JP2928658B2 - Optimal route search device for mobile robots - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE SakuSo this invention probe the optimal path of a suitable mobile robot applied to a mobile robot, such as self-reliance unmanned vehicle
About the location.

[0002]

2. Description of the Related Art In a mobile robot, various studies have been made on the problem of what route should be optimally taken from a certain point to a destination, and various methods have been proposed. For example, data corresponding to a map is stored in a memory built in the mobile robot, and an optimal route is determined based on the data. That is, may be stored in advance in the memory of the linear distance and node connections between specific points on the map (node), then the node V _j to go from node V _i which is currently located in the following manner decide.

(1) A route from a start node S to a destination node G is searched by a known method such as a vertical search or a horizontal search. Then, the next candidate node V node V _j to be next passes from the connected to the intermediate node V _i node
Select _jx . (2) from the next candidate node V _jx, choose the closest node V _j current to node V _i. (3) This operation is repeated to determine a route from the departure node S to the destination node G. In this case, the route has the minimum total route length and is considered to be the optimal route.

[0004] However, in the above-described conventional route search method, a true optimal route may not be obtained. This is because, when the next node V _j is selected from the next candidate nodes V _jx , the node closest to the current node V _i is selected, so that a case may occur in which the path goes away from the target node G. Because. Further, in the vertical search, there is a case where the target node G is not reached at any time. On the other hand, in the horizontal search, there is a problem that it takes a long time to search for the target node G. this is,
This is because if the distance between the nodes is equal, the number of branches to be searched increases geometrically.

Therefore, the present applicant has previously described the intermediate node V
_In selecting the next node V _j from _i , the following evaluation functions H ₁ (V _i , V _jx ), H ₂ (V _i , V _jx ) and H ₃
Three methods for performing a route search using (V _i , V _jx ) have been proposed. (A) The first method _{H 1 (V i, V jx} ) = W a1 · A 1 (V jx, G) + W b1 · B 1 (V i, V jx) ...... (1) where, A ₁ ( _{V jx, G) = D (} V jx, G) / L 1 ...... (2) B 1 (V i, V jx) = D (V i, V jx) / L 1 ...... (3) where, D (V _jx, G) is the distance of the distance between the next candidate node V _jx and destination node _{G, D (V i, V} jx) of the current node V _i and the next candidate node V _jx, L ₁ is the node Is the maximum distance between Therefore, the variables A ₁ (V _jx , G) and B ₁ (V
_i, V _jx) has been normalized by the distance L _1, 0 to
Taking a value of 1, the smaller the value, the higher the evaluation. 0 is the best and 1 is the worst. W _a1 and W _b1 are variables A ₁ (V _jx ,
G) and B ₁ (V _i , V _jx ). For details of the technique described above, refer to Japanese Patent Application Laid-Open No. Sho 62-32516 which discloses a method for searching for an optimum route for a mobile robot, which has been previously proposed by the present applicant.

(B) Second method H ₂ (V _i , V _jx ) = W _a2 · A ₂ (V _jx , G) + W _b2 · B ₂ (V _i , V _jx ) (4) A ₂ (V _jx , G) = D (V _jx , G) / L ₂ (5) B ₂ (V _i , V _jx ) = D (V _i , V _jx ) / πL ₂ (6) Here, L ₂ is a map (that is, a moving area of the mobile robot).
Is the distance of the diagonal line of the rectangle circumscribing. Therefore, the variable A ₂ (V _jx , G) is normalized by the distance L ₂ and the variable B 2
₂ (V _i , V _jx ) is normalized by a value obtained by multiplying the distance L ₂ by the pi, and takes a value of 0 to 1. The smaller the value, the higher the evaluation. 0 is the best and 1 is the worst. It is. W _a2 and W _b2 are variables A ₂ (V _jx , G) and B ₂ (V _i ,
V _jx ). For details of the above-mentioned technology, refer to Japanese Patent Application Laid-Open No. Sho 62-32517, which is a method of searching for an optimum route for a mobile robot previously proposed by the present applicant.
Gazette).

(C) Third method H ₃ (V _i , V _jx ) = W _a3 · A ₃ (V _i , V _jx ) + W _b3 · B ₃ (θ) (7) where A ₃ (V _i, V _jx) is a variable according to the distance information between nodes V _i and node V _jx, are set to evaluate as follows: candidate node V _jx approaching the destination node G is improved. Also, B ₃ (theta) is the traveling direction (i.e., the direction of a line connecting the current node V _i and the previous node) at the node V _i and, formed between a line connecting the node V _i and node V _j This is a variable indicating a loss determined with respect to the angle θ. This angle θ
Is obtained by calculation from the coordinates of each node, and the left and right turns are set to 90 ° without a sign according to the rotation direction. The smaller the angle θ, that is, the closer to a straight line, the better the evaluation of the variable B ₃ (θ). W _a3 and W _b3 are variables A
₃ (V _i , V _jx ) and B ₃ (θ) are weighting coefficients. For details of the above-described technique, refer to the publication (Japanese Unexamined Patent Publication No. 62-32518) of a method for searching for an optimal route of a mobile robot previously proposed by the present applicant.

[0008]

The above-mentioned conventional methods for searching for an optimum route for a mobile robot all search for the shortest distance route as the optimum route. However, the optimal route in the mobile robot is the shortest time route. Further, the conventional optimal route search method has the following problems in the following cases.

Now, suppose that the mobile robot at node 7 on the map shown in FIG. 6 searches node 7 → node 8 → node 3 → node 4 → node 5 as a route when moving to node 5. In this case, since the node 7 → node 8 and the node 8 → node 3 both travel straight along the path, the node 8 travels on a curve. Therefore, node 8
When evaluating the node 3 from the above, it is appropriate to evaluate the direction conversion as an angle change by the conventional optimal route search method.

[0010] However, when the mobile robot at the node 7 on the map shown in FIG. 6 moves to the node 5, the route of the node 7 → node 8 → node 9 → node 4 → node 5
Is searched, the path between the node 8 and the node 9 is straight, and the path between the node 9 and the node 4 is traversing (forward and backward). At this time, at the node 9, although it differs depending on the traveling mechanism of the mobile robot, a large time loss is caused as compared with traveling on a curve. The same applies to the node 4.

The problem here is that in the conventional optimal route search method, a case where the vehicle travels in a curve at the node 8 and a case where the vehicle temporarily stops at the node 9 and then moves to the traversing posture and then proceeds to the node 4 actually occur. Despite the differences in travel times
In terms of angle change, the evaluation is the same, and an evaluation different from reality is performed. The present invention has been made under such a background, and an optimal path search for a mobile robot capable of searching for a path in the shortest time and evaluating a change in a traveling direction as a time corresponding to an actual operation.
It is an object to provide a cable device .

[0012]

SUMMARY OF THE INVENTION According to the present invention, a rotation angle
Curve running to change and run, and stop once and wheels
First storage means for storing map data indicating a moving area of a mobile robot capable of traversing in which the angle of the mobile robot is changed to a right angle; second storage means for storing travel speed data of the mobile robot; a unit angle time data required to rotate the angle, and direction change time data required to perform the transverse travel, proceeding candidate sites and current location
Travel between the current location and the candidate location
A third storage means for storing the velocity data of Rutoki, based on the map data, a distance calculation means for obtaining a linear distance data indicating the respective linear distance to the destination from the plurality of candidate locations, the straight line First dividing means for dividing the distance data by the traveling speed data and obtaining linear distance traveling time data as a result; and dividing the distance data into the traveling data.
Data, and the resulting travel time data
Dividing means, when traveling on a curve, the unit angle time data
Rotation angle time data obtained based on the
Based on the distance travel time data and the travel time data,
To obtain the evaluation value of the route search,
Direction change time data, the straight-line distance travel time data,
Evaluation value calculation means for obtaining an evaluation value of a route search based on the travel time data; and selection means for selecting a route having the smallest evaluation value among the routes to the plurality of candidate locations. It is characterized by the following.

[0013]

According to the above method, the shortest time route which is the original optimum route as the mobile robot can be searched.

[0014]

An embodiment of the present invention will be described below with reference to the drawings. Figure 1 is a block diagram showing a configuration of a mobile robot according to the optimum route exploration search device of a mobile robot according to an embodiment of the present invention. In this figure, 1 is traveling device of a mobile robot, 2 traveling device 1 A CPU (Central Processing Unit) 3 for controlling the traveling is a memory in which map information relating to the map shown in FIG. 2 is stored.

The map information is composed of node numbers, connection information of each node, and information between nodes (scenes), and is basically created, added and deleted by the control station. In some cases, the scene is created by running the mobile robot itself. The created map information is transferred to each mobile robot using the IC card as a recording medium, and the mobile robot stores the transferred map information in the memory 3 backed up by a battery. The data structure of the map information must be the same between the control station and the mobile robot, and the contents must be the same.

Here, the data structure of the map information is shown in FIGS. FIG. 3 is a node entry table in which the node numbers of registered nodes are stored in the order of registration.
In an area that is not registered, 0 is stored as a node number. This node number can be arbitrarily determined by the user.

FIG. 4 shows a network table, which stores network information such as world coordinates of nodes and connection information between nodes. The memory capacity for one node is equal to the number of nodes that can be registered with a fixed length. In FIG. 4, the node X coordinate and the Y coordinate are the world coordinates (mm) of the node, and the work point X coordinate and the work point Y coordinate are the world coordinates (mm) of the work point.

The connection node numbers 0 to 3 are the numbers of the nodes connected to this node (in this case, the nth registered node), and the node numbers connected from the area of the node number 0 are When the number of nodes is three or less, 0 is stored in the remaining area. The scene offsets 0 to 3 indicate the byte number of the scene between the nodes corresponding to the connected node numbers 0 to 3 in a scene table described later.

FIG. 5 shows a scene table in which the states of the walls between the nodes and the running conditions are stored. In FIG.
The linear distance between nodes is a distance (mm) actually traveled by the mobile robot obtained from the coordinates, and the measured distance between nodes (mm) has a value only when the node is not a straight line, and is 0 when it is a straight line. In the one-way instruction, 0 indicates no one-way traffic, 1 indicates an end point from a start point, and 2 indicates a start point from an end point. In the running posture, 0 indicates forward / backward movement and 1 indicates traversing. The forward running offset and the backward running offset are distances (mm) from the reference line, where + indicates left with respect to the traveling direction and-indicates right with respect to the traveling direction. The speed limit is 0 and the maximum speed (m / h
our). The effective direction of the beam sensor is 0 for both directions, 1 for the forward direction, 2 for the reverse direction, and 3 for both directions.

Here, the node serving as a reference for the route search is V
_i, then the node V _j should proceed, the node V _i evaluation for selecting the node V _j from the function H (V _i, V _jx) is defined by the following equation. H (V _i , V _jx ) = T _gx + T _lx + T _dx + T _ox (8) In this equation, V _jx is the candidate node to be selected next, and T _gx is the candidate node V _jx and the destination (goal). time to linear distance running, T _lx node time to travel from V _i to the candidate node V _jx, T _dx node V _i attitude change time for traveling from the position to the candidate node V _jx of, T _ox other travel It is time to limit.

[0021] Now, consider the case to evaluate the candidate node V _jx at node V _i. Each parameter of the evaluation can be calculated as shown below. T _gx = (node V _i and the linear distance goal) / (the maximum speed of the mobile robot) (9) node V _i and the linear distance goal is calculated from both coordinates. The maximum speed of the mobile robot is stored in the memory 3. This T _gx indicates how the candidate node V _jx is approaching the goal.

[0022] T _lx = (node V _i and the distance between the candidate node V _jx) / (node speed when traveling from V _i to the candidate node V _jx) ...... (10) node V _i and the candidate node V speed when traveling from the distance and the node V _i between _jx to the candidate node V _jx both are stored in the scene table. When traveling on a curve, T _dx = time required for rotating by a unit angle × angle (1)
1) When the running direction changes, T _dx = the time required to change the running direction ... (12) T _ox is used to reflect a special situation. For example, if there is an automatic door in the middle to go to a certain node and it is necessary to wait for the automatic door to open, the time until the automatic door opens is set to _Tox .

As described above, the evaluation value is calculated for each node, and the node having the smallest value is selected to search for the optimum route of the mobile robot. As described above, according to the above-described embodiment, the shortest time route that is the original optimal route as the mobile robot can be searched. Also, (（H (V _i , V _jx ) -ΣT _gx ) can be calculated as the running time from the start point to the goal. Further, the change in the traveling direction can be evaluated as a time during which an actual operation is performed. In addition, the time loss due to special situations on the passage can be taken into account.

[0024]

As described above, according to the present invention, there is an effect that a shortest time route which is an original optimum route as a mobile robot can be searched. Further, there is an effect that a change in the traveling direction can be evaluated as a time corresponding to an actual operation.

[Brief description of the drawings]

1 is a block diagram showing the applied mobile robot constituting the optimum route exploration search device of a mobile robot according to an embodiment of the present invention.

FIG. 2 is a conceptual diagram showing an example of a map according to an embodiment of the present invention.

FIG. 3 is a diagram showing a configuration of one node entry table of map information.

FIG. 4 is a diagram showing a configuration of one network table of map information.

FIG. 5 is a diagram showing a configuration of one scene table of map information.

FIG. 6 is a conceptual diagram of a map for explaining a problem of a conventional method for searching for an optimum route for a mobile robot.

[Explanation of symbols]

 1 traveling device 2 CPU 3 memory

Claims (1)

(57) [Claims] 1. A curve running in which the rotation angle is changed and the vehicle runs.
Line, and the side that stops once and changes the angle of the wheel to right angle
First storage means for storing map data indicating a moving area of the mobile robot capable of traveling in line, second storage means for storing travel speed data of the mobile robot, and at least a unit for rotating the unit angle. a unit angle time data such a direction change time data required to perform the transverse travel, proceeding distance data between the candidate site and are here
And speed data when traveling between the current location and the candidate location.
A third storage means for storing a over data, based on the map data, a distance calculation means for obtaining a linear distance data indicating the respective linear distance to the destination from the plurality of candidate locations, the linear distance data Dividing by the traveling speed data, the first division means to obtain the linear distance traveling time data as a result, the distance data is divided by the traveling data, as a result
Second division means for obtaining travel time data, and when traveling on a curve, based on the unit angle time data
Obtained rotation angle time data and the linear distance travel time
Based on the data and the travel time data.
An evaluation value is obtained, and when traveling in a transverse direction, the direction change time data and the straight line
Based on the distance travel time data and the travel time data,
Come, the evaluation value calculating means for calculating an evaluation value of the route search, the smallest value of the path to the plurality of the candidate site
Selecting means for selecting a route having an evaluation value .