JP2018198033A - Mobile body control method, mobile body controller, and program - Google Patents

Abstract

To provide a mobile body control method and a mobile body controller capable of not only adding that there is an obstacle but also adding cases where the obstacle behaves in a different manner depending on time and automatically performing acceleration and deceleration and direction change processing of a mobile body until reaching a destination in the shortest time.SOLUTION: A mobile body control method of the present invention includes: defining a state vector indicating transition destination at the time when performing an acceleration and deceleration control in a three-dimensional direction from a three-dimensional position relative to an initial position of the mobile body and a three-dimensional speed; obtaining as to the case where all acceleration and deceleration controls for sequential state vectors for which acceleration and deceleration controls are performed at fixed intervals from a departure place to a final destination place have been controlled; selecting the sequential state vectors having the shortest arrival time among the set of obtained state vectors; and performing the sequential acceleration and deceleration controls from the departure place to the destination place according to the selected state vectors.SELECTED DRAWING: Figure 1

Description

  The present disclosure relates to a technique for controlling the operation of a moving object.

  Conventionally, as a method for indicating the shortest route to a destination, a method for optimizing a route search to the destination using learning control such as reinforcement learning has been proposed (see, for example, Non-Patent Document 1). .)

Hideaki Uchida, Hideki Fujii, Shinobu Yoshimura, Yukiyo Arai et al., “Learning Route Selection for Changes in Road Networks”, IPSJ Journal, vol. 53, no. 11, pp. 2409-2418, 2012.

  However, Non-Patent Document 1 is a technique for selecting a route, and does not propose an acceleration / deceleration method for reaching the shortest distance.

  On the other hand, in recent years, the development of sensors and the processing performance of the moving body itself have been improved, and it has become possible to acquire the location of obstacles and moving circumstances that change with time in real time, It is thought that the optimal mobile control to the destination can be determined from the starting stage.

  From this background, the present invention not only takes into account that there is an obstacle from the starting point to the destination, but can also take into account when the obstacle behaves differently depending on the time, moving to the destination in the shortest time It is an object of the present invention to provide a moving body control method and a moving body control apparatus capable of automatically performing acceleration / deceleration and direction change processing of a body.

  In order to achieve the above object, a mobile object control method according to the present invention is a three-dimensional acceleration / deceleration control with respect to a three-dimensional relative position and a three-dimensional speed from an initial position of a mobile object. Define a state vector that indicates the transition destination when executing, and obtain a series of state vectors that finally perform acceleration / deceleration control from the departure point and finally reach the destination when all acceleration / deceleration control is performed. Then, a series of state vectors having the shortest arrival time is selected from the obtained set of state vectors, and a series of acceleration / deceleration control from the starting point to the destination is executed according to the state vector.

Specifically, the moving body control method according to the present invention includes:
A state grasping procedure for obtaining the position of the moving body and the position of the obstacle with respect to the moving body;
A vector acquisition procedure for acquiring a transition destination in which the moving body moves by accelerating and decelerating the speed from a current position based on a map and the position of the obstacle, and a state vector indicating the speed at the transition destination in three dimensions; ,
Using the state vector as a key, the arrival time for the moving body to reach the transition destination, the immediately preceding state vector indicating the transition of the moving body to the current position in three dimensions, and the immediately preceding state vector to the state vector A storage procedure for storing the acceleration / deceleration of the speed in the database as a record;
After repeating the vector acquisition procedure and the storage procedure until the moving body moves to the destination set in the map, a series of times that reach the destination using dynamic programming is minimized. A control procedure for taking out the state vector from the database and executing acceleration / deceleration and direction change control from the starting point to the destination for the moving body;
I do.

Moreover, the moving body control device according to the present invention includes:
A state grasping unit for obtaining the position of the moving body and the position of the obstacle with respect to the moving body;
Based on the map and the position of the obstacle, obtain a transition vector in which the moving body moves by accelerating / decelerating the speed from the current position and a state vector indicating the speed at the transition destination in three dimensions,
Using the state vector as a key, the arrival time for the moving body to reach the transition destination, the immediately preceding state vector indicating the transition of the moving body to the current position in three dimensions, and the immediately preceding state vector to the state vector Acceleration / deceleration of speed is stored in the database as a record,
After repeatedly obtaining the state vector and storing the record in the database until the mobile body moves to the destination set on the map, the arrival time to the destination is determined using dynamic programming. A path control unit that takes out a series of the state vectors that are the shortest from the database, and executes acceleration / deceleration and direction change control from the starting point to the destination for the moving body;
Is provided.

  The mobile object control method and apparatus finds a series of state vectors having the shortest arrival time from a set of state vectors indicating a destination, and uses the series of state vectors to control the movement of the mobile object. Can reach the destination in the shortest possible time. Further, when an obstacle is recognized, the obstacle can be avoided because a state vector that contacts the obstacle is not acquired. Furthermore, it is possible to cope with a case where an obstacle moves by repeating this control periodically until the moving body arrives at the destination.

  Therefore, the present invention not only takes into account that there is an obstacle, but can also take into account when the obstacle behaves differently depending on time, and automatically performs acceleration / deceleration and direction change processing of the moving body to the destination in the shortest time. It is possible to provide a moving body control method and a moving body control apparatus that can be performed in the above manner.

  The moving body control method and apparatus according to the present invention, when there are a plurality of state vectors having the same transition destination, are used as a state vector for acquiring a shorter arrival time for the mobile body to reach the transition destination. Features. Records to be stored can be reduced, and the processing speed can be improved.

  The moving body control method and apparatus according to the present invention obtains the state vector again when the position of the moving body is different from the acceleration / deceleration and direction change control path, stores the record in the database again, and again. The acceleration / deceleration and direction change control are executed. It is also possible to deal with a case where the moving body is displaced from the route due to disturbance or the like.

  The program according to the present invention is a program for executing the mobile body control method. The mobile control method and apparatus according to the present invention can be realized by a computer and a program, and can be recorded on a recording medium or provided through a network.

  The present invention not only takes into account the presence of obstacles, but also takes into account when the obstacles behave differently depending on time, and automatically performs acceleration / deceleration and direction change processing of the moving body to the destination in the shortest time. It is possible to provide a movable body control method and a movable body control apparatus.

It is a block diagram explaining the mobile body control apparatus which concerns on this invention. It is a figure explaining the procedure which the path | route control determines so that the arrival time may become the shortest which the mobile body control apparatus which concerns on this invention performs. s (#) represents an arbitrary state vector. Further, it is assumed that c (s (q)) + f (s (q), a ′)> c (s (r)) + f (s (r), a *). It is a flowchart explaining the moving body control method which concerns on this invention.

  Embodiments of the present invention will be described with reference to the accompanying drawings. The embodiments described below are examples of the present invention, and the present invention is not limited to the following embodiments. In the present specification and drawings, the same reference numerals denote the same components.

FIG. 1 is a block diagram illustrating a mobile control apparatus according to this embodiment. This mobile control device
A state grasping unit 11 for acquiring the position of the moving body and the position of the obstacle with respect to the moving body;
Based on the map and the position of the obstacle, obtain a transition vector in which the moving body moves by accelerating / decelerating the speed from the current position and a state vector indicating the speed at the transition destination in three dimensions,
Using the state vector as a key, the arrival time for the moving body to reach the transition destination, the immediately preceding state vector indicating the transition of the moving body to the current position in three dimensions, and the immediately preceding state vector to the state vector The speed acceleration / deceleration is stored in the database 13 as a record,
After repeatedly obtaining the state vector and storing the record in the database 13 until the moving body moves to the destination set on the map, the arrival time to the destination is determined using dynamic programming. A route control unit 12 that takes out a series of the state vectors that are the shortest from the database 13 and executes acceleration / deceleration and direction change control from the starting point to the destination with respect to the moving body;
Is provided.

FIG. 3 is a flowchart for explaining the moving body control method of the present embodiment. This mobile control method
The state grasping unit 11 performs a state grasping procedure S11 for acquiring the position of the moving body and the position of the obstacle with respect to the moving body,
The route control unit 12
Based on the map and the position of the obstacle, a vector acquisition procedure S12 for acquiring a transition vector in which the moving body moves by accelerating / decelerating the speed from the current position and a state vector indicating the speed at the transition destination in three dimensions. When,
Using the state vector as a key, the arrival time for the moving body to reach the transition destination, the immediately preceding state vector indicating the transition of the moving body to the current position in three dimensions, and the immediately preceding state vector to the state vector Storage procedure S13 for storing the acceleration / deceleration of the speed in the database as a record;
After repeating the vector acquisition procedure S12 and the storage procedure S13 until the moving body moves to the destination set on the map, a series of times that the arrival time to the destination is minimized by using dynamic programming. A control procedure S14 for extracting the state vector from the database and executing acceleration / deceleration and direction change control from the starting point to the destination for the moving body;
I do.

  The mobile body autonomous control performed by the mobile body control device is achieved by using two modules of the state grasping unit 11 and the path control unit 12. Furthermore, a database 13 connected to the inside or outside of the apparatus is required. The state grasping unit 11 acquires the current position of the moving body, position information such as an obstacle, and the path control unit 12 performs the optimum moving body path control (acceleration / deceleration control) including avoidance of the obstacle to the destination. ) Using dynamic programming. In the multi-agent traffic flow simulator MATE described in Non-Patent Document 1, dynamic programming is used for route selection to the destination, but it is not applied to acceleration / deceleration control.

[Status grasping section]
The state grasping unit 11 obtains position information and the like using an existing technique. For example, the state grasping unit 11 includes a position and speed of the moving body, a sensor attached to the moving body with an obstacle ahead of the moving body, GPS information, and pre-input information (two-dimensional or three-dimensional maps, Rules).

[Route control unit]
The route control unit 12 determines the route control of the moving body from the departure point to the destination based on the information of the state grasping unit 11, and performs acceleration / deceleration control periodically (for example, every second).

ｖ_ｘ＋ａ_ｘは、ｘ軸方向の速度ｖ_ｘをａ_ｘ加速することによって得られる速度である。ｙ軸方向の速度ｖ_ｙ及びｚ軸方向の速度ｖ_ｚに関しても同様である。
Ｌ_ｘ＋ｖ_ｘ＋ａ_ｘは、Ｌ_ｘの位置についてｘ軸方向の速度ｖ_ｘをａ_ｘ加速することによって変化する、ｘ軸上の位置を示している。ｙ軸上の位置Ｌ_ｙ及びｚ軸上の位置Ｌ_ｚも同様である。
つまり、数１の状態ベクトルは、各軸方向における、出発地からの３次元上の相対位置と当該相対位置に遷移する際の速度変化の組み合わせを示している。 Here, L _x , L _y , and L _z are the three-dimensional relative positions from the initial position of the moving body, and v _x , v _y , and v _z are the velocities in the three-dimensional direction. After acquiring from the grasping unit, the state vector of the transition destination when executing the acceleration / deceleration control a _x , a _y , a _z in the three-dimensional direction is defined as follows.

v _x + a _x is a speed obtained by accelerating the speed v _x in the x-axis direction by a _x . The same applies to the y-axis velocity v _y and z-axis velocity v _z.
L _x + v _x + a _x indicates a position on the x-axis that is changed by a _x acceleration of the velocity v _x in the x-axis direction with respect to the position of L _x . position L _z on position L _y and z axis on the y-axis is the same.
That is, the state vector of Formula 1 indicates a combination of a three-dimensional relative position from the starting point and a speed change when transitioning to the relative position in each axial direction.

  The path control unit 12 stores the time to reach the state vector, acceleration control, and the state vector before acceleration control (previous state vector) in the database 13 using the state vector as a key. Since the route control unit 12 determines the acceleration / deceleration guideline that can minimize the arrival time to the destination with each state vector, the route control unit 12 uses the dynamic programming method from the state vector of the departure point to determine the purpose. Achieve.

The state vector of Equation 1 is expressed by a simplified notation s, and the arrival time taken to transition to the state vector s is defined as c (s). The state vector s from the acceleration _{a = (a x, a y} , a z) ' when a transition is made to, c (s' s by) becomes c (s) + f (s , a). Here, f (s, a) is a time taken for acceleration / deceleration from s to a. a = 0 indicates (0, 0, 0).

The route control unit 12 stores these pieces of information in the database 13 in a format such as Equation 2.

When there are a plurality of state vectors having the same transition destination, the route control unit 12 sets the state vector that acquires the shorter arrival time for the moving body to reach the transition destination.
When the state transition is repeated, there is a possibility that the state vector s ′ may also transition from a state vector other than the state vector s. For example, consider a transition from state vector s * to s ′ by acceleration / deceleration of a *. At this time, when the record of the state vector s ′ satisfies c (s ′)> c (s *) + f (s *, a *), c (s ′) is changed to c (s *) + f (s *, a *) And s and a are replaced with s * and a *.

  FIG. 2 is a diagram for explaining records and state vectors stored in the database 13 when the route control unit 12 performs all of the above procedures from the starting point and recursively executes them. Finally, the route control unit 12 extracts a state vector set indicating the destination from the record of the database 13, and finds the key and value of the state vector having the shortest arrival time (c (s) takes the smallest value). Acceleration / deceleration control is executed based on a series of state vectors from the departure point to the destination. Thereby, the route control unit 12 can control the moving body so that the arrival time from the departure place to the destination is the shortest.

  For avoiding an obstacle or the like, the path control unit 12 determines whether or not the state vector s touches the obstacle, and prevents the transition to the state vector s that may touch the obstacle. Note that c (s) means the time from the departure point to the transition to s. Therefore, by taking this value into consideration, there is an obstacle whose appearance pattern differs depending on time in a specific state vector s. The obstacle can be avoided.

  When the position of the moving body is different from the acceleration / deceleration and direction change control route, the route control unit 12 acquires the state vector again, stores the record in the database again, and executes acceleration / deceleration and direction change control again. . If the moving object deviates from the intended route due to a collision with another moving object or an unexpected event during the route control, the route control unit 12 calculates the route again.

The moving body control method and apparatus according to the present invention can control a moving body that can be controlled by a computer from the starting point to the destination in the shortest time, and can autonomously operate a small airplane or car that can be operated from the outside. It can be applied to control.
In addition, it is possible to combine other collision avoidance methods and accident response methods with the moving body control method and apparatus according to the present invention.

11: State grasping unit 12: Path control unit 13: Database

Claims (7)

A state grasping procedure for obtaining the position of the moving body and the position of the obstacle with respect to the moving body;
A vector acquisition procedure for acquiring a transition destination in which the moving body moves by accelerating and decelerating the speed from a current position based on a map and the position of the obstacle, and a state vector indicating the speed at the transition destination in three dimensions; ,
Using the state vector as a key, the arrival time for the moving body to reach the transition destination, the immediately preceding state vector indicating the transition of the moving body to the current position in three dimensions, and the immediately preceding state vector to the state vector A storage procedure for storing the acceleration / deceleration of the speed in the database as a record;
After repeating the vector acquisition procedure and the storage procedure until the moving body moves to the destination set in the map, a series of times that reach the destination using dynamic programming is minimized. A control procedure for taking out the state vector from the database and executing acceleration / deceleration and direction change control from the starting point to the destination for the moving body;
A moving body control method.   2. The vector acquisition procedure according to claim 1, wherein when there are a plurality of state vectors with the same transition destination, a state vector for acquiring a shorter arrival time for the moving body to reach the transition destination is used. The moving body control method described in 1.   3. The moving body control method according to claim 1, wherein when the position of the moving body is different from the acceleration / deceleration and direction change control path during the control procedure, the moving body control method according to claim 1 is executed again. A state grasping unit for obtaining the position of the moving body and the position of the obstacle with respect to the moving body;
Based on the map and the position of the obstacle, obtain a transition vector in which the moving body moves by accelerating / decelerating the speed from the current position and a state vector indicating the speed at the transition destination in three dimensions,
Using the state vector as a key, the arrival time for the moving body to reach the transition destination, the immediately preceding state vector indicating the transition of the moving body to the current position in three dimensions, and the immediately preceding state vector to the state vector Acceleration / deceleration of speed is stored in the database as a record,
After repeatedly obtaining the state vector and storing the record in the database until the mobile body moves to the destination set on the map, the arrival time to the destination is determined using dynamic programming. A path control unit that takes out a series of the state vectors that are the shortest from the database, and executes acceleration / deceleration and direction change control from the starting point to the destination for the moving body;
A moving body control apparatus comprising:   5. The path control unit according to claim 4, wherein when there are a plurality of state vectors having the same transition destination, the path control unit obtains a state vector that acquires a shorter arrival time for the moving body to reach the transition destination. The moving body control device described in 1.   When the position of the moving body is different from the acceleration / deceleration and direction change control route, the path control unit obtains the state vector again, stores the record in the database again, and renews the acceleration / deceleration and direction change. Control is performed, The mobile body control apparatus of Claim 4 or 5 characterized by the above-mentioned.   The program for performing the moving body control method in any one of Claim 1 to 3.

Images