Apparatus for fast path search by learning heuristic function and method thereof

1. An apparatus for fast path search, comprising:

an input unit to receive current location information and destination information from a user, and to receive a search request for a path to the received destination information; and a control unit to search a map for a path from a start node corresponding to the current location information to a destination node corresponding to the destination information using a heuristic evaluation function,

wherein the heuristic evaluation function is expressed as a summation of a first cost to move from the start node to a current node on the path to the destination node, and a second cost to move from the current node to the destination node, the second cost including a cost that is calculated by multiplying a lineal distance from the current node to the destination node by a distance weight, and the path from the start node to the destination node is selected as a path for avoiding an obstacle on the map, and

wherein the distance weight divides the map into a plurality of spaces and indicates a ratio of a distance between two divided spaces to an average lineal distance.

2. The apparatus of claim 1 , wherein:

the plurality of spaces are formed by clustering nodes on the map; and

the second cost is based on the distance weight which corresponds to a clustered space of the current node and a clustered space of the destination node.

3. The apparatus of claim 2 , further comprising:

a learning unit to generate a heuristic set for the distance weight and update the heuristic set as path searching to the destination node is performed a predetermined number of times.

4. A method for fast path search, comprising:

receiving destination information from a user to receive a search request for a path to the received destination information;

obtaining current location information of the user when the search request for the path is received; and

searching a map for a path from a start node corresponding to the current location information to a destination node corresponding to the destination information using a heuristic evaluation function,

wherein the heuristic evaluation function is expressed as a summation of a first cost to move from the start node to a current node on the path and a second cost to move from the current node to the destination node, the second cost including a cost that is calculated by multiplying a lineal distance from the current node to the destination node by a distance weight, and the path from the start node to the destination node is selected as a path for avoiding an obstacle on the map,

wherein the distance weight divides the map into a plurality of spaces and indicates a ratio of a distance between two divided spaces to an average lineal distance.

5. The method of claim 4 , wherein:

the plurality of spaces are formed by clustering nodes on the map; and

the second cost is based on the distance weight which corresponds to a clustered space of the current node and a clustered space of the destination node.

6. The method of claim 5 , further comprising:

generating a heuristic set for the distance weight; and

updating the heuristic set as path searching to the destination node is performed a predetermined number of times.