability levels as a selected self-location detection means, and determines a definite self-location based on the self-location calculated using the result detected by the environment detection sensor of the selected self-location detection means at the time of interest.

[0013] According to the present disclosure, it is possible to provide a method for creating a reliability map in which reliability, which is an index showing the certainty of a self-position determined when multiple sensors are available, is registered in association with each position within a specified area, and a method for determining a self-position using the reliability map.

[0014] FIG. 1 is a block diagram showing a schematic configuration of a mobile robot according to an embodiment of the present invention. FIG. 2 is a block diagram showing the hardware configuration of a self-location determination device. FIG. 3 is a block diagram showing an example of the functional configuration of a self-location determination device. FIG. 4 is a diagram for explaining an environment map and a reliability map. FIG. 5 is a diagram for explaining a reliability map. FIG. 6 is a diagram for explaining a reliability map for each environment detection sensor. FIG. 7 is a diagram for explaining a reliability map for each environment detection sensor. FIG. 8 is a flowchart showing the flow of a self-location determination process. FIG. 9 is a flowchart showing the flow of a tentative self-localization process. FIG. 10 is a flowchart showing the flow of a reliability acquisition process. FIG. 11 is a flowchart showing the flow of a self-location determination process without reliability. FIG. 12 is a flowchart showing the flow of a self-location determination process with reliability. FIG. 13 is a diagram for explaining another example of a reliability map. FIG. 14 is a flowchart showing the flow of a reliability acquisition process in Modification Example 1.

[0015] An example of an embodiment of the present disclosure will be described below with reference to the drawings. Note that the same or equivalent components and parts in each drawing are given the same reference numerals. Also, the dimensions and proportions of the drawings are exaggerated for the convenience of explanation and may differ from the actual proportions.

[0016] 1 is a block diagram showing the schematic configuration of a mobile robot 100 according to this embodiment. The mobile robot 100 includes environment detection sensors 50A and 50B and a self-localization device 10. The self-localization device 10 includes a reliability map creation device 10A. The mobile robot 100 also includes a drive mechanism and control unit for autonomous travel, but detailed description of these will be omitted in this embodiment.

[0017] The environment detection sensors 50A and 50B are sensors that detect the status of the arrangement of objects around the mobile robot 100. The number of environment detection sensors 50A and 50B is not limited to the example shown in FIG. 1 and may be three or more. In this embodiment, the environment detection sensor 50A is a LiDAR (Light Detection and Ranging) sensor, and the environment detection sensor 50B is an image sensor. The image sensor may be a monocular camera, a stereo camera, an RGBD camera that combines an RGB camera and a distance sensor, a multi-camera, or the like. Each camera may be a normal camera with a viewing angle of 60 degrees or less, or a wide-angle camera. Alternatively, the camera may be an event camera that outputs only the changed pixel values. Hereinafter, when the environment detection sensor 50A and the environment detection sensor 50B are described without distinction, they will be referred to as the "environment detection sensor 50."

[0018] Fig. 2 is a block diagram showing the hardware configuration of the self-positioning device 10. As shown in Fig. 2, the self-positioning device 10 has a CPU (Central Processing Unit) 12, a memory 14, a storage device 16, an input device 18, an output device 20, a storage medium reader 22, and a communication I / F (Interface) 24. Each component is connected to each other via a bus 26 so as to be able to communicate with each other.

[0019] The storage device 16 stores a self-positioning program for executing the self-positioning process. The CPU 12 is a central processing unit that executes various programs and controls each component. That is, the CPU 12 reads the program from the storage device 16 and executes the program using the memory 14 as a work area. The CPU 12 controls each component and performs various arithmetic processes in accordance with the program stored in the storage device 16.

[0020] The memory 14 is configured with a RAM (Random Access Memory) and serves as a working area to temporarily store programs and data. The storage device 16 is configured with a ROM (Read Only Memory), a HDD (Hard Disk Drive), an SSD (Solid State Drive), etc., and stores various programs including the operating system and various data.

[0021] The input device 18 is a device for performing various inputs, such as a keyboard or a mouse. The output device 20 is a device for outputting various information, such as a display or a printer. A touch panel display may be used as the output device 20 to function as the input device 18.

[0022] The storage medium reader 22 reads data stored in various storage media such as CD (Compact Disc)-ROM, DVD (Digital Versatile Disc)-ROM, Blu-ray Disc, USB (Universal Serial Bus) memory, etc., and writes data to the storage media. The communication I / F 24 is an interface for communicating with other devices, and uses standards such as Ethernet (registered trademark), FDDI, and Wi-Fi (registered trademark).

[0023] Next, the functional configuration of the self-location determining device 10 will be described. FIG. 3 is a block diagram showing an example of the functional configuration of the self-location determining device 10. As shown in FIG. 3, the self-location determining device 10 includes, as its functional configuration, a tentative self-location identifying unit 32, a self-location determining unit 34, a reliability calculating unit 36, and a registering unit 38. Note that the self-location determining unit 34, the reliability calculating unit 36, and the registering unit 38 also function as functional components of the reliability map creating device 10A. Each functional configuration is realized when the CPU 12 reads out a self-location determining program stored in the storage device 16, expands it into the memory 14, and executes it.

[0024] Additionally, a predetermined storage area of ​​the self-positioning device 10 stores an environment map 40 and a reliability map 42. The environment map 40 shows the location of objects around the travel path of the mobile robot 100. The reliability map 42 is a map in which reliability, which is an index indicating the accuracy of the self-position calculated by the self-positioning device 10 (described later), and information identifying the environment detection sensor 50 corresponding to that reliability (≦, referred to as a "sensor ID") are registered in association with a position on the map. The environment detection sensor 50, the self-positioning unit 34, and the environment map 40 are examples of the self-position detection means of the present disclosure.

[0025] 4 shows a schematic diagram of the environment map 40 and the reliability map 42. The reliability map 42 is a grid-like map that divides the area shown by the environment map 40 into multiple location sections. The size of one grid of the reliability map 42 can be set arbitrarily depending on the type of mobile robot 100, the environment in which the mobile robot 100 travels, and other factors. For example, if the mobile robot 100 is an indoor mobile robot, one grid may be 5 cm wide, and if it is an automobile, one grid may be 1 m wide.

[0026] As shown in the upper diagram of Figure 4, the mobile robot 100 determines its own position while traveling through the environment shown in the environmental map 40. Therefore, as shown in the lower diagram of Figure 4, reliability is registered in the grids corresponding to the route traveled by the mobile robot 100 (the diagonally shaded grids in the lower diagram of Figure 4). The reliability may be a value between 0 and 1, for example, with values ​​closer to 1 indicating higher reliability. Note that a value (e.g., -1) is set in the white grids of the reliability map 42 in the lower diagram of Figure 4 to indicate that no reliability is registered.

[0027] FIG. 5 shows a schematic diagram of the reliability map 42 and an enlarged portion thereof. Each grid registers the detected orientation of the environment detection sensor 50 when the mobile robot 100 travels through the area corresponding to that grid, and the reliability at that time. The detected orientation corresponds to the posture of the mobile robot 100. For example, if the environment detection sensor 50 is facing the front (traveling direction) of the mobile robot 100, the detected orientation is the azimuth angle of the traveling direction. The detected orientation is quantized into eight directions, for example, up, down, left, right, and each diagonal direction. Note that the quantization of the detected orientation is not limited to eight directions, and may be quantized into four directions, for example, up, down, left, right, and each diagonal direction.

[0028] The reliability map 42 may include multiple maps for each environment detection sensor 50, or a single map in which the reliability corresponding to each environment detection sensor 50 and the sensor ID of that environment detection sensor 50 are registered at each position on the map. That is, registering the reliability and sensor ID in association with the position on the reliability map 42 is not limited to directly registering the reliability and sensor ID with the position on the reliability map 42, but also includes indirectly registering the reliability and sensor ID with the position. For example, in the case of a reliability map 42 for each environment detection sensor 50, the reliability value and information identifying the reliability map 42 (hereinafter referred to as "reliability map ID") may be registered in association with the position on the reliability map 42, and the association between the reliability map ID and the sensor ID may be registered separately from the reliability map 42. Alternatively, the reliability value and information identifying the reliability (hereinafter referred to as "reliability ID") may be directly registered in association with the position on the reliability map 42, and the association between the reliability ID and the sensor ID may be registered separately from the reliability map 42. In addition, information specifying a position (such as a serial number of the position data, hereinafter referred to as "position data ID") may be registered in association with a position on the reliability map 42, and the position data ID and a reliability value may be registered in association with each other separately from the reliability map 42, and the position data ID and a sensor ID may also be registered in association with each other separately.

[0029] The tentative self-localization unit 32 periodically acquires sensor values ​​from each environment detection sensor 50. In this embodiment, it is assumed that the sensor values ​​from each environment detection sensor 50 are acquired synchronously, but synchronization is not essential. The tentative self-localization unit 32 determines a tentative self-localization, which is a position on the reliability map 42 that corresponds to the location of the mobile robot 100 at a time of interest, estimated based on the mobile robot's 100 travel history.

[0030] For example, the tentative self-localization unit 32 may acquire an odometry signal, which is the rotational angle of the left and right wheels of the mobile robot 100. The tentative self-localization unit 32 may then convert the odometry signal into the translational velocity and angular velocity of the mobile robot 100, and determine the tentative self-localization by integrating the self-localization signal with the self-localization signal from the previous period. Alternatively, the tentative self-localization unit 32 may determine the tentative self-localization by estimating the self-localization at a given time based on the movement of the mobile robot 100 estimated from the history of past self-localizations. Alternatively, the tentative self-localization unit 32 may determine the tentative self-localization by using a method such as SLAM (Simultaneous Localization and Mapping) that uses an environment map 40 and sensor values. Furthermore, for example, the tentative self-location identifying unit 32 may identify, as the tentative self-location, location information measured by a GPS (Global Positioning System), an IMU (Inertial Measurement Unit), etc. Note that the method for identifying the tentative self-location may be a method that is simpler than the method for calculating the self-location described later.

[0031] When a reliability level is registered in association with a tentative self-location, and a detected direction of the environment detection sensor 50 corresponding to a reliability level that satisfies a predetermined criterion indicating high reliability among the registered reliability levels is a detection direction that the environment detection sensor 50 can detect at the time of interest, the self-location determination unit 34 determines the reliability level as a valid reliability level. The self-location determination unit 34 selects the environment detection sensor 50 for which a valid reliability level is registered for the tentative self-location in the reliability map 42 as a selected sensor, and calculates the self-location using the sensor value detected by the selected sensor at the time of interest for the detected direction registered in association with the reliability level. The self-location determination unit 34 calculates the self-location using, for example, a method such as SLAM using the environment map 40 and the sensor value. The self-location determination unit 34 then determines the average of the self-locations calculated for each selected sensor as the confirmed self-location.

[0032] The reliability that satisfies a predetermined standard may be, for example, a reliability equal to or greater than a threshold. The threshold may be, for example, 1 / (number of environment detection sensors 50) × a. Note that a may be any value (e.g., 0.8) such that a≦1. In this embodiment, since the number of environment detection sensors 50 is two, the threshold is 0.4. For example, if the reliability of both the environment detection sensors 50A and 50B is 1 / 2, both are equal to or greater than the threshold. Therefore, the self-location determination unit 34 determines the average of the self-locations calculated based on the sensor values ​​of the environment detection sensors 50A and 50B as the confirmed self-location. Furthermore, for example, if the reliability of the environment detection sensor 50A is 99 / 100 and the reliability of the environment detection sensor 50B is 1 / 100, only the sensor value of the environment detection sensor 50A is used.

[0033] In addition, if the reliability is not registered in the grid of the reliability map 42 corresponding to the tentative self-position, the self-position determination unit 34 determines the definitive self-position by averaging the self-positions calculated by the multiple environment detection sensors 50 that deviate from the tentative self-position within a predetermined range.

[0034] Furthermore, if a reliability associated with the tentative self-position is registered but the reliability that satisfies the predetermined criteria is not registered, or if the detected direction registered together with the reliability that satisfies the predetermined criteria is not a detected direction that the environment detection sensor 50 can detect at the time of interest, the self-position determination unit 34 determines the tentative self-position of the mobile robot 100 at the time of interest, estimated based on the driving history of the mobile robot 100, as the confirmed self-position.

[0035] The reliability calculation unit 36 ​​calculates the reliability based on the tentative self-location, the confirmed self-location determined based on the tentative self-location, and the environment map 40. Specifically, when the environment map 40 and the confirmed self-location are given, the reliability calculation unit 36 ​​calculates the probability that the tentative self-location will be identified for each environment detection sensor 50, and calculates the reliability by normalizing the calculated probability.

[0036] For example, in the case of the environment detection sensor 50A (LiDAR), n points are observed, and the angle of point i with respect to the detection direction of the LiDAR is defined as θ ｉ , the distance between the LiDAR and point i is d ｉ Then, the reliability = (point 1 is at distance d １ (Probability of being observed at point 2) * (Probability of being observed at point 2) ２ Probability of being observed at point n) * ... * (point n is at distance d ｎ More specifically, the reliability (LiDAR) may be calculated using the following formula (1):

[0037]

[0038] Note that A is a coefficient for normalization, which is obtained by multiplying the Gaussian distribution coefficient 1 / √(2πσ) n times. ｍ is an observation value estimated from the environment map 40, and the self-position T ｉ Angle θ with respect to the detection direction ｉ is the distance to the first obstacle hit when light is projected in the direction of σ. σ is the standard deviation of the measurement value, and represents the magnitude of the error in the distance measurement.

[0039] In the case of the environment detection sensor 50B (image sensor), n points are observed, and the angle of point i with respect to the detection direction of the image sensor is expressed as θｉ , the pixel position of point i is p ｉ Then, the reliability = (point 1 is the pixel position p １ (Probability of being observed at pixel position p) * (Probability of being observed at pixel position p) ２ (Probability of being observed at point n) * ... * (Probability of being observed at point n) ｎ More specifically, the reliability (image sensor) may be calculated using the following formula (2):

[0040]

[0041] In addition, p ｍ is an observation value estimated from the environment map 40, and the self-position T ｉ In the image captured of the environment shown in the environment map 40, the pixel position p of the observed image is ｉ The corresponding pixel positions are searched for by feature point matching using ORB (Oriented FAST and Rotated BRIEF) feature amounts.

[0042] For example, when using reliability that depends on the environment detection sensor 50, such as the difference between an observed image and an image corresponding to the environment map 40, it is not possible to appropriately compare reliability between the environment detection sensors 50. In this embodiment, by calculating reliability using probability as described above, the reliability can be used as an index that is common to multiple environment detection sensors 50. This makes it possible to select an appropriate environment detection sensor 50 when selecting an environment detection sensor 50 to be used at each location.

[0043] The registration unit 38 registers information specifying the detection direction corresponding to the reliability of the selected sensor, along with the calculated reliability and the sensor ID of the selected sensor, in association with a grid of the reliability map 42 corresponding to the confirmed self-location. If a reliability has already been registered in the corresponding grid, the registration unit 38 updates the registered reliability with the newly calculated reliability.

[0044] This creates a reliability map 42 in which reliability and sensor IDs are registered in association with map positions. For example, the lower diagram of FIG. 6 shows an example of a reliability map of the mobile robot's position determined while the mobile robot 100 moves along a path in an environment such as the upper diagram of FIG. 6 , where the LiDAR observation range is 240 degrees forward and the image sensor camera is facing directly in the direction of movement. Circles indicate the reliability calculated for that position, with open circles indicating high reliability and closed circles indicating low reliability. The arrows next to the circles indicate the detection direction. For example, in locations with no topographical features, such as where only the base of a shelf can be observed, the accuracy of LiDAR's calculation of the position decreases, resulting in low reliability. On the other hand, for image sensors, in locations with few distinctive landmarks, such as where only walls are visible, the accuracy of LiDAR's calculation of the position decreases, resulting in low reliability.

[0045] Furthermore, for example, in the case where the LiDAR observation range is 360 degrees and the image sensor camera is an omnidirectional camera, an example of a reliability map of the self-position determined while moving along a route in an environment such as that shown in the upper diagram of Fig. 7 is shown in the lower diagram of Fig. 7. Note that in Fig. 7, objects indicated by diagonal lines, such as a chair, a toy box, and objects around the toy box (assuming toys have been taken out of the toy box), are objects with uncertain positions, i.e., objects that are prone to movement. If the presence of such objects causes the object's position at the time of self-position determination to change from the position of the object shown in the environment map 40, the calculation accuracy of the self-position in the vicinity will decrease, and the reliability will also decrease, in both the LiDAR and the image sensor cases.

[0046] 7, since the detection direction of both the LiDAR and the image sensor is 360 degrees, no arrows are added to the circles. In such cases, the reliability map 42 is registered with the reliability either corresponding to all quantized directions or without corresponding to a detection direction. When obtaining the reliability from such a reliability map 42, the registered reliability can be treated as valid regardless of the direction of the detection direction at the time of interest.

[0047] In addition, the lower diagram of Figure 8 shows an example of a reliability map of the self-position determined while moving along a route in an environment such as that shown in the upper diagram of Figure 8, where the LiDAR observation range is 240 degrees forward and the image sensor camera is facing directly in the direction of movement of the mobile robot 100. In this case, as with the example of Figure 6, the LiDAR reduces the accuracy of the self-position calculation in locations without topographical features, such as those surrounded by shelves and walls, resulting in a low reliability. On the other hand, the image sensor reduces the accuracy of the self-position calculation in locations with few distinctive landmarks, such as those where only walls are visible, resulting in a low reliability.

[0048] Next, the operation of the self-positioning device 10 according to this embodiment will be described.

[0049] 9 is a flowchart showing the flow of the self-positioning process executed by the CPU 12 of the self-positioning device 10. The CPU 12 reads out the self-positioning program from the storage device 16, loads it into the memory 14, and executes it, causing the CPU 12 to function as each functional component of the self-positioning device 10, and the self-positioning process shown in FIG.

[0050] In step S10 , the tentative self-localization unit 32 acquires sensor values ​​from each environment detection sensor 50 .

[0051] Next, in step S20, a tentative self-localization process is executed. The tentative self-localization process will now be described with reference to FIG.

[0052] In step S22, the tentative self-localization unit 32 identifies the trajectory, which is the traveling history of the mobile robot 100, based on the sensor values ​​of the environment detection sensor 50, and calculates the self-localization so that the trajectory and the environment map 40 fit best.

[0053] Next, in step S24, the tentative self-localization unit 32 determines whether the error in the fitting in step S22 is equal to or greater than a predetermined threshold. If the error is equal to or greater than the threshold, the process proceeds to step S26, and if the error is less than the threshold, the process proceeds to step S32.

[0054] In step S26, the tentative self-localization unit 32 acquires the odometry signal. Next, in step S28, the tentative self-localization unit 32 converts the odometry signal into the translational velocity and angular velocity of the mobile robot 100. Next, in step S30, the tentative self-localization unit 32 multiplies the self-location determined in the previous cycle by the result of the conversion in step S28 to calculate the self-location in the current cycle, and identifies the calculated self-location as the tentative self-location.

[0055] On the other hand, in step S32, the tentative self-location determination unit 32 determines the self-location calculated by fitting with the environment map 40 in step S22 as the tentative self-location, the tentative self-location determination process ends, and the process returns to the self-location determination process (Figure 9).

[0056] Next, in step S40, a reliability acquisition process is executed. The reliability acquisition process will now be described with reference to FIG.

[0057] In step S42, the self-location determination unit 34 quantizes the tentative self-location identified in step S20 using the grid width of the reliability map 42. The self-location determination unit 34 also quantizes the detected orientation at the tentative self-location using the granularity of the detected orientation registered in the reliability map 42 (eight directions in this example).

[0058] Next, in step S44, the self-location determination unit 34 determines whether or not a reliability corresponding to the quantized detected direction at the tentative self-location is registered in a grid of the reliability map 42 corresponding to the quantized tentative self-location. If a corresponding reliability is registered, the process proceeds to step S46, and if not, the process proceeds to step S48.

[0059] In step S46, the self-position determining unit 34 acquires the reliability determined to be registered in step S44, and returns to the self-position determining process (FIG. 9).

[0060] Meanwhile, in step S48, the self-location determination unit 34 extracts, from nearby grids present within a predetermined range from the grid in the reliability map 42 corresponding to the quantized tentative self-location, a grid in which a reliability corresponding to the detection orientation in a predetermined range including the quantized detection orientation at the tentative self-location is registered. The predetermined range for the grid may be, for example, eight grids surrounding the grid corresponding to the quantized tentative self-location. Furthermore, the predetermined range for the detection orientation may be, for example, a direction of 45 degrees forward and backward from the quantized detection orientation.

[0061] Next, in step S50, the self-position determining unit 34 determines whether or not the grid extracted in step S48 exists. If the extracted grid exists, the process proceeds to step S52, and if not, the process proceeds to step S54.

[0062] In step S52, the self-location determining unit 34 obtains, from the grid closest to the grid corresponding to the quantized tentative self-location, the reliability associated with the detected orientation within a predetermined range including the quantized detected orientation at the tentative self-location. Meanwhile, in step S54, the self-location determining unit 34 returns a message indicating that there is no reliability associated with the tentative self-location, and returns to the self-location determining process ( FIG. 9 ).

[0063] Next, in step S60, the self-location determination unit 34 determines whether or not a reliability has been registered in the grid of the reliability map 42 corresponding to the tentative self-location. If the reliability acquisition process in step S40 returns a message indicating that there is no reliability, it determines that the reliability has not been registered, and the process proceeds to step S70. If the reliability has been acquired in the reliability acquisition process in step S40, it determines that the reliability has been registered, and the process proceeds to step S80.

[0064] In step S70, the unreliable self-location determination process is executed, which will now be described with reference to FIG.

[0065] In step S72, the self-location determination unit 34 calculates a self-location for each environment detection sensor 50, for example, by a technique such as SLAM, based on the sensor values ​​of each environment detection sensor 50 and the environment map 40. Next, in step S74, the self-location determination unit 34 compares the tentative self-location identified in step S20 with the self-location calculated in step S72, and excludes any self-location that differs from the tentative self-location by more than a predetermined threshold as an outlier. Next, in step S76, the self-location determination unit 34 averages the self-locations calculated in step S72 that were not excluded as outliers in step S74, and determines this as a confirmed self-location, and returns to the self-location determination process ( FIG. 9 ).

[0066] On the other hand, in step S80, a reliable self-location determination process is executed. Here, the reliable self-location determination process will be described with reference to FIG.

[0067] In step S82, the self-position determination unit 34 determines whether or not there is an environment detection sensor 50 whose reliability is equal to or greater than a predetermined threshold among the environment detection sensors 50 whose reliability is acquired in the reliability acquisition process in step S40. If there is an environment detection sensor 50 whose reliability is equal to or greater than the threshold, that environment detection sensor 50 is selected, and the process proceeds to step S84; if there is no environment detection sensor 50, the process proceeds to step S88.

[0068] In step S84, the self-location determination unit 34 calculates the self-location for each selected sensor based on the sensor value and the environment map 40, for example, by a technique such as SLAM. Next, in step S86, the self-location determination unit 34 averages the self-locations calculated in step S84 above to determine the averaged self-location as a confirmed self-location, and returns to the self-location determination process ( FIG. 9 ). On the other hand, in step S88, the tentative self-location identified in step S20 above is determined as a confirmed self-location, and returns to the self-location determination process ( FIG. 9 ).

[0069] Next, in step S90, the reliability calculation unit 36 ​​calculates the probability that a tentative self-position based on the sensor values ​​of each environment detection sensor 50 will be identified when the environment map 40 and the confirmed self-position are given, and calculates the reliability by normalizing the calculated probability.

[0070] Next, in step S92, the registration unit 38 determines whether or not a reliability has already been registered for the grid corresponding to the tentative self-location. If the reliability has not been registered, the process proceeds to step S94, and if the reliability has already been registered, the process proceeds to step S96.

[0071] In step S94, the registration unit 38 registers the reliability corresponding to the selected sensor, together with information specifying the detection direction of the selected sensor used to calculate the reliability, in the corresponding grid of the reliability map 42. Meanwhile, in step S96, the registration unit 38 updates the reliability already registered in the corresponding grid with the newly calculated reliability, and the self-location determination process then ends.

[0072] The self-position determination process is repeatedly executed each time a sensor value is input from the environment detection sensor 50 at a fixed period.

[0073] As described above, the self-localization device according to this embodiment determines a tentative self-localization, which is a position on a reliability map corresponding to the mobile robot's location at a time of interest, estimated based on the mobile robot's travel history. Furthermore, when reliability values ​​associated with the tentative self-localization are registered, the self-localization device selects an environmental detection sensor corresponding to a reliability value that meets a predetermined criterion indicating high reliability among the registered reliability values ​​as a selected sensor. The self-localization device then determines a final self-localization based on a self-localization calculated using the detection results of the selected sensor at the time of interest. The self-localization device also calculates the reliability based on the tentative self-localization, the final self-localization determined based on the tentative self-localization, and the environment map. The self-localization device then registers the calculated reliability and the sensor ID of the selected sensor in association with a position on the reliability map corresponding to the final self-localization. This makes it possible to create a reliability map in which reliability values, which are indicators of the accuracy of a self-localization determined when multiple sensors are available, are registered in association with each position within a specified area, and to determine a self-localization with high reliability using the reliability map.

[0074] Specifically, as described with reference to Figures 6 to 8, depending on the type of environmental detection sensor, there are positions where the estimation accuracy of the self-location is high or low. In this embodiment, as described above, a reliability map can be created in which reliability and sensor IDs are registered in association with map positions, making it possible to identify positions where the accuracy of the self-location estimation may be low depending on the environmental detection sensor. Furthermore, since the cause of the low accuracy can be identified based on the type of environmental detection sensor and the environment map, measures can be taken to eliminate the cause. One possible measure is to place objects that can serve as landmarks, such as posting posters in locations where the accuracy of the image sensor is likely to be low. Furthermore, by creating such a reliability map, the environmental detection sensor to be used when determining the self-location can be appropriately selected, allowing a highly reliable self-location to be calculated.

[0075] <Modification 1> In the above embodiment, a grid-shaped reliability map has been described, but this is not limiting. For example, coordinates may be set on the reliability map, and positions on the reliability map may be represented using those coordinates. For example, as shown in Fig. 14, the reliability of each environment detection sensor may be stored in association with each coordinate. In the example of Fig. 14, x and y are coordinates on the reliability map, and θ is the detection orientation.

[0076] With reference to FIG. 15, the reliability acquisition process executed in step S40 of the self-position determination process (FIG. 9) in the first modification will be described.

[0077] In step S242, the self-position determination unit searches, for example, a list-format reliability map such as that shown in FIG. 14, for data in which the difference between the coordinates (x, y) in the reliability map and the identified tentative self-position is within a predetermined range, and the difference between the detected orientation θ in the reliability map and the detected orientation of the environment detection sensor at the time of interest is within a predetermined range.

[0078] Next, in step S244, the self-location determining unit determines whether or not one or more pieces of data were found in step S242. If one or more pieces of data were found, the process proceeds to step S246, and if no data was found, the process proceeds to step S248.

[0079] In step S246, the self-location determining unit identifies data from the retrieved data that includes (x, y, θ) closest to the tentative self-location and the detected direction at the time of interest, and obtains the reliability of the corresponding environment detection sensor in that data. On the other hand, in step S248, the self-location determining unit returns a message indicating that there is no corresponding reliability, and the reliability obtaining process ends.

[0080] <Variation 2> In the above embodiment, a self-location determination device that determines its own location while creating (updating) a reliability map has been described. However, the reliability map may be created and updated offline, and the reliability map created offline may be fixed and used during the self-location determination process. In this case, a reliability map creation device that creates a reliability map offline may execute a reliability map creation process including steps S10, S70, and S90 to S96 of Figure 9. Furthermore, a self-location determination device that uses a fixed reliability map may execute a self-location determination process including steps S10 to S80 of Figure 9.

[0081] Regardless of whether the reliability creation map is created offline or online, the process of updating the reliability already registered is not essential.

[0082] Furthermore, the reliability map created by the reliability creation device is not limited to use in the self-location determination device, and the created reliability map may be visualized and presented to the user, thereby urging the user to take action in areas where the accuracy of the environment detection sensor is likely to be poor, as described above.

[0083] In addition, the self-positioning process executed by the CPU in the above embodiment by reading the software (program) may be executed by various processors other than the CPU. Examples of processors in this case include PLDs (Programmable Logic Devices) whose circuit configuration can be changed after manufacture, such as FPGAs (Field-Programmable Gate Arrays), and dedicated electrical circuits, which are processors having a circuit configuration designed specifically for executing specific processes, such as ASICs (Application Specific Integrated Circuits). Furthermore, the self-positioning process may be executed by one of these various processors, or by a combination of two or more processors of the same or different types (for example, multiple FPGAs, a combination of a CPU and an FPGA, etc.). Furthermore, the hardware structure of these various processors is, more specifically, an electrical circuit that combines circuit elements such as semiconductor elements.

[0084] In the above embodiment, the self-positioning program is pre-stored (installed) in a storage device, but this is not limiting. The program may be provided in a form stored in a storage medium such as a CD-ROM, DVD-ROM, Blu-ray disc, or USB memory. The program may also be downloaded from an external device via a network.

[0085] The following are additional notes regarding this disclosure.

[0086] (Supplementary Item 1) A method for creating a reliability map used to determine the self-location of a mobile robot equipped with multiple self-location detection means including an environment detection sensor that detects the situation regarding the placement of objects around the mobile robot, an environment map that shows the placement of objects around a travel path of the mobile robot, and a calculation unit that calculates the self-location of the mobile robot from the detection results of the environment detection sensor and the environment map, wherein the reliability map is a map in which reliability, which is an index indicating the accuracy of the self-location calculated by the self-location detection means, and information identifying the self-location detection means corresponding to the reliability, are registered in association with positions on the map, and the method for creating the reliability map includes: a tentative self-location determination step of determining a tentative self-location, which is a position on the reliability map that corresponds to the location of the mobile robot at a time of interest estimated based on the mobile robot's travel history; a self-location determination step of selecting a self-location detection means, when the reliability is registered in association with the tentative self-location, that corresponds to a reliability that satisfies a predetermined standard indicating high reliability among the registered reliabilities, and determining a confirmed self-location based on the self-location calculated using a result detected at the time of interest by the environment detection sensor of the selected self-location detection means; a reliability calculation step of calculating the reliability based on the tentative self-location, the confirmed self-location determined based on the tentative self-location, and the environment map; and a registration step of registering the reliability calculated in the reliability calculation step and information identifying the self-location detection means used to calculate the reliability, in association with a position in the reliability map corresponding to the confirmed self-location.

[0087] (Supplementary Item 2) The reliability map is a map in which information specifying a detection direction corresponding to the reliability of the environment detection sensor included in the self-location detection means is registered in association with a position on the map, along with information specifying the reliability and the self-location detection means corresponding to the reliability; the self-location determination step, when the reliability is registered in association with the tentative self-location and, if the environment detection sensor of the self-location detection means corresponding to the registered reliability that satisfies the predetermined criterion is detectable by the environment detection sensor at the time of interest, selects the self-location detection means as a selected self-location detection means and determines the final self-location based on the self-location calculated using a result of detection by the environment detection sensor of the selected self-location detection means for the detection direction at the time of interest; and the registration step registers information specifying the detection direction corresponding to the reliability of the environment detection sensor included in the self-location detection means, along with the reliability calculated in the reliability calculation step and information specifying the self-location detection means corresponding to the reliability.

[0088] (Supplementary Item 3) The reliability map creation method according to Supplementary Item 1 or Supplementary Item 2, wherein the reliability map is divided into a plurality of position sections, and a position on the reliability map is represented by information that identifies the position section.

[0089] (Supplementary Item 4) The reliability map creating method according to Supplementary Item 1 or Supplementary Item 2, wherein coordinates are set in the reliability map, and positions on the reliability map are represented using the coordinates.

[0090] (Appendix 5) A reliability map creation method described in any one of appendixes 1 to 4, in which, if the reliability is not registered for the tentative self-position, the self-position determination step determines the final self-position by averaging the self-positions calculated by the multiple self-position detection means that deviate from the tentative self-position within a predetermined range.

[0091] (Supplementary Item 6) The method for creating a reliability map according to Supplementary Item 2, wherein the self-location determination step determines the location of the mobile robot at the time of interest, estimated based on the mobile robot's travel history, as the final self-location, when the reliability has been registered in association with the tentative self-location but no reliability satisfying the predetermined criterion has been registered, or when the detected orientation registered together with the reliability satisfying the predetermined criterion is not a detected orientation that can be detected by the environment detection sensor of the self-location detection means at the time of interest.

[0092] (Supplementary Item 7) A reliability map creation method described in any one of Supplementary Items 1 to 6, in which the reliability calculation step calculates the probability that the tentative self-location will be identified for each self-location detection means when the environmental map and the determined self-location are given, and calculates the reliability by normalizing the calculated probability.

[0093] (Supplementary Item 8) A reliability map creation method according to any one of Supplementary Items 1 to 7, wherein the reliability map is a plurality of reliability maps for each of the self-location detection means, or a single reliability map in which, at each position on the map, information identifying the self-location detection means is registered together with the reliability corresponding to the self-location detection means.

[0094] REFERENCE SIGNS LIST 10 Self-positioning device 10A Reliability map creating device 12 CPU 14 Memory 16 Storage device 18 Input device 20 Output device 22 Storage medium reading device 24 Communication I / F 26 Bus 32 Provisional self-position identifying unit 34 Self-positioning unit 36 ​​Reliability calculation unit 38 Registration unit 40 Environment map 42 Reliability map 50, 50A, 50B Environment detection sensor 100 Mobile robot

Claims

1. A method for creating a reliability map used to determine the self-location of a mobile robot equipped with multiple self-location detection means including an environment detection sensor that detects the situation regarding the placement of objects around the mobile robot, an environment map that shows the placement of objects around the mobile robot's travel path, and a calculation unit that calculates the mobile robot's self-location from the detection results of the environment detection sensor and the environment map, wherein the reliability map is a map in which reliability, which is an index indicating the accuracy of the self-location calculated by the self-location detection means, and information identifying the self-location detection means corresponding to the reliability, are registered in association with positions on the map, and the method for creating the reliability map comprises: a tentative self-location determination step of determining a tentative self-location, which is a position on the reliability map that corresponds to the mobile robot's position at a time of interest estimated based on the mobile robot's travel history; a self-location determination step of selecting a self-location detection means, when the reliability is registered in association with the tentative self-location, that corresponds to a reliability that satisfies a predetermined standard indicating high reliability among the registered reliabilities, and determining a confirmed self-location based on the self-location calculated using a result detected at the time of interest by the environment detection sensor of the selected self-location detection means; a reliability calculation step of calculating the reliability based on the tentative self-location, the confirmed self-location determined based on the tentative self-location, and the environment map; and a registration step of registering the reliability calculated in the reliability calculation step and information identifying the self-location detection means used to calculate the reliability, in association with a position in the reliability map corresponding to the confirmed self-location.

2. The reliability map creation method according to claim 1, wherein the reliability map is a map in which information specifying the reliability and the self-location detection means corresponding to the reliability, as well as information specifying the detection direction corresponding to the reliability of the environment detection sensor included in the self-location detection means, is registered in association with a position on the map; the self-location determination step, when the reliability is registered in association with the tentative self-location and if the environment detection sensor of the self-location detection means corresponding to the registered reliability that satisfies the predetermined criterion is detectable by the environment detection sensor at the time of interest, selects the self-location detection means as a selected self-location detection means and determines the definite self-location based on the self-location calculated using the detection result of the environment detection sensor of the selected self-location detection means for the detection direction at the time of interest; and the registration step registers information specifying the detection direction corresponding to the reliability of the environment detection sensor included in the self-location detection means, as well as the reliability calculated in the reliability calculation step and information specifying the self-location detection means corresponding to the reliability.

3. A reliability map creation method according to claim 1 or claim 2, wherein the reliability map is divided into a plurality of position sections, and a position on the reliability map is represented by information specifying the position section.

4. A reliability map creation method according to claim 1 or 2, wherein coordinates are set in the reliability map, and positions on the reliability map are expressed using the coordinates.

5. A reliability map creation method as described in claim 1 or claim 2, wherein, if the reliability is not registered for the tentative self-position, the self-position determination step determines the definite self-position by averaging the self-positions calculated by the self-position detection means that deviate from the tentative self-position within a predetermined range.

6. A reliability map creation method as described in claim 2, wherein the self-location determination step determines the location of the mobile robot at the time of interest, estimated based on the mobile robot's travel history, as the final self-location, when the reliability is registered in association with the tentative self-location but a reliability that meets the specified criteria is not registered, or when the detected orientation registered together with the reliability that meets the specified criteria is not a detected orientation that can be detected by the environment detection sensor of the self-location detection means at the time of interest.

7. A reliability map creation method as described in claim 1 or claim 2, wherein the reliability calculation step calculates the probability that the tentative self-location will be identified for each self-location detection means when the environmental map and the determined self-location are given, and calculates the reliability by normalizing the calculated probability.

8. A reliability map creation method as described in claim 1 or claim 2, wherein the reliability map is a plurality of reliability maps for each of the self-location detection means, or a single reliability map in which information identifying the self-location detection means is registered along with the reliability corresponding to the self-location detection means at each position on the map.

9. A self-localization method based on a reliability map for a mobile robot equipped with multiple self-localization means including an environment detection sensor that detects the status of the placement of objects around the mobile robot, an environment map that shows the placement of objects around the mobile robot's travel path, and a calculation unit that calculates the mobile robot's self-localization from the detection results of the environment detection sensor and the environment map, wherein the reliability map is a map in which reliability, which is an index indicating the accuracy of the self-localization calculated by the self-localization means, and information identifying the self-localization means corresponding to the reliability, are registered in association with positions on the map, the self-localization method comprising: a tentative self-localization step of determining a tentative self-localization, which is a position on the reliability map that corresponds to the mobile robot's location at a given time point, estimated based on the mobile robot's travel history; and a self-localization step of, if the reliability is registered in association with the tentative self-localization, selecting the self-localization means corresponding to a reliability that meets a predetermined criterion indicating a high reliability among the registered reliability levels as a selected self-localization means, and determining a definitive self-localization based on the self-localization calculated using the results detected by the environment detection sensor of the selected self-localization means at the given time point. A self-location determination method including:

10. A self-location determination method as described in claim 9, further comprising: a reliability calculation step of calculating the reliability based on the tentative self-location, the confirmed self-location determined based on the tentative self-location, and the environmental map; and a registration step of registering the reliability calculated in the reliability calculation step and information identifying the self-location detection means used to calculate the reliability, in association with a position on the reliability map corresponding to the confirmed self-location.

11. A reliability map creation device that creates a reliability map used to determine the self-location of a mobile robot, the reliability map being equipped with multiple self-location detection means including an environment detection sensor that detects the situation regarding the placement of objects around the mobile robot, an environment map that shows the placement of objects around the mobile robot's travel path, and a calculation unit that calculates the mobile robot's self-location from the detection results of the environment detection sensor and the environment map, wherein the reliability map is a map in which reliability, which is an index indicating the accuracy of the self-location calculated by the self-location detection means, and information identifying the self-location detection means corresponding to the reliability, are registered in association with positions on the map, and the reliability map creation device comprises: a tentative self-location determination unit that determines a tentative self-location, which is a position on the reliability map that corresponds to the mobile robot's position at a time of interest estimated based on the mobile robot's travel history; a self-location determination unit that, when the reliability is registered in association with the tentative self-location, selects the self-location detection means that corresponds to a reliability that satisfies a predetermined standard indicating high reliability among the registered reliabilities as a selected self-location detection means, and determines a confirmed self-location based on the self-location calculated using a result detected at the time of interest by the environment detection sensor of the selected self-location detection means; a reliability calculation unit that calculates the reliability based on the tentative self-location, the confirmed self-location determined based on the tentative self-location, and the environment map; and a registration unit that registers the reliability calculated in the reliability calculation unit and information specifying the self-location detection means used to calculate the reliability, in association with a position on the reliability map that corresponds to the confirmed self-location.

12. A self-location determination device that determines a self-location based on a reliability map of a mobile robot, the device having multiple self-location detection means including an environment detection sensor that detects the status of the placement of objects around the mobile robot, an environment map that shows the placement of objects around the mobile robot's travel path, and a calculation unit that calculates the mobile robot's self-location from the detection results of the environment detection sensors and the environment map, wherein the reliability map is a map in which reliability, which is an index indicating the accuracy of the self-location calculated by the self-location detection means, and information identifying the self-location detection means corresponding to the reliability, are registered in association with positions on the map, and the self-location determination device comprises: a tentative self-location determination unit that determines a tentative self-location, which is a position on the reliability map that corresponds to the mobile robot's position at a time of interest estimated based on the mobile robot's travel history; a self-location determination unit that, when the reliability is registered in association with the tentative self-location, selects the self-location detection means corresponding to a reliability that satisfies a predetermined standard indicating high reliability among the registered reliability levels as a selected self-location detection means, and determines a definitive self-location based on the self-location calculated using the result detected at the time of interest by the environment detection sensor of the selected self-location detection means.

13. A reliability map creation program for creating a reliability map used to determine the self-location of a mobile robot equipped with multiple self-location detection means, including an environment detection sensor that detects the status of the placement of objects around the mobile robot, an environment map that shows the placement of objects around the mobile robot's travel path, and a calculation unit that calculates the mobile robot's self-location from the detection results of the environment detection sensor and the environment map, wherein the reliability map is a map in which reliability, which is an index indicating the accuracy of the self-location calculated by the self-location detection means, and information identifying the self-location detection means corresponding to the reliability, are registered in association with positions on the map, and the program is configured to include: a tentative self-location determination unit that determines a tentative self-location, which is a position on the reliability map that corresponds to the mobile robot's position at a time of interest estimated based on the mobile robot's travel history; a self-location determination unit that, when the reliability is registered in association with the tentative self-location, selects the self-location detection means that corresponds to a reliability that meets a predetermined criterion indicating high reliability among the registered reliabilities as a selected self-location detection means, and determines a confirmed self-location based on the self-location calculated using the result detected at the time of interest by the environment detection sensor of the selected self-location detection means; a reliability calculation unit that calculates the reliability based on the tentative self-location, the confirmed self-location determined based on the tentative self-location, and the environment map; and a registration unit that registers the reliability calculated in the reliability calculation unit and information that identifies the self-location detection means used to calculate the reliability, in association with a position on the reliability map that corresponds to the confirmed self-location.

14. A self-location determination program for a mobile robot that determines its own location based on a reliability map of the mobile robot, the program having multiple self-location detection means including an environment detection sensor that detects the status of the placement of objects around the mobile robot, an environment map that shows the placement of objects around the mobile robot's travel path, and a calculation unit that calculates the mobile robot's own location from the detection results of the environment detection sensor and the environment map, wherein the reliability map is a map in which reliability, which is an index indicating the accuracy of the self-location calculated by the self-location detection means, and information identifying the self-location detection means corresponding to the reliability, are registered in association with positions on the map, the program comprising: a tentative self-location determination unit that determines a tentative self-location, which is a position on the reliability map that corresponds to the mobile robot's location at a time of interest, estimated based on the mobile robot's travel history; and When the reliability is registered in association with the tentative self-location, the self-location detection means corresponding to the reliability that satisfies a predetermined standard indicating high reliability among the registered reliabilities is set as a selected self-location detection means, and the self-location determination program functions as a self-location determination unit that determines a definitive self-location based on the self-location calculated using the result detected at the time of interest by the environment detection sensor of the selected self-location detection means.

Images