KR100994075B1 - Method for planning an optimal path in a biped robot - Google Patents

Abstract

The present invention relates to a method for designing an optimal path of a pedestrian robot, and more particularly, to an optimal path planning method for efficient driving by avoiding obstacles in consideration of the characteristics of the pedestrian robot. An optimal path planning method for a walking robot according to an embodiment of the present invention includes the steps of initializing, calculating a path cost, creating a candidate list, and calculating an angle cost, a heuristic cost, and a total cost of each candidate node. And creating an open list, creating a closed list, and determining whether the next node is a target node.

Walking robot, A * algorithm, optimal path, minimization of rotation, radius of rotation, path cost

Description

Optimal path planning method for pedestrian robots {METHOD FOR PLANNING AN OPTIMAL PATH IN A BIPED ROBOT}

The present invention relates to a method for designing an optimal path of a pedestrian robot, and more particularly, to an optimal path planning method for efficient driving by avoiding obstacles in consideration of the characteristics of the pedestrian robot.

In general, a method for moving a walking robot to a target point includes acquiring the position information of the robot, a self position estimation step of finding the current position of the robot based on the acquired position information, and expressing the entire driving environment. It is divided into the map making stage and the route planning stage in which the optimal route to the target point is searched using the environmental map and the self location information. Typically, the Potential Field Method, Vector Field Histogram (VFH), and A * (referred to as "Ast") algorithms are used in the path planning step.

The potential field method is a method of expressing the search space as a sum of stimulus affecting the robot. In this method, the stimulus at the target position is expressed in attractive force, and the stimulus at the obstacle position is expressed in repulsive force. The algorithm used in the potential field method is simple, so it is possible to quickly search for the optimum path, and there is an advantage that it can be used without limitation of the type of sensor. However, when two obstacles are adjacent to each other, since two repulsive forces generated in the obstacle can be recognized as one repulsive force, there is a problem that it is impossible to pass between the two obstacles.

In order to solve the above problems, J. Borenstein and Y. Koren, "THE VECTOR FIELD HISTOGRAM-FAST OBSTACLE AVOIDANCE FOR MOBILE ROBOTS", IEEE Journal of Robotics and Automation V7 , No. 3, June 1991, pp. 278-288 ) proposed the vector field method. The vector field method uses a histogram grid of two-dimensional rectangular coordinates. The vector field method uses a two-step data reduction process to calculate the control commands of the robot. In the first step, the histogram grid is reduced to a one-dimensional polar histogram constructed around the robot's momentary position. A value representing the polar obstacle density (POD) in that direction is included in each sector of the polar histogram. In a second step, the algorithm selects the most suitable sector with the lowest movie object density among polar histogram sectors. However, there is a disadvantage that it is difficult to use a wide range of maps.

The A * algorithm refers to an algorithm for searching for an optimal path from a starting point to a target point by using information about a known environment using a grid map. This algorithm is a method of searching for an optimal path having a minimum cost by adding the shortest distance information from the starting point to the current position of the robot and heuristic information from the current position of the robot to the target point.

These methods have several problems. However, the biggest problem of these methods is that it does not take into account the rotatable position of the walking robot, the radius of rotation of the walking robot, the size of the obstacles and the direction of the target. For example, the walking robot may not be able to rotate in place, and may take a lot of time to rotate even when it is possible to rotate in place. In addition, in order to perform the in-situ rotation of the walking robot, complicated configuration and control is required, resulting in an increase in cost and an increase in weight. In addition, the direction of the target may be considered important when the walking robot grasps and moves an object in a specific direction at the target point. Therefore, in consideration of these various characteristics, there is a need to develop an optimal path planning method for pedestrian robots that can avoid obstacles.

The present invention was devised to solve the above problems, and an object of the present invention is to provide an optimal path planning method for a walking robot including a step of calculating the angular cost to minimize the rotation of the walking robot.

Another object of the present invention is to provide an optimal path planning method of a walking robot considering the rotation radius of the walking robot.

Still another object of the present invention is to provide an optimal path planning method of a walking robot in consideration of the direction of the target and the direction of the walking robot.

An optimal path planning method of a walking robot according to an embodiment of the present invention comprises the steps of: initializing and receiving information about a departure node, a target node, a previous node, a current node, and an obstacle node; Calculating a path cost, inputting a candidate node which is not an obstacle node among a plurality of candidate nodes in the vicinity of the current node in the candidate list, calculating an angular cost of each candidate node entered in the candidate list; Calculating a heuristic cost of each candidate node input to the candidate list, calculating a total cost of each candidate node input to the candidate list according to a path cost, an angular cost, and a heuristic cost, and storing the result in each candidate node; Entering a candidate node in an open list, updating the total cost of the node already entered in the open list with the total cost for the current node, Designating a candidate node having a minimum total cost among the nodes in the open list as the next node and inputting the candidate node to the closed list, and determining whether the next node is the target node, and in the step of determining the target node, the next node is the target node. Then returns the entire path entered in the closed list by back tracking, returns to the path cost calculation step if the next node is not the target node, and the next node in the next node input step to the current node in the path cost calculation step. Are treated.

Preferably, the embodiment of the present invention further includes the step of modifying the path of the inflection point in consideration of the rotatable position of the walking robot.

In one embodiment of the present invention, the angle cost is calculated by the angle between the previous node and the next node around the current node, the heuristic cost is the distance cost from the current node to each candidate node and from the candidate node to the target node Calculated by the sum of costs. In addition, the total cost is calculated by the following formula:

f (n) = g (n) + C _a (n) * α + C _h (n) * (1-α),

Where f (n) is the total cost,

g (n) is the path cost,

C _a (n) is the angle cost,

C _h (n) is the heuristic cost,

α is the weight of the angle cost, and α is a real number greater than 0 and less than 1.

According to another embodiment of the present invention, an optimal path planning method of a walking robot receives and initializes a starting node, a target node, a previous node, a current node, and an obstacle node, and considers the direction of the target node and the direction of the walking robot. Setting a temporary target node having a predetermined rotation radius with the target node, calculating a path cost from the starting node to the current node, and a candidate that is not an obstacle node among a plurality of candidate nodes in the vicinity of the current node. Inputting a node to the candidate list, calculating an angular cost of each candidate node entered in the candidate list, calculating a heuristic cost of each candidate node entered in the candidate list, path costs, angular costs, and Calculating the total cost of each candidate node entered in the candidate list according to the heuristic cost and storing the candidate cost in each candidate node; Entering the list, updating the total cost of the node already entered in the open list to the total cost for the current node, and inputting the candidate node with the lowest total cost among the nodes in the open list as the next node in the closed list. And determining whether the next node is a temporary target node. In the step of determining a temporary target node, if the next node is a temporary target node, the entire path input in the closed list including the path from the temporary target node to the target node is included. Returned by back tracking, if the next node is not a temporary target node, the process returns to the path cost calculation step, and the next node of the next node input step is treated as the current node of the path cost calculation step.

Preferably, another embodiment of the present invention further comprises the step of modifying the path of the inflection point in consideration of the rotatable in place of the walking robot.

In another embodiment of the present invention, the angular cost is calculated by the angle between the previous node and the next node about the current node, and the heuristic cost is the distance cost from the current node to each candidate node and from the candidate node to the target node. Calculated by the sum of costs. In addition, the total cost is calculated by the following formula:

f (n) = g (n) + C _a (n) * α + C _h (n) * (1-α),

Where f (n) is the total cost,

g (n) is the path cost,

C _a (n) is the angle cost,

C _h (n) is the heuristic cost,

α is the weight of the angle cost, and α is a real number greater than 0 and less than 1.

According to the first embodiment of the present invention, by minimizing the rotation on the path by the optimal path planning method of the walking robot including the step of calculating the angular cost, the walking robot, which cannot be rotated in place, can avoid the obstacle. You can plan. In addition, by using a coordinate system represented by a grid for the entire terrain, it is possible to plan the optimal route in consideration of the size of the walking robot and obstacles. According to the optimal path planning method of the walking robot according to the second embodiment of the present invention, by setting an appropriate temporary target point for the target, it is possible to plan the optimal path in consideration of the specific direction of the target. As a result, when the walking robot can be rotated in place, the moving time can be shortened. In addition, there is no need to consider the configuration related to the in-situ rotation in the manufacture of the walking robot.

Hereinafter, with reference to the accompanying drawings will be described in detail the optimal path planning method of the pedestrian bots according to the first and second embodiments of the present invention.

1 is a schematic view showing a coordinate system and components used in the present invention.

As shown in FIG. 1, the walking robot 110 includes a heading part 111 indicating a moving direction of the walking robot 110, and the target 130 includes a gripping part 131 indicating a specific direction of the target. Equipped. The x-axis increases to the right and the y-axis increases to the bottom. In the present invention, a map composed of a square grid of an appropriate size is used based on information on obstacles and targets. However, the present invention is not limited to this, and may be composed of any polygonal grid.

2 is a view showing a direction meter used in the present invention, Figure 3 is a view showing the relationship between each node used to calculate the angular cost of the walking robot. Hereinafter, the "current node" is a node for the current position of the walking robot, the "previous node" is the node immediately before the current node, and the "next node" is the node of the current node that the walking robot will walk from the current node later. Immediately after this node. However, if the current position of the walking robot is the same as the starting position, the current node may be the same as the starting node.

As shown in FIG. 2, the upper hemisphere is 0 ° to -180 ° counterclockwise with respect to the x-axis, and the lower hemisphere is 0 ° to + 180 ° clockwise. When calculating the angle, the direction indicator is applied about the current node. As shown in Fig. 3, the angle θ _n-1 of the previous node with respect to the current node is obtained as follows:

,

,

Where N _{n -1} (x _{n -1} , y _{n -1} ) is the coordinates of the previous node,

N _n (x _n , y _n ) is the coordinate of the current node.

The angle of the candidate node θ _{n + 1} with respect to the current node is obtained as follows:

,

,

Where N _n (x _n , y _n ) is the coordinates of the current node,

N _{n +1} (x _{n +1} , y _{n +1} ) is the coordinate of the candidate node.

The angle [theta] between the previous node and the candidate node around the current node is obtained as follows:

.

.

Detailed description of the angular cost will be described later.

4 is a flowchart illustrating an optimal path planning method of a walking robot according to a first embodiment of the present invention.

As shown in FIG. 4, the optimal path planning method of the walking robot according to the first embodiment of the present invention includes receiving and initializing information about a starting node, a target node, a previous node, a current node, and an obstacle node ( S101, calculating a path cost from the starting node to the current node (S102), inputting a candidate node which is not an obstacle node among a plurality of candidate nodes in the vicinity of the current node (S103); Calculating an angular cost of each candidate node input to the candidate list (S104), calculating a heuristic cost of each candidate node input to the candidate list (S105), and a path cost, an angular cost, and a heuristic cost. In operation S106, the total cost of each candidate node input to the candidate list is calculated and stored in each candidate node, and the candidate node is entered into an open list, and each of the candidates is already entered in the open list. Updating the node with the total cost for the current node (S107), selecting a candidate node having the minimum total cost among the nodes in the open list as the next node, removing the list from the open list, and entering the closed list (S108); And determining whether the next node is the target node (S109), and if the next node is the target node, returns the entire path entered in the closed list by back tracking, and if the next node is not the target node, the path cost. Returning to the calculation step S102, the next node of the closed list input step S108 is treated as the current node of the path cost calculation step S102.

The initialization step S101 is a step of receiving and initializing information on a departure node, a target node, a previous node, a current node, and an obstacle node. The "starting node" is a node at the point where the walking robot first starts walking, the "target node" is a node at the final target point of the walking robot, and the "obstacle node" is a node where static obstacles are located. The obstacle node may be composed of a plurality of grids in consideration of the size of the obstacle. The closed list in this step S101 is where the optimal path selected in consideration of the total cost is stored, and the current node is stored as a part of the optimal path.

The path cost calculation step S102 is a step of calculating the path cost from the starting node to the current node. Route costs can be calculated either by simple distance costs or by considering distances and angles. Hereinafter, "cost" refers to a value given to a predetermined value with respect to a required time or a predetermined distance and converted into a numerical value thereof. For example, "angle cost" is a conversion value for the distance between each node, and "distance cost" is a conversion value with respect to the time taken for the walking robot to rotate by a predetermined angle (θ).

The candidate node input step S103 is a step of inputting a plurality of candidate nodes excluding obstacle nodes among candidate nodes in the vicinity of the current node to the candidate list. By default, candidate nodes are eight grids surrounding the current node. However, candidate nodes that are not obstacle nodes are input to the candidate list. This is to avoid the step of calculating various costs for the obstacle node.

The angle cost calculation step S104 is a step of calculating and storing the angle cost of each candidate node input to the candidate list. The angular cost is calculated based on the angle θ formed between the previous node and the candidate node with respect to the current node. Angle (theta) is calculated | required by Formula (3). If the angle θ is 180 °, the angle cost is minimum. If the angle θ is smaller than 180 °, the angle cost increases. If the angle θ is larger than 180 °, the angle cost increases.

The heuristic cost calculation step S105 is a step of calculating the distance cost from the current node to the target node. The heuristic cost is divided into the distance cost from the current node to the candidate node and the distance cost from the candidate node to the target node. These distance costs can be calculated by the linear distance between each node. Heuristic costs can be expressed as:

C _h (n) = C _h1  + C _h2 ,

Where C _h (n) is the heuristic cost,

C _h1 is the distance cost from the current node to the candidate node,

C _h2 is the distance cost from the candidate node to the target node.

The total cost calculation step S107 is a step of calculating the total cost of each candidate node according to the path cost, the angular cost, and the heuristic cost. The total cost can be calculated by the sum of the costs. On the other hand, the total cost can be expressed as the sum of the path costs from the originating node to the current node and the predicted costs from the current node to the target node:

f (n) = g (n) + h (n),

Where f (n) is the total cost,

g (n) is the path cost from the originating node to the current node,

h (n) is the predicted cost from the current node to the target node.

In this case, the forecast cost can be expressed as the sum of the angular cost and the heuristic cost:

h (n) = C _a (n) + C _h (n),

Where C _a (n) is the angle cost

C _h (n) is the heuristic cost.

The predicted cost can be expressed as follows by weighting the angular cost so that the walking robot has the minimum rotation:

h (n) = C _a (n) * α + C _h (n) * (1-α)

Where α is a weight and α is a real number greater than 0 and less than 1.

α is a weighting coefficient of the angular cost and represents a specific gravity for rotation minimization. The weight may be appropriately selected in consideration of the rotatable position of the walking robot. That is, if the weight is 0.5 or more, rotation minimization is taken into account. If the weight is 0.5 or less, the shortest path is considered. The total cost calculated for each candidate node is stored in each candidate node entered in the candidate list.

The open list input and update step (S107) is a step of inputting candidate nodes into the open list, and updating the total cost of each node that has already been entered in the open list to the total cost for the current node. In this step S107, the candidate node in which the total cost is stored is deleted from the candidate list and entered in the open list. There may be candidate nodes already entered in the open list. However, since these candidate nodes store the total cost for the previous node, they are updated with the total cost for the current node. Thereafter, the candidate nodes in the open list are sorted in order of total cost.

The closed list input step (S108) is a step of designating the next node and storing it in the closed list. In this step S108, the candidate node having the lowest total cost is designated as the next node. The candidate node designated as the next node is deleted from the open list, and the remaining candidate nodes remain in the open list. After that, the next node is entered in a closed list and becomes part of the full path.

The next node determination step (S109) is a step of determining whether the next node is a target node. In this step S109, if the next node is the target node, the entire path composed of the nodes input to the closed list is returned by back tracking. In this step (S109), if the next node is not the target node, the process returns to step S102 and the above-described steps are repeated until the next node matches the target node. The next node of step S108 is treated as the current node below step S102.

Figure 5 is a schematic diagram showing the optimal path of the walking robot without considering the rotation of the walking robot and the size of the walking robot, Figure 6 is a schematic diagram showing the optimal path of the walking robot considering the rotation of the walking robot and the size of the walking robot. to be. Hereinafter, the inflection point refers to a rotation point at which the change of direction occurs in the path.

Since the optimal path shown in FIG. 5 is adjacent to the obstacle 120, the walking robot 110 may collide with the obstacle 120, and the rotational potential of the walking robot 110 is not considered. In order to avoid these collisions, the grid size of the map should be set in consideration of the size of the walking robot 110. And, the obstacle node should be determined to accommodate the size of the entire obstacle (120). In this case, the optimal path includes three inflection points.

The optimal path shown in FIG. 6 considers the size of the walking robot 110 and the size of the obstacle 120, and considers the minimization of rotation according to the first embodiment of the present invention. In this case, the optimal path includes two inflection points. The optimal path of the walking robot 210 of FIG. 6 has reduced the number of inflection points compared to the optimum path of FIG. 5 without minimizing the rotation, the size of the walking robot, and the size of obstacles.

FIG. 7 is a schematic view showing an optimal path of a walking robot that can rotate in place, and FIG. 8 is a schematic view of an optimal path of a walking robot that cannot rotate in place.

In the case of the walking robot 110 capable of rotating in place as shown in FIG. 7, the optimal path planning method according to the first embodiment of the present invention can shorten the time required for rotation by minimizing rotation. In the case of the walking robot 110 which cannot rotate in place as shown in FIG. 8, the walking robot 110 has a predetermined rotation radius R. As shown in FIG. The walking robot 110 is rotated after going straight by the rotation radius (R) to enter the next path. Therefore, the inflection point of the entire path should be corrected in consideration of the rotation radius (R).

9 is a flowchart illustrating an optimal path planning method of a walking robot according to a second embodiment of the present invention.

The optimal path planning method of the walking robot according to the second embodiment of the present invention includes a step (S201) of receiving and initializing a starting node, a target node, a previous node, a current node, and an obstacle node, the direction of the target node, and Setting a temporary target node having a predetermined rotation radius with the target node in consideration of the directionality of the walking robot (S202), calculating a path cost from the starting node to the current node (S203), and surrounding the current node Inputting a candidate node which is not an obstacle node among a plurality of candidate nodes in a candidate list (S204), calculating an angle cost of each candidate node input in the candidate list (S205), and inputting the candidate list Computing the heuristic cost of each candidate node (S206), and calculating the total cost of each candidate node entered in the candidate list according to the path cost, the angular cost, and the heuristic cost to store in each candidate node. Step S207, inputting the candidate node to the open list, updating the total cost of the node already entered in the open list with the total cost for the current node (S208), and the candidate with the lowest total cost among the nodes in the open list; The step S209 of designating a node as a next node and inputting it to the closed list, and determining whether the next node is a temporary target node (S210), and in the step S210 of determining a temporary target node, the next node is a temporary target. If it is a node, the entire path entered in the closed list including the path from the temporary target node to the target node is returned by back tracking. If the next node is not the temporary target node, the process returns to the path cost calculating step (S203). The next node of the next node input step S210 is treated as the current node of the path cost calculation step S203.

In the present embodiment, the description of the same parts as in the first embodiment will be omitted, and the description for the parts different from the first embodiment will be focused. Steps S203 to S209 of the second embodiment are substantially the same as steps S102 to S108 of the first embodiment.

FIG. 10 shows eight temporary target points having a radius of rotation with respect to the target, and FIG. 11 is a schematic view showing a path when the temporary target point is "140a".

Step S202 is a step of setting a temporary target point in consideration of the direction of the gripper of the target. As shown in FIG. 10, the eight temporary target points 140a to 140h are distanced by a rotation radius in the direction of the grip portion 131 of the target 130. The angle between the temporary target points 140a to 140h with respect to the target 130 is equal to 45 °. However, the present invention is not limited to this angle, and may be set to any other angle, and the angles between the temporary target points need not be the same.

As shown in FIG. 11, when the walking robot 110 reaches the temporary target point 140a, the walking robot reaches the target 130 by rotating after going straight by the rotation radius R. FIG. In this case, the heading portion 111 of the walking robot 110 is facing the holding portion 131 of the target 130.

12 is a schematic view showing the optimum path of the walking robot in consideration of the direction of the target node and the radius of rotation of the walking robot.

When the temporary target point is set in consideration of the directionality of the target 130 and the rotation radius of the walking robot 110, a path is searched for in accordance with the method according to the first embodiment. If the next node is the temporary target point in step S210, a path from the temporary target point to the target is formed to return the final path by back tracking. If the next node is not the temporary target point in step S201, it is repeatedly executed until the next node coincides with the temporary target point according to the method according to the first embodiment.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the invention. It will be clear to those who have knowledge of.

1 is a schematic view showing a coordinate system and components used in the present invention.

2 is a view showing an aroma meter used in the present invention.

Figure 3 is a schematic diagram showing the relationship between the nodes used to calculate the angular cost of the walking robot.

4 is a flowchart illustrating an optimal path planning method of a walking robot according to a first embodiment of the present invention.

5 is a schematic view showing the optimal path of the walking robot without considering the size of the walking robot and the size of the obstacle, Figure 6 is a schematic view showing the optimal path of the walking robot considering the rotation minimization of the walking robot and the size of the walking robot.

FIG. 7 is a schematic view showing an optimal path of a walking robot that can rotate in place, and FIG. 8 is a schematic view of an optimal path of a walking robot that cannot rotate in place.

9 is a flowchart illustrating a method for planning an optimal path of a walking robot in consideration of a rotation radius of the walking robot.

10 is a schematic view showing eight temporary target points having a radius of rotation with respect to a target.

FIG. 11 is a schematic diagram illustrating an optimal path when the temporary target point is “140a” of FIG. 10.

12 is a schematic view showing the optimum path of the walking robot in consideration of the direction of the target node and the radius of rotation of the walking robot.

<Description of the symbols for the main parts of the drawings>

110: walking robot

111: heading part

120: obstacle

130: target point

131: holding part

Claims (12)

Receiving and initializing information about a departure node, a target node, a previous node, a current node, and an obstacle node; Calculating a route cost from the departure node to the current node; Inputting a candidate node, which is not an obstacle node, among a plurality of candidate nodes in the vicinity of the current node to a candidate list; Calculating an angular cost of each candidate node input to the candidate list; Calculating a heuristic cost of each candidate node input to the candidate list; Calculating a total cost of each candidate node input to the candidate list according to the path cost, angle cost, and heuristic cost, and storing the total cost in each candidate node; Inputting the candidate node into an open list, updating the total cost of the node already entered in the open list with the total cost for the current node; Designating a candidate node having a minimum total cost among the nodes of the open list as a next node and inputting the candidate node into a closed list; Determining whether the next node is the target node; Modifying the path of the inflection point in consideration of the possibility of in-situ rotation of the walking robot, In the step of determining the target node, if the next node is the target node, the entire path input to the closed list is returned by back tracking, and if the next node is not the target node, the path cost calculation step returns The next node of the next node input step is treated as the current node of the path cost calculation step. Optimal path planning method of walking robot. delete The method of claim 1, wherein the angle cost is calculated by an angle formed by the previous node and the next node with respect to the current node. Optimal path planning method of walking robot. The method of claim 1, wherein the heuristic cost is calculated by a sum of a distance cost from the current node to each candidate node and a cost from the candidate node to the target node. Optimal path planning method of walking robot. The method of claim 1, wherein the total cost is calculated by the following formula: f (n) = g (n) + C _a (n) * α + C _h (n) * (1 -α), Where f (n) is the total cost, g (n) is the path cost, C _a (n) is the angle cost, C _h (n) is the heuristic cost, α is the weight of the angular cost, and α is a real number greater than zero and less than one. Receiving and initializing a starting node, a target node, a previous node, a current node, and an obstacle node; Setting a temporary target node having a predetermined radius of rotation with the target node in consideration of the direction of the target node and the direction of the walking robot; Calculating a route cost from the departure node to the current node; Inputting a candidate node, which is not an obstacle node, among a plurality of candidate nodes in the vicinity of the current node to a candidate list; Calculating an angular cost of each candidate node input to the candidate list; Calculating a heuristic cost of each candidate node input to the candidate list; Calculating a total cost of each candidate node input to the candidate list according to the path cost, angle cost, and heuristic cost, and storing the total cost in each candidate node; Inputting the candidate node into an open list, updating the total cost of the node already entered in the open list with the total cost for the current node; Designating a candidate node having a minimum total cost among the nodes of the open list as a next node and inputting the candidate node into a closed list; Determining whether the next node is the temporary target node; Comprising the step of modifying the path of the inflection point taking into account the possibility of rotation of the walking robot, In the step of determining the temporary target node, if the next node is the temporary target node, the entire path inputted to the closed list including the path from the temporary target node to the target node is returned by back tracking. If the next node is not a temporary target node, the process returns to the path cost calculation step, and the next node of the next node input step is treated as the current node of the path cost calculation step. Optimal path planning method of walking robot. delete The method of claim 6, wherein the angle cost is calculated by an angle formed by the previous node and the next node with respect to the current node. Optimal path planning method of walking robot. 7. The method of claim 6, wherein the heuristic cost is calculated by the sum of the distance cost from the starting node to the current node and the distance cost from the current node to the next node. Optimal path planning method of walking robot. The method of claim 6, wherein the total cost is calculated by the following formula: f (n) = g (n) + C _a (n) * α + C _h (n) * (1-α), Where f (n) is the total cost, g (n) is the path cost, C _a (n) is the angle cost, C _h (n) is the heuristic cost, α is a weighting coefficient of the angle cost, and α is a real number larger than 0 and smaller than 1. Receiving and initializing information about a departure node, a target node, a previous node, a current node, and an obstacle node; Calculating a route cost from the departure node to the current node; Inputting a candidate node, which is not an obstacle node, among a plurality of candidate nodes in the vicinity of the current node to a candidate list; Calculating an angular cost of each candidate node input to the candidate list; Calculating a heuristic cost of each candidate node input to the candidate list; Calculating a total cost of each candidate node input to the candidate list according to the path cost, angle cost, and heuristic cost, and storing the total cost in each candidate node; Inputting the candidate node into an open list, updating the total cost of the node already entered in the open list with the total cost for the current node; Designating a candidate node having a minimum total cost among the nodes of the open list as a next node and inputting the candidate node into a closed list; Determining whether the next node is the target node, In the step of determining the target node, if the next node is the target node, the entire path inputted to the closed list is returned by back tracking, and if the next node is not the target node, the process returns to the path cost calculation step. The next node of the next node input step is treated as a current node of the path cost calculation step, The optimal path planning method of the walking robot, the total cost is calculated by the following formula: f (n) = g (n) + C _a (n) * α + C _h (n) * (1 -α), Where f (n) is the total cost, g (n) is the path cost, C _a (n) is the angle cost, C _h (n) is the heuristic cost, α is the weight of the angular cost, and α is a real number greater than zero and less than one. Receiving and initializing a starting node, a target node, a previous node, a current node, and an obstacle node; Setting a temporary target node having a predetermined radius of rotation with the target node in consideration of the direction of the target node and the direction of the walking robot; Calculating a route cost from the departure node to the current node; Inputting a candidate node, which is not an obstacle node, among a plurality of candidate nodes in the vicinity of the current node to a candidate list; Calculating an angular cost of each candidate node input to the candidate list; Calculating a heuristic cost of each candidate node input to the candidate list; Calculating a total cost of each candidate node input to the candidate list according to the path cost, angle cost, and heuristic cost, and storing the total cost in each candidate node; Inputting the candidate node into an open list, updating the total cost of the node already entered in the open list with the total cost for the current node; Designating a candidate node having a minimum total cost among the nodes of the open list as a next node and inputting the candidate node into a closed list; Determining whether the next node is the temporary target node; In the step of determining the temporary target node, if the next node is the temporary target node, the entire path entered in the closed list, including the path from the temporary target node to the target node, is returned by back tracking. If the next node is not a temporary target node, the process returns to the path cost calculation step, and the next node of the next node input step is treated as a current node of the path cost calculation step, The optimal path planning method of the walking robot, the total cost is calculated by the following formula: f (n) = g (n) + C _a (n) * α + C _h (n) * (1-α), Where f (n) is the total cost, g (n) is the path cost, C _a (n) is the angle cost, C _h (n) is the heuristic cost, α is a weighting coefficient of the angle cost, and α is a real number larger than 0 and smaller than 1.

Images