JP2020107024A - Moving object, learned element, and learned element generation method - Google Patents

Abstract

To provide a moving object having a feature relating to a speed of the moving object that can move at a speed suitable for a surrounding condition and can achieve both high-speed movement and securement of safety.SOLUTION: A moving object 1 comprises: a sensor 11 that acquires ambient information; a storage unit 12 for storing learned elements that are a plurality of sets of learning results of training input information which is information around the moving object 1 and training output information which is a speed corresponding to the training input information; a speed acquisition unit 13 that applies ambient information to the learned element to acquire a speed; a current position acquisition unit 15; a moving mechanism 14; and a movement controller 17 that controls the moving mechanism 14 using a current position. The movement control unit 17 also controls a speed of the moving object 1 so as to be equal to or lower than the acquired speed. In this way, the moving object 1 can move at a speed suitable for the surrounding state, thereby, both high speed movement and securement of safety can be achieved.SELECTED DRAWING: Figure 1

Description

The present invention relates to a moving body or the like that moves autonomously.

In a conventional moving body, an obstacle existing around the moving body is detected, and when the obstacle is detected, the moving body is decelerated or stopped, for example, in order to prevent a collision with the obstacle ( For example, see Patent Document 1).

JP, 2005-103227, A

When the moving body is moving in a moving area with good visibility, a distant obstacle can be detected and a collision with the obstacle can be easily avoided, so that the moving speed can be increased. On the other hand, when the moving body is moving in a moving area where visibility is not good, there is a possibility that an obstacle may jump out from a position that is a blind spot for the moving body. Considering such dangers, in order to improve safety, the maximum speed is set according to the jumping out of the obstacle from the blind spot, regardless of the moving area of the moving body. It was necessary to set the speed so that stop and deceleration could be performed appropriately. As a result, there is a problem that the moving speed of the moving body is reduced as a whole, and efficient movement cannot be realized.

The present invention has been made to solve the above problems, and an object of the present invention is to provide a moving body and the like that can achieve both high-speed movement and ensuring safety.

In order to achieve the above-mentioned object, the moving body according to the present invention is a moving body that moves autonomously, and has training input information that is information about the surroundings of the moving body and a speed of the moving body that corresponds to the surrounding information. A storage unit in which a learner learned by using a plurality of sets of certain output information for training is stored, a sensor for acquiring ambient information that is information around the moving body, and applying the ambient information to the learner. The speed acquisition unit that acquires the speed of the moving body corresponding to the surrounding information, the current position acquisition unit that acquires the current position of the moving body, the moving mechanism that moves the moving body, and the moving mechanism using the current position. And a movement control unit that controls the movement control unit, and the movement control unit controls the speed of the moving body to be equal to or lower than the speed acquired by the speed acquisition unit.
With such a configuration, the speed of the moving body can be controlled so as to be equal to or lower than the speed according to the surrounding situation of the moving body. Therefore, for example, in a moving environment in which the possibility of deceleration or stop is low and the visibility is good, the speed can be increased and high-speed movement can be realized. Further, for example, in a moving environment in which the possibility of deceleration or stop is high and visibility is poor, low-speed movement can be performed, and safety can be ensured. In this way, it is possible to achieve both high-speed movement and ensuring safety.

Further, in the mobile object according to the present invention, the sensor may be a distance measurement sensor that measures the distance to surrounding objects in a plurality of directions.
With such a configuration, the speed can be acquired by using the surrounding information according to the distance measurement result.

Further, the learning device according to the present invention is a learning result of a plurality of sets of input information for training which is information around the moving body and output information for training which is the speed of the moving body according to the surrounding information. When a surrounding information, which is information about the surroundings of a moving body that moves autonomously, is applied, the velocity of the moving body can be acquired according to the surrounding information.
With such a configuration, by using this learning device, it becomes possible to easily acquire the speed according to the situation around the moving body. Further, by moving the moving body at a speed equal to or lower than the speed thus acquired, it becomes possible to achieve both high-speed movement of the moving body and ensuring safety.

Further, the learning device manufacturing method according to the present invention is a learning device manufacturing method that is processed using a training information receiving unit, a learning device manufacturing unit, and a learning device output unit, wherein the training information receiving unit is a mobile unit. The training information receiving step of receiving a plurality of pairs of training input information that is the surrounding information and the training output information that is the speed of the moving body according to the surrounding information, and the learning device manufacturing unit, the training input information. And a learning device manufacturing step of manufacturing a learning device by learning a plurality of sets of training output information, and a learning device output step of the learning device output unit outputting the learning device.
With such a configuration, it becomes possible to manufacture a learning device that is used to acquire the speed according to the situation around the moving body.

Further, in the learning device manufacturing method according to the present invention, the learning device manufacturing method is further processed by using the speed receiving unit, the acceleration generating unit, and the speed changing unit, and the speed receiving unit corresponds to the training input information. The speed receiving step of receiving the speed of the moving body, the acceleration generating step of the acceleration generating section generating the acceleration from the speed received in the speed receiving step, and the speed changing section of the acceleration of deceleration generated in the acceleration generating step. If the absolute value exceeds the threshold, the speed change step of changing the speed accepted in the speed accepting step to a speed smaller than the speed is further provided, and in the training information accepting step, the speed change in the speed changing step is performed. If the above is performed, the training output information, which is the speed changed in the speed changing step, may be received.
With such a configuration, it becomes possible to automatically generate the training information used for learning according to the change in the speed of the moving body.

According to the present invention, it is possible to provide a moving body and the like that can achieve both high-speed movement and ensuring safety.

FIG. 3 is a block diagram showing a configuration of a moving body according to an embodiment of the present invention. Block diagram showing the configuration of the learning device manufacturing apparatus according to the embodiment The flowchart which shows operation|movement of the mobile body by the same embodiment. Flowchart showing the operation of the learning device manufacturing apparatus according to the embodiment The figure which shows an example of the mobile body which has a distance measurement sensor in the same embodiment. The figure which shows an example of the mobile body which has the image sensor in the same embodiment. The figure which shows an example of the movement situation of the mobile body in the same embodiment. The figure which shows an example of the movement situation of the mobile body in the same embodiment. The figure which shows an example of the movement situation of the mobile body in the same embodiment.

Hereinafter, the moving body according to the present invention will be described with reference to the embodiments. Note that, in the following embodiments, the constituent elements and steps given the same reference numerals are the same or correspond to each other, and the repeated description may be omitted. The moving body according to the present embodiment can move at a speed according to the surroundings of the moving body and thus can perform high-speed movement when the possibility of deceleration or stop is low. When is high, it is possible to move slowly.

FIG. 1 is a block diagram showing a configuration of a moving body 1 according to this embodiment. The moving body 1 according to the present embodiment moves autonomously, and includes a sensor 11, a storage unit 12, a speed acquisition unit 13, a moving mechanism 14, a current position acquisition unit 15, and an obstacle detection unit. 16 and a movement control unit 17. The moving body 1 moving autonomously may mean that the moving body 1 does not move according to an operation instruction received from a user or the like, but moves to a destination according to its own judgment. The destination may be, for example, manually determined or automatically determined. Further, the movement to the destination may or may not be performed along the movement route, for example. Further, moving to the destination according to its own judgment may be, for example, moving to the destination by the moving body 1 itself judging the traveling direction, movement, stop, and the like. Alternatively, for example, the moving body 1 may move so as not to collide with an obstacle. The moving body 1 may be, for example, a dolly or a moving robot. The robot may be, for example, an entertainment robot, a monitoring robot, a transfer robot, a cleaning robot, or a robot that captures moving images and still images. , Other robots may be used.

The sensor 11 acquires ambient information that is information about the periphery of the moving body 1. The surrounding information acquired by the sensor 11 is preferably information in front of the moving direction of the moving body 1. Further, in order to be able to detect an obstacle on the front right side in the moving direction of the moving body 1 and an obstacle on the left side in the front direction, the surrounding information includes a predetermined angular range in front of the moving direction of the moving body 1. It is more preferable that at least information (for example, a range of at least 45 degrees to 120 degrees or the like) is included. The sensor 11 may be a distance measuring sensor that measures distances to surrounding objects in a plurality of directions as shown in FIG. 5A, and as shown in FIG. It may be an image sensor to be acquired. 5A and 5B are plan views of the moving body 1 viewed from above. When the sensor 11 is a distance measuring sensor, the surrounding information is the distance for each predetermined angle. When the sensor 11 is an image sensor, the surrounding information is a captured image.

When the moving body 1 can move in any direction, it is preferable that when the sensor 11 is a distance measurement sensor, distance measurement is performed in all directions, and when the sensor 11 is an image sensor. It is preferable that a captured image in any direction can be acquired. This is because any direction of the moving body 1 can be in front of the moving direction. The image sensor that can acquire the captured image in any direction may be, for example, an image sensor that can acquire the captured images in all directions, or an image sensor that can pan (horizontal) about 360 degrees. Note that, as described above, even when the sensor 11 can acquire information in any direction around the moving body 1, the surrounding information acquired by the sensor 11 is information in front of the moving direction. Is preferred.

The distance measuring sensor may be, for example, a laser sensor, an ultrasonic sensor, a distance sensor using microwaves, a distance sensor using a stereo image taken by a stereo camera, or the like. The laser sensor may be a laser range sensor (laser range scanner). Incidentally, those distance measuring sensors are already known, and their explanations are omitted. In the present embodiment, the case where the sensor 11 is a distance measuring sensor and the distance measuring sensor is a laser range sensor will be mainly described. Further, the sensor 11 may have one laser range sensor, or may have two or more laser range sensors. In the latter case, two or more laser range sensors may cover all directions. In addition, when the distance measuring sensor is an ultrasonic sensor or a distance sensor using microwaves, the distances in a plurality of directions may be measured by rotating the distance measuring direction of the distance measuring sensor. Distances in a plurality of directions may be measured by using a plurality of distance measuring sensors arranged at. Measuring the distances in a plurality of directions may be, for example, measuring distances in a plurality of directions at predetermined angular intervals for a predetermined angular range or the entire circumference (360 degrees). The angular interval may be constant, for example, 1 degree interval, 2 degree interval, 5 degree interval, or the like. The information obtained from the distance measuring sensor may be, for example, a distance to a peripheral object regarding each of a plurality of azimuth angles with respect to a certain direction of the moving body 1. By using the distance, it becomes possible to know what kind of object exists around the moving body 1 in the local coordinate system of the moving body 1.

The image sensor may be, for example, a CCD image sensor or a CMOS image sensor. The image sensor may include, for example, an optical system such as a lens for forming an image on the image sensor. The image sensor may be a monocular or a binocular (stereo camera). The image sensor usually captures a moving image, that is, captures consecutive image frames. It is preferable that the frame rate is sufficiently higher than the moving speed. For example, the frame rate may be about 30 fps.

The storage unit 12 stores a learner learned by using a plurality of sets of training input information that is information about the surroundings of the moving body and training output information that is the speed of the moving body according to the surrounding information. To be done. A set of training input information and training output information may be referred to as training information. The learning device may be, for example, a learning result of a neural network (NN), and may be a learning result of a support vector machine (SVM) or a support vector regression (SVR). Alternatively, it may be the learning result of other machine learning.

The neural network may be, for example, a convolutional neural network (CNN: Convolutional Neural Network), or may be another neural network (for example, a neural network including fully connected layers). If the neural network has at least one intermediate layer (hidden layer), the learning of the neural network may be considered to be deep learning. When a neural network is used for machine learning, the number of layers of the neural network, the number of nodes in each layer, the type of each layer (eg, convolutional layer, fully connected layer, etc.) can be selected appropriately. Good. In addition, the number of nodes in the input layer and the output layer is usually determined by the input and the output included in the training information.

Note that the learning device stored in the storage unit 12 may mean that the learning device itself (for example, a function that outputs a value with respect to an input, a model of a learning result, or the like) is stored. It may well be that information such as parameters necessary for configuring the learning device is stored. Even in the latter case, the learning device can be configured by using the information such as the parameters, so that it can be considered that the learning device is substantially stored in the storage unit 12. In the present embodiment, a case where the learning device itself is stored in the storage unit 12 will be mainly described.

The process of storing the learning device in the storage unit 12 does not matter. For example, the learning device may be stored in the storage unit 12 via a recording medium, or the learning device transmitted via a communication line or the like may be stored in the storage unit 12. The storage in the storage unit 12 may be temporary storage in a RAM or the like, or long-term storage. The storage unit 12 can be realized by a predetermined recording medium (for example, a semiconductor memory, a magnetic disk, an optical disk, etc.).

Here, the generation of the learning device will be described. FIG. 2 is a block diagram showing the configuration of the learning device manufacturing apparatus 2 according to this embodiment. The learning device manufacturing apparatus 2 according to the present embodiment includes a speed receiving unit 21, an acceleration generating unit 22, a speed changing unit 23, a training information receiving unit 24, a learning device manufacturing unit 25, and a learning device output unit 26. Equipped with.

The moving body used when acquiring the training information (teacher data) used to generate the learning device may be a moving body that moves autonomously or may be a moving body that does not move autonomously. When acquiring training information using a moving body that moves autonomously, the speed of the moving body at a certain time is accepted by the speed receiving unit 21, and the training is information about the surroundings of the moving body at that time. The training input information may be received by the training information receiving unit 24. When the training information is acquired using a moving body that moves autonomously, the moving body 1 may be used as the moving body, for example. When the training information is acquired using a moving body that moves autonomously, the movement of the moving body may be movement in a real space or may be movement in simulation. In any case, the training information is acquired in an environment in which there is a moving obstacle (for example, another moving body or a person), that is, an environment in which the obstacle may jump out of the blind spot. Is preferred. Further, it is preferable that the training information is acquired in the same moving environment as the moving environment in which the mobile body 1 moves, or in a similar moving environment. Even when the training information is acquired by the simulation, it is preferable that the simulation be performed in a virtual moving environment that is the same environment as the moving environment in which the mobile body 1 moves. A moving body that autonomously moves to acquire training information may move along a pre-generated route or reach a destination such as a dynamic window approach. The route may be moved while being generated in real time. In any case, when a moving object that moves to acquire training information detects an obstacle, it can decelerate, stop, or detour so as not to collide with the obstacle. It is suitable.

When the route is generated in advance, for example, the moving route may be generated by the state space search method, or the moving route may be generated by the node search method. When the route is generated by the state space search method, known algorithms such as RRT (Rapidly-exploring Random Tree), Laplace potential method, and thermal potential method are used by using map information indicating the movable region of the moving body. You may search for a route using. In the state space search method, a movement route is acquired so as to avoid an obstacle that does not move in a space where nodes or the like are not set in advance. When the route search is performed by the node search method, map information in which nodes and links are set in advance is used, and known algorithms such as Dijkstra method, A* algorithm, Bellman-Ford method, Warshall-Floyd method, etc. are used. You may search for a route using. In the node search method, a movement route connecting predetermined nodes is acquired.

When the moving body that moves to acquire the training information does not move autonomously, for example, even if the operator uses a remote control device that is connected to the moving body by wire or wirelessly, Good. In this case, the moving direction and speed of the moving body are operated by the operator. Further, in this case, even if an obstacle jumps out from a position that is a blind spot for the moving body, the moving speed of the moving body is set by the operator so that the moving speed of the moving body can be appropriately avoided. It is preferably controlled. That is, it is preferable that the operation is performed at a high speed in a moving environment with good visibility and at a lower speed than a moving environment with good visibility in a moving environment with poor visibility.

In the present embodiment, a case will be mainly described in which a moving body that moves to acquire training information moves autonomously. Further, in the present embodiment, a case where the learning device manufacturing apparatus 2 accepts training input information that is surrounding information and speed in real time directly from a moving body that is moving to acquire training information is mainly described. Explained.

The speed receiving unit 21 receives the speed of the mobile body corresponding to the training input information received by the training information receiving unit 24. That is, the speed of the mobile body at the time when the information around the mobile body, which is the input information for training, is acquired is accepted. This speed may be a scalar amount having a magnitude of 0 or more, that is, a speed magnitude in the forward direction. Further, when the received speed is a vector, it may be converted into a scalar. The speed may be stored in a recording medium (not shown). The speed reception unit 21 passes the received speed to the acceleration generation unit 22 and the speed change unit 23.

The speed accepting unit 21 may receive the speed transmitted via a wired or wireless communication line, for example. The speed accepting unit 21 may or may not include a device (for example, a modem or a network card) for performing the acceptance. The speed accepting unit 21 may be implemented by hardware or software such as a driver that drives a predetermined device.

The acceleration generation unit 22 generates acceleration from the speed received by the speed reception unit 21 and passes it to the speed change unit 23. In the generation of this acceleration, the speed accepted by the speed accepting unit 21 in the past may also be used. For example, when the speed reception unit 21 receives the speed V(T) at each time interval ΔT, the acceleration generation unit 22 may calculate the acceleration A(T) at the time T according to the following equation.
A(T)=(V(T)-V(T-1))/ΔT

The time T is an integer indicating the time, and the time interval between adjacent times, for example, the time interval from the time T-1 to the time T is ΔT. Further, when the first speed is accepted, there is no speed earlier than that by ΔT, and thus the acceleration cannot be calculated. Therefore, in that case, for example, the acceleration may not be calculated, or the acceleration may be zero.

The speed changing unit 23 determines whether the absolute value of the deceleration acceleration generated by the acceleration generating unit 22 exceeds a threshold value. That is, the speed changing unit 23 determines whether the generated acceleration indicates deceleration and whether the absolute value of the acceleration exceeds the threshold value. If the acceleration has a negative value, it may be determined that the acceleration indicates deceleration, and if not, it may be determined that the acceleration does not indicate deceleration. When the generated acceleration indicates deceleration and the absolute value of the acceleration exceeds the threshold value, the speed changing unit 23 sets the speed accepted by the speed accepting unit 21 to be higher than the speed. Change to a smaller speed. That is, the speed is changed so as to decelerate. When the absolute value of the generated deceleration acceleration is equal to the threshold value, the speed changing unit 23 may or may not change the speed.

When the absolute value of the deceleration acceleration exceeds the threshold value, the moving body that is autonomously moving is performing rapid deceleration, for example, by detecting an obstacle that has jumped out of a position that is a blind spot for the moving body. Respond to the situation. Therefore, the threshold value may be set to a value that cannot be taken by movement in a situation where no obstacle jumps out of the blind spot. Further, in such a situation, the speed before deceleration is not suitable for the surrounding circumstances and is too high, so the speed is changed to a smaller speed. Therefore, it is preferable that the change in speed results in a high speed in a moving environment with good visibility and a low speed in a moving environment with poor visibility as compared with a moving environment with good visibility.

Further, the changed speed may be one or more predetermined values, or may be a non-predetermined value. If the speed after the change is one value, the one speed can appropriately avoid the collision with the obstacle even if the obstacle jumps out from the position that becomes the blind spot. It may be set to a speed that allows it. That rate is usually the rate of creep. Further, when the changed speed has a plurality of predetermined values, for example, the changed speed may decrease as the absolute value of the deceleration acceleration increases. If the speed after change is not predetermined, for example, the speed after change is a value obtained by multiplying the speed before change by a coefficient smaller than 1 such as 0.3 or 0.5. Good. The coefficient may be, for example, a predetermined value, or may be a coefficient that becomes smaller as the absolute value of the deceleration acceleration increases.

When the speed is changed, the speed changing unit 23 passes the changed speed to the training information receiving unit 24. When the speed is not changed, the speed is not changed, that is, the speed. The speed accepted by the acceptance unit 21 may be passed to the training information acceptance unit 24.

Further, when the absolute value of the generated deceleration acceleration does not exceed the threshold value, the speed changing unit 23 performs fractional processing on the speed accepted by the speed accepting unit 21 so that the training output information is discrete. It may be a finite number. In this case, even when the absolute value of the generated deceleration acceleration does not exceed the threshold value, the speed changing unit 23 makes the speed accepted by the speed accepting unit 21 smaller than the speed, for example. May be rounded to The fractional treatment may be, for example, if the speed of the moving body is 40 (m/min) or more, round it to 40 (m/min), and set it to 30 (m/min) or more and less than 40 (m/min). For example, rounding to 30 (m/min), rounding to 20 (m/min) or more and less than 30 (m/min) may be performed. It should be noted that the fractional processing also changes the accepted speed to a smaller speed, but the degree of the change depends on the speed when the absolute value of the generated deceleration acceleration exceeds the threshold value. May be smaller than the degree of change. Further, for speeds equal to or lower than the minimum unit of rounding, for example, rounding up may be performed so as to be the minimum unit, or rounding may not be performed. For example, as described above, when rounding is performed in units of 10 (m/min), a speed less than 10 (m/min) may be rounded up to 10 (m/min), or as it is. It does not have to be changed.

The training information accepting unit 24 accepts a plurality of sets of training input information that is information about the surroundings of the moving body and training output information that is the speed of the moving body according to the surrounding information. The training input information received by the training information receiving unit 24 is preferably the same type of information as the surrounding information acquired by the sensor 11 of the mobile unit 1. That is, when the sensor 11 obtains the distance to the object around the moving body 1, the training input information is also preferably the distance to the object around the moving body. When the captured image around the moving body 1 is acquired, the training input information is also preferably the captured image around the moving body. When the speed changing unit 23 changes the speed, the training information receiving unit 24 receives the training output information which is the speed changed by the speed changing unit 23, and the speed changing unit 23 changes the speed. If not, the training output information, which is the speed accepted by the speed accepting unit 21, may be accepted. In addition, as described above, when the speed changing unit 23 performs the speed fraction processing, the training output information may be the speed after the fraction processing. Further, the speed that is the output information for training may be a scalar amount. It is preferable that the output information for training becomes high speed in a moving environment with good visibility, and becomes slower than in a moving environment with good visibility in a moving environment with poor visibility.

The training information accepting unit 24 may or may not include a device for accepting. Further, the training information accepting unit 24 may be realized by hardware, or may be realized by software such as a driver that drives a predetermined device.

Normally, the speed is accepted and the acceleration is generated for each ΔT, and when the acceleration satisfies a predetermined condition, the speed is also changed. Then, the training information receiving unit 24 receives training information including, for each ΔT, training input information that is information about the surroundings of the moving body and training output information that is the speed of the moving body. The speed that is the output information for training may be the moving speed of the moving body, or may be the speed after the moving speed is changed as described above. The training information reception unit 24 receives a plurality of training information by repeating a series of movements by the moving body and receiving the training information repeatedly. The received plurality of pieces of training information may be stored in a recording medium (not shown).

The learning device manufacturing unit 25 manufactures a learning device by learning a plurality of sets of training input information and training output information. This learning device is the result of machine learning of a plurality of sets of input information for training which is information around the moving body and output information for training which is the speed of the moving body according to the surrounding information. Therefore, for example, when the surrounding information that is the information about the surroundings of the moving body 1 that autonomously moves is applied to the learning device, the speed of the moving body 1 according to the surrounding information can be acquired. .. Details of this learning will be described later.

The learning device output unit 26 outputs the learning device manufactured by the learning device manufacturing unit 25. The learning device output by the learning device output unit 26 is stored in the storage unit 12 of the mobile unit 1 shown in FIG. Here, this output may be, for example, transmission to a predetermined device via a communication line, storage in a recording medium, or delivery to another component. The learning device output unit 26 may or may not include a device that outputs (for example, a communication device). Further, the learning device output unit 26 may be realized by hardware, or may be realized by software such as a driver that drives those devices.

In addition, although the case where the learning device manufacturing apparatus 2 receives the information on the speed and the surroundings in real time from the moving body that moves to acquire the training information has been mainly described, but the learning device manufacturing apparatus 2 does not have to do so. Such information may be previously stored in a recording medium (not shown). Then, the learning device manufacturing apparatus 2 may accept the information by reading the speed and surrounding information from the recording medium sequentially or collectively. In addition, the learning device manufacturing apparatus 2 may receive the speed and surrounding information transmitted from another device by sequentially or collectively receiving the information.

Moreover, although the case where the learning device manufacturing apparatus 2 also receives the speed has been described, the learning device manufacturing apparatus 2 does not have to do so. For example, as described above, when the moving body that moves to acquire the training information is remotely controlled by the operator so as to be an ideal movement, the speed acquired according to the movement and The training information receiving unit 24 may receive a set of information about the surroundings. Also in this case, fractional processing may be performed on the speed. The reception may be reception of information transmitted via a wired or wireless communication line, or may be reception of information read from a predetermined recording medium (for example, a magnetic disk or a semiconductor memory). In this case, the learning device manufacturing apparatus 2 may not include the speed accepting unit 21, the acceleration generating unit 22, and the speed changing unit 23.

Further, in order to acquire the training information, when a plurality of autonomously moving mobile bodies are actually moved or are moved on a simulation, the speed reception unit 21 and the training information reception unit 24 make a plurality of movements. You may accept information from your body. By doing so, it becomes possible to collect more training information in a shorter period of time and improve the efficiency of collecting training information.

Next, machine learning by the learning device manufacturing unit 25 will be described. This machine learning is supervised learning, neural network learning, support vector machine learning, support vector regression learning, or other machine learning. When the training output information has a small number of stages, the machine learning may be, for example, a neural network or a support vector machine that performs multilevel classification. As a case where the training output information has a small number of stages, for example, there is a case where the speed changing unit 23 is also performing fractional processing of the speed. On the other hand, when the training output information is a large number of multistages or real numbers in a predetermined range, the machine learning may be support vector regression. As a case where the training output information is a large number of multi-stages or a real number in a predetermined range, for example, there is a case where the speed changing unit 23 does not perform the speed fraction processing.

Here, the training input information will be described separately as follows: (1) information corresponding to a distance measurement result, and (2) information corresponding to a captured image.

(1) When the training input information corresponds to the distance measurement result In this case, the training input information may be the distance measurement result itself. When the training input information is the distance measurement result itself, a set of distances to surrounding objects in each distance measurement direction may be the training input information. For example, when the sensor 11 is a laser range sensor and the distance measurement is performed in the range of 120 degrees for every 1 degree in the forward direction of movement, 120 distance measurement results become the training input information. The distance measurement result may be normalized by preprocessing, for example. For example, the longest measurement distance of the laser range sensor may be "1", and the shortest measurement distance may be "0".

Further, using the distance measurement result, a set of information indicating whether or not an obstacle exists within a predetermined threshold distance is generated for each distance measurement direction, and the generated information is also used as training input information. Good. For example, when the sensor 11 is a laser range sensor and distance measurement is performed in the range of 120 degrees for every 1 degree ahead of the movement direction, when the measurement result is less than or equal to the threshold distance for each distance measurement direction, the movement is performed. The information (for example, “1” may be displayed) indicating that an obstacle exists near the body, and when the measurement result exceeds the threshold distance, indicates that there is no obstacle near the moving body. The training input information may be generated to be information (for example, “0” may be used). In this case, for example, 120 pieces of binary information indicating whether or not an obstacle is present near the moving body becomes the training input information.

When learning using the training input information (1) is performed by a neural network, each layer of the neural network may be, for example, a fully connected layer.

(2) When the training input information corresponds to the captured image In this case, the training input information may be the captured image itself. In this case, for example, learning by a convolutional neural network may be performed. The convolutional neural network may have the same configuration as the convolutional neural network used in object category recognition. Then, instead of the object category, the training output information having a plurality of stages of speed may be used. Each layer of the convolutional neural network used in the object category recognition may have a plurality of convolutional layers in order from the input layer side, for example. Also, a pooling layer may be present after the at least one convolutional layer. Further, one or more fully-bonded layers may be present after the plurality of convolutional layers. Also, a softmax layer may exist immediately before the output layer.

In any of the above (1) and (2), when the learning device is a neural network, for example, it is preferable to update the parameters so that overfitting does not occur. Learning may be performed using a method or the like. A known method can be used as a learning method in machine learning, and a detailed description thereof will be omitted.

The training input information and the training output information used in the machine learning may be information that has been preprocessed as described above. The training input information may be subjected to preprocessing such as normalization or binary conversion. Further, for example, the training output information may be subjected to preprocessing such as fraction processing.

Returning to FIG. 1, the speed acquisition unit 13 applies the surrounding information acquired by the sensor 11 to the learning device stored in the storage unit 12, thereby acquiring the speed of the mobile body 1 corresponding to the surrounding information. To do. Specifically, the speed acquisition unit 13 can acquire the speed, which is the output of the learning device, by inputting the ambient information acquired by the sensor 11 to the learning device. When inputting the ambient information to the learning device, for example, the ambient information itself may be input, or the ambient information may be input after performing a predetermined preprocessing. For example, as described above, when the training input information is normalized or converted into binary information and learning is performed, the same applies when the ambient information is applied to the learning device. You may apply it, after processing.

The moving mechanism 14 moves the moving body 1. The moving mechanism 14 may or may not be one that can move the moving body 1 in all directions, for example. Being able to move in all directions means being able to move in any direction. The moving mechanism 14 may include, for example, a traveling unit (for example, wheels) and a driving unit (for example, a motor or an engine) that drives the traveling unit. Further, the moving mechanism 14 may have a mechanism capable of acquiring the speed of the wheels or the like, for example, an encoder or the like. When the moving mechanism 14 is capable of moving the moving body 1 in all directions, the traveling portion may be an omnidirectional moving wheel (for example, an omni wheel, a mecanum wheel, etc.). Since a known mechanism can be used as the moving mechanism 14, detailed description thereof will be omitted.

The current position acquisition unit 15 acquires the current position of the mobile body 1. The acquisition of the current position may be performed, for example, by using wireless communication, may be performed by using a measurement result of a distance to a surrounding object, or may be performed by capturing an image of the surroundings. , Other means capable of obtaining the current position may be used. As a method of acquiring the current position using wireless communication, for example, a method using GPS (Global Positioning System), a method using indoor GPS, a method using a nearest wireless base station, and the like are known. In addition, for example, as a method of acquiring the current position by using a measurement result of a distance to a surrounding object or capturing an image of a surrounding area, for example, SLAM (Simultaneous Localization and Mapping) is known. You may use the method. As the measurement result of the distance to the surrounding object, for example, the measurement result of the sensor 11 which is the distance measuring sensor may be used. Further, as the surrounding captured image, for example, a captured image acquired by the sensor 11 which is an image sensor may be used. Further, when a map created in advance (for example, a measurement result of the distance to the surrounding object or a map having a captured image) is stored, the current position acquisition unit 15 uses the map to display the surrounding area. The current position may be acquired by specifying the position corresponding to the measurement result of the distance to the object, and the current position may be acquired by shooting the surrounding image and using the map to specify the position corresponding to the shooting result. The position may be acquired. Further, the current position acquisition unit 15 may acquire the current position using, for example, an autonomous navigation device. In addition, it is preferable that the current position acquisition unit 15 acquires the current position including the direction (direction) of the moving body 1. The direction may be indicated by an azimuth angle measured clockwise, for example, with 0 degrees north as north, or may be indicated by information indicating another direction. The orientation may be acquired by an electronic compass or a geomagnetic sensor.

The obstacle detection unit 16 detects an obstacle existing around the moving body 1. For example, the obstacle detection unit 16 may detect an obstacle by using a distance measurement result regarding the periphery of the moving body 1, or may detect an obstacle by using a captured image of the periphery of the moving body 1. When detecting an obstacle using the distance measurement result, and when the sensor 11 is a distance measurement sensor, the obstacle detection unit 16 detects the obstacle using the distance measurement result acquired by the sensor 11. You may go. In addition, when the moving body 1 has a distance measuring sensor for detecting an obstacle, the obstacle detecting unit 16 detects the obstacle using the distance measuring result acquired by the distance measuring sensor. Good. Further, in the case of detecting an obstacle using the captured image, and when the sensor 11 is an image sensor, the obstacle detection unit 16 uses the captured image acquired by the sensor 11 to detect the obstacle. You can go. In addition, when the moving body 1 has an image sensor for detecting an obstacle, the obstacle detection unit 16 may perform obstacle detection using a captured image obtained by the image sensor.

When the obstacle detection is performed using the distance measurement result, the obstacle detection unit 16 detects that an object exists near the moving body 1 based on the distance measured by the distance measurement sensor, for example. The object may be detected as an obstacle. For example, when the distance to the surrounding object measured by the distance measuring sensor becomes equal to or less than a predetermined threshold, the obstacle detection unit 16 may detect the object as an obstacle. The distance to the surrounding object may be, for example, the distance from the distance measuring sensor, the distance from the outer edge of the moving body 1, or the position where the outer edge of the moving body 1 is virtually expanded. It may be a distance from, or a distance from another reference. The threshold value may be one, or may be two or more.

When the obstacle detection is performed using the captured image, the obstacle detection unit 16 detects, for example, that an object is present near the moving body 1 from the captured image captured by the image sensor. The object may be detected as an obstacle. The object recognition may be performed by, for example, pattern matching or the like, segmentation using a convolutional neural network, or another method. Whether or not the object included in the captured image exists near the moving body 1 may be determined using the type of the object and the size (for example, length or area) of the object in the captured image. .. For example, when the actual size of the object included in the captured image is large (for example, the object is another moving object), the object moves even if the size of the object in the captured image is not so large. It may be determined that it exists near the body 1. On the other hand, for example, when the actual size of the object included in the captured image is small (for example, when the object is a small corrugated cardboard box), the object is only displayed when the size of the object in the captured image is large. It may be determined that the mobile body 1 exists near the mobile body 1. A known method can be used as a method of detecting an obstacle, and a detailed description thereof will be omitted.

The movement control unit 17 controls the movement of the moving body 1 by controlling the movement mechanism 14 using the current position acquired by the current position acquisition unit 15. The movement control may be control such as the movement direction of the moving body 1 or the start/stop of movement. For example, when the moving route is set, the movement control unit 17 may control the moving mechanism 14 so that the moving body 1 moves along the moving route. More specifically, the movement control unit 17 may control the movement mechanism 14 so that the current position acquired by the current position acquisition unit 15 is along the movement route. In addition, the movement control unit 17 may control movement using a map. Further, the movement control unit 17 may perform movement control while dynamically generating a route, such as a dynamic window approach. Since the control of the moving mechanism 14 by the movement control unit 17 is known, its detailed description is omitted.

The movement control unit 17 controls the speed of the moving body 1 to be equal to or lower than the speed acquired by the speed acquisition unit 13. Therefore, for example, the movement control unit 17 sets the acquired speed to be the acquired speed when the command speed is higher than the speed acquired by the speed acquisition unit 13, or sets the speed to be less than the acquired speed. The command speed may be changed. When the commanded speed is a speed vector, each component of the speed vector, which is the commanded speed, is multiplied by a positive real number smaller than 1 so that the size of the speed vector is equal to or smaller than the acquired speed. May be. In addition, for example, the movement control unit 17 may generate the command speed such that the command speed is equal to or less than the speed acquired by the speed acquisition unit 13. When the commanded speed is a speed vector, the commanded speed may be generated so that the magnitude of the speed vector is equal to or less than the acquired speed. From the viewpoint of reducing the influence on the movement control unit 17, when the command speed is higher than the acquired speed, the command speed is changed to the acquired speed, and in other cases, the speed is changed. Is preferably not performed. In this case, the movement control unit 17 generates a command speed so as to move at a high speed (that is, to move at a speed assuming that there is no obstacle around) when generating the command speed. Is preferred.

The movement control unit 17 also controls the movement mechanism 14 so as to prevent collision with the obstacle detected by the obstacle detection unit 16. The movement control unit 17 controls the moving mechanism 14 so that the moving body 1 stops when an obstacle existing in a stop area, which is an area set around the moving body 1, is detected, for example. If no obstacle is detected in the stop area, but an obstacle existing in the deceleration area, which is an area set outside the stop area, is detected, the moving body 1 is faster than the speed at that time. The moving mechanism 14 may be controlled to decelerate. When the moving body 1 is decelerated, the speed after deceleration may or may not be determined. In the latter case, for example, deceleration may be performed so that the speed is determined relative to the speed at that time (for example, 50% speed). Further, the movement control unit 17 may control the movement mechanism 14 so that the moving body 1 passes through a route that bypasses the detected obstacle, for example, in order to avoid collision with the detected obstacle. In that case, the movement control unit 17 may newly generate a movement route that does not pass through the detected position of the obstacle and perform a movement according to the movement route.

Next, the operation of the mobile unit 1 will be described with reference to the flowchart of FIG.
(Step S101) The movement control unit 17 determines whether or not to move. If the movement is to be performed, the process proceeds to step S102. If not, the process of step S101 is repeated until it is determined that the movement is performed. It should be noted that the movement control unit 17 may determine that the movement is to be performed, for example, when autonomous movement is performed according to the movement route. Further, the movement control unit 17 may determine that the movement is not performed, for example, when the destination is reached.

(Step S102) The current position acquisition unit 15 acquires the current position.

(Step S103) The sensor 11 acquires ambient information about the moving body 1.

(Step S104) The speed acquisition unit 13 acquires the speed corresponding to the surrounding information by applying the surrounding information acquired at Step S103 to the learning device stored in the storage unit 12.

(Step S105) The movement control unit 17 controls the movement of the moving body 1 by controlling the movement mechanism 14 using the current position acquired in step S102. In this movement control, the movement control unit 17 performs control so that the speed of the moving body 1 does not exceed the speed acquired in step S104, that is, the speed is equal to or lower than the speed. This movement control may be, for example, control of autonomous movement toward the destination. By repeating the control of step S105, the moving body 1 may reach the destination from the departure place.

(Step S106) The obstacle detection unit 16 determines whether or not there is an obstacle in the area where the obstacle is detected. If there is an obstacle, the process proceeds to step S107, and if not, the process returns to step S101 to continue the movement.

(Step S107) The movement control unit 17 performs control according to the detection of the obstacle in step S106. The control is, for example, control for stopping the moving body 1 or control for decelerating. Then, the process returns to step S101. When the moving body 1 is decelerated in response to the detection of the obstacle, the movement control unit 17 limits the upper limit of the moving speed to that after deceleration, and then returns to step S101 to continue the movement. May be. In that case, the upper limit of the moving speed may be released when the obstacle is no longer detected (NO in step S106). Further, when the moving body 1 is stopped in response to the detection of the obstacle, the movement control unit 17 continues the stop until the obstacle detection unit 16 detects the obstacle, and the obstacle is not detected. After that (for example, after a human being as an obstacle moves), the process may return to step S101 to restart the movement.

The order of processing in the flowchart of FIG. 3 is an example, and the order of each step may be changed as long as the same result can be obtained. For example, when the current position is acquired using the surrounding information, the current position may be acquired (step S102) after the acquisition of the surrounding information (step S103). Further, in the flowchart of FIG. 3, the processing is ended by powering off or interruption for aborting the processing.

Next, the operation of the learning device manufacturing apparatus 2 will be described with reference to the flowchart of FIG. In this flowchart, a case will be described in which the learning device manufacturing apparatus 2 receives speed and surrounding information in real time from a moving body that moves autonomously to acquire training information.

(Step S201) The speed reception unit 21 determines whether or not the speed of a moving body that moves autonomously for acquiring training information has been received. Then, if the speed is accepted, the process proceeds to step S202. If not, the process of step S201 is repeated until the speed is accepted.

(Step S202) The acceleration generation unit 22 uses the received speed to generate acceleration.

(Step S203) The speed changing unit 23 determines whether the acceleration indicates deceleration and the absolute value thereof exceeds the threshold value. If the acceleration indicates deceleration and the absolute value thereof exceeds the threshold value, the process proceeds to step S204, and if not, the process proceeds to step S205.

(Step S204) The speed changing unit 23 changes the speed accepted in step S201 to a speed smaller than the speed.

(Step S205) The training information receiving unit 24 receives training information. The training information is input information for training, which is information about the surroundings of the moving body, corresponding to the speed accepted in step S201, the speed accepted by the speed accepting unit 21, or the speed changed by the speed changing unit 23. And the output information for training that is The received training information may be stored in a recording medium (not shown).

Even if it is determined to be NO in step S203, the speed changing unit 23 may perform fractional processing of speed. In that case, the training output information received in step S205 may be the speed changed by the speed changing unit 23.

(Step S206) The training information reception unit 24 determines whether to end the reception of training information. Then, if the acceptance of the training information is ended, the process proceeds to step S207, and if not, the process returns to step S201. Note that the training information reception unit 24 may determine that the reception of the training information is ended when, for example, a predetermined number of pieces of training information have been received, and the movement of the mobile body for acquisition of the training information is performed. It may be determined that the acceptance of the training information is terminated when is ended.

(Step S207) The learning device manufacturing unit 25 manufactures the learning device by performing machine learning using the plurality of training information received by the training information receiving unit 24.

(Step S208) The learning device output unit 26 outputs the learning device manufactured in step S207. Then, a series of processes for manufacturing the learning device ends.

Note that the order of processing in the flowchart of FIG. 4 is an example, and the order of each step may be changed as long as the same result can be obtained.

Next, operations of the moving body 1 and the learning device manufacturing apparatus 2 according to the present embodiment will be described using a specific example. In this specific example, the sensor 11 is assumed to be a laser range sensor.

First, in order to acquire the training information, a plurality of moving bodies that move autonomously are operated in the same moving area (for example, in a certain factory). Then, it is assumed that the learning device manufacturing apparatus 2 receives the moving speed and the surrounding information. Further, in this specific example, when the absolute value of the deceleration acceleration exceeds the threshold value, the accepted speed is changed to 10 (m/min), and even if not, 10, 20, It is assumed that the fraction of the received speed is rounded down so that the speed becomes the closest speed of either 30 or 40 (m/min). Specifically, the speed 35 (m/min) is changed to the speed 30 (m/min). In addition, when the accepted speed is less than 10 (m/min), the rounding-up process to 10 (m/min) shall be performed. Specifically, the speed of 8.5 (m/min) is changed to the speed of 10 (m/min).

When the mobile body is moved to acquire the training information, the speed of the mobile body and the result of distance measurement of the surroundings when moving at the speed are transmitted to the learning device manufacturing apparatus 2 accordingly. When the velocity is received, the acceleration is generated (steps S201 and S202), and when the acceleration is the deceleration and the absolute value exceeds the threshold, the velocity is changed to 10 (m/min), Even if not, the fraction processing is performed as described above (steps S203 and S204). Then, the training information including the changed speed and the surrounding distance measurement result received from the moving body is received by the training information receiving unit 24 and accumulated in a recording medium (not shown) (step S205). In this way, the training information will be sequentially accumulated (steps S201 to S206). After that, when the moving body that has moved to acquire the training information stops moving, the learning device is manufactured using the accumulated training information and is output (steps S207 and S208). This learning device outputs one of speeds of 10, 20, 30, and 40 (m/min) in response to the input of the distance measurement result around the moving body 1.

After the learning device manufactured in this way is stored in the storage unit 12, the moving body 1 operates. The mobile unit 1 autonomously moves from a predetermined starting point to a destination. In addition, the moving body 1 is assumed to generate a command speed (for example, a command speed in the case where it is assumed that there is no obstacle in the surroundings) for moving the route from the departure place to the destination at a high speed. In the autonomous movement, the current position is acquired and the distance measurement result of the surroundings is acquired (steps S101 to S103). After that, the speed is acquired by inputting the result of distance measurement of the surroundings to the learning device (step S104).

As shown in FIG. 6A, the moving body 1 is moving between the obstacles B1 and B2, but when there is no obstacle ahead of the moving direction indicated by the arrow in the figure, for example, The output from the learning device according to the input of the distance measurement result is 40 (m/min). Therefore, the movement control unit 17 controls the movement so that the speed becomes 40 (m/min) or less (step S105). That is, when the command speed exceeds 40 (m/min), it is changed to 40 (m/min), and when the command speed is 40 (m/min) or less, the command speed remains unchanged. Will be used.

On the other hand, as shown in FIG. 6B, the moving body 1 is moving between the obstacles B1 and B2, and a part of the obstacles B1 and B2 is present in front of the moving direction, so that the blind spot is reduced. When the distances to the obstacles B1 and B2 are not so close, the output from the learning device corresponding to the input of the distance measurement result is 20 (m/min), for example. Therefore, the movement control unit 17 controls the movement so that the speed becomes 20 (m/min) or less (step S105).

In addition, as shown in FIG. 6C, the moving body 1 is moving between the obstacles B1 and B2, and a part of the obstacles B1 and B2 is present in front of the moving direction, so that the blind spot is When there is an object and the distance to the obstacle B2 is short, the output from the learning device corresponding to the input of the distance measurement result is 10 (m/min), for example. Therefore, the movement control unit 17 controls the movement so that the speed is 10 (m/min) or less (step S105).

Further, when an obstacle is detected during the movement of the moving body 1, the movement control is performed so that the collision with the obstacle is avoided according to the detection of the obstacle (step S106). , S107).

In this specific example, when the command speed of the movement control unit 17 is lower than the speed obtained by using the learning device, the command speed is used as it is, and the command speed is obtained by using the learning device. When the commanded speed is higher than the specified speed, the commanded speed is changed to the speed obtained by using the learning device. However, the speed may not be so. For example, movement control may be performed so as to move at the speed obtained by applying the ambient information to the learning device.

As described above, according to the moving body 1 according to the present embodiment, it is possible to travel at an appropriate speed obtained by applying the surrounding information to the learning device. Therefore, for example, in a moving environment in which the possibility of deceleration or stop is low and the visibility is good, the speed can be increased and high-speed movement can be realized. In addition, for example, in a moving environment where there is a high possibility of deceleration or stoppage and the visibility is not good, it is possible to run at low speed, and avoid collision even if another moving body etc. jumps out of the blind spot. It is possible to ensure safety. In this way, it is possible to achieve both high-speed movement and securing of safety for the moving body 1, and efficient movement can be safely realized.

Further, according to the learning device manufacturing apparatus 2 according to the present embodiment, the learning device used when the moving body 1 is moved is generated according to the actual movement of the moving body for acquiring the training information or the movement on the simulation. You will be able to. Further, by providing the speed changing unit 23, the speed can be appropriately changed according to the acceleration of the moving body that moves to acquire the training information. Further, the speed changing unit 23 can also perform a fractional process regarding the speed of the moving body, and the speed output from the learning device can be set to a plurality of stages.

In the present embodiment, the surrounding information and the surrounding information of the moving body, which is the training input information used at the time of learning, may be information in time series within a certain range. For example, the latest ambient information and the ambient information that is ΔT earlier than that may be applied to the learning device. Further, similarly, the training input information including the latest surrounding information of the moving body and the surrounding information before ΔT may be used for learning. By doing so, it is considered that the speed can be acquired with higher accuracy.

In addition, in the present embodiment, when the moving body that moves to acquire the training information moves autonomously, the received speed is changed when a predetermined condition regarding acceleration is satisfied. Then I explained, but it doesn't have to be. For example, by performing a moving average process on the speed of a moving body that moves autonomously, the speed of the moving body is smoothed in time series. For example, when the moving body is suddenly decelerated, the speed after the moving average corresponding to the position where the sudden deceleration is performed will decrease according to the speed after the sudden deceleration. Therefore, when the speed after moving average is used as the speed of the moving body that moves autonomously, the training input information that is the surrounding information acquired by the moving body and the training that is the speed after moving average are used. The training output information may be received by the training information receiving unit 24.

Further, in the above embodiments, each process or each function may be realized by being centralized by a single device or a single system, or distributed by a plurality of devices or multiple systems. It may be realized by

Further, in the above-described embodiment, when information is exchanged between the respective constituent elements, for example, when two constituent elements for exchanging the information are physically different, one of the constituent elements is used. It may be performed by outputting the information and receiving the information by the other component, or when the two components that transfer the information are physically the same, one component The processing may be performed by shifting from the processing phase corresponding to the above to the processing phase corresponding to the other component.

In addition, in the above-described embodiment, information related to processing executed by each component, for example, information that each component has received, acquired, selected, generated, transmitted, or received. Information such as thresholds, mathematical expressions, addresses, etc. used by each component in processing may be held in a recording medium (not shown) temporarily or for a long period of time, even if not specified in the above description. In addition, the storage of information in the recording medium (not shown) may be performed by each component or the storage unit (not shown). Further, the reading of information from the recording medium (not shown) may be performed by each component or the reading unit (not shown).

Further, in the above-described embodiment, when the information used in each component or the like, for example, the information such as the threshold value or the address used in the process of each component or various setting values may be changed by the user, Even if not explicitly stated in the description, the user may or may not be able to change the information as appropriate. When the user can change the information, the change is realized by, for example, a reception unit (not shown) that receives a change instruction from the user and a change unit (not shown) that changes the information according to the change instruction. May be. The reception of the change instruction by the reception unit (not shown) may be reception from an input device, reception of information transmitted via a communication line, or reception of information read from a predetermined recording medium. ..

Further, in the above-described embodiment, when two or more constituent elements included in the mobile unit 1 or the learning device manufacturing apparatus 2 have a communication device, an input device, or the like, the two or more constituent elements are physically single. It may have a device or it may have a separate device.

Further, in the above embodiment, each component may be configured by dedicated hardware, or a component that can be implemented by software may be implemented by executing a program. For example, each component can be realized by a program execution unit such as a CPU reading and executing a software program recorded in a recording medium such as a hard disk or a semiconductor memory. When executing the program, the program execution unit may execute the program while accessing the storage unit or the recording medium. The program may be executed by being downloaded from a server or the like, or may be executed by reading the program recorded in a predetermined recording medium (for example, an optical disk, a magnetic disk, a semiconductor memory, etc.). Good. Further, this program may be used as a program that constitutes a program product. Moreover, the computer that executes the program may be singular or plural. That is, centralized processing may be performed or distributed processing may be performed.

Further, it is needless to say that the present invention is not limited to the above-mentioned embodiments, and various modifications can be made, which are also included in the scope of the present invention.

As described above, according to the moving body and the like of the present invention, it is possible to obtain the effect of achieving both high-speed movement and ensuring safety, and is useful as, for example, a moving body that moves autonomously.

1 moving body, 2 learning device manufacturing apparatus, 11 sensor, 12 storage unit, 13 speed acquisition unit, 14 moving mechanism, 15 current position acquisition unit, 17 movement control unit, 21 speed reception unit, 22 acceleration generation unit, 23 speed change Part, 24 training information acceptance part, 25 learning device manufacturing part, 26 learning device output part

Claims (5)

A mobile body that moves autonomously,
A storage unit that stores a learning device that is learned by using a plurality of sets of training input information that is information about the surroundings of a moving body and output information for training that is the speed of the moving body according to the surrounding information. When,
A sensor that acquires ambient information that is information around the moving body,
By applying the surrounding information to the learning device, a speed acquisition unit that acquires the speed of the moving body corresponding to the surrounding information,
A current position acquisition unit for acquiring the current position of the moving body,
A moving mechanism for moving the moving body,
A movement control unit that controls the movement mechanism using the current position,
The movement control unit controls the movement speed of the movement body to be equal to or lower than the speed acquired by the speed acquisition unit. The moving body according to claim 1, wherein the sensor is a distance measuring sensor that measures distances to surrounding objects in a plurality of directions. A training device which is a training result of a plurality of sets of training input information which is information around the moving body and training output information which is the speed of the moving body according to the surrounding information,
A learning device capable of acquiring the speed of the moving body according to the surrounding information when the surrounding information that is the surrounding information of the moving body that autonomously moves is applied. A learning device manufacturing method that is processed using a training information receiving unit, a learning device manufacturing unit, and a learning device output unit,
A training information receiving step in which the training information receiving unit receives a plurality of sets of training input information that is information about the surroundings of the moving body and output information for training that is the speed of the moving body according to the surrounding information; ,
A learning device manufacturing step in which the learning device manufacturing unit manufactures a learning device by learning a plurality of sets of the training input information and the training output information,
A learning device output step in which the learning device output section outputs the learning device. The learning device manufacturing method is further processed by using a speed receiving unit, an acceleration generating unit, and a speed changing unit,
A speed receiving step in which the speed receiving unit receives the speed of the moving body corresponding to the training input information;
An acceleration generation step in which the acceleration generation section generates acceleration from the speed received in the speed reception step;
When the absolute value of the deceleration acceleration generated in the acceleration generating step exceeds a threshold value, the speed changing unit changes the speed accepted in the speed accepting step to a speed smaller than the speed. Further comprising steps,
The learning device manufacturing method according to claim 4, wherein in the training information receiving step, when the speed is changed in the speed changing step, output information for training which is the speed changed in the speed changing step is received. ..

Images