Research Article
Multiagent Task Planning Based on Distributed Resource Scheduling under Command and Control Structure
Algorithm 2
MultiAgent-HTN algorithm flow in distributed environment.
(1) | Procedure MA-HTN | (2) | set , initial state | (3) | set , initial task set | (4) | set , domain | (5) | initial operator | (6) | set the empty plan | (7) | | (8) | loop | (9) | if is empty then return | (10) | nondeterministically choose any in | (11) | if is a primitive task then | (12) | choose a ground instance for with the smallest among the available resources | (13) | | (14) | | (15) | | (16) | send to the other planners | (17) | if | (18) | | (19) | conflict actions | (20) | | (21) | , then | (22) | if | (23) | delete the rest in | (24) | | (25) | delete in and add into | (26) | else if | (27) | delete in | (28) | | (29) | Backtrack | (30) | else delete in and add into | (31) | if is a nonprimitive task then | (32) | choose a method to decompose into subtasks | (33) | delete in and add into | (34) | if receiving a new in of and | (35) | | (36) | conflict actions | (37) | | (38) | | (39) | if | (40) | delete the rest in | (41) | | (42) | Backtrack to task node at action | (43) | else if | (44) | delete in | (45) | | (46) | delete in and add into | (47) | repeat | (48) | end MA-HTN |
|