Construction of pursuit strategies with using Lyapunov functions
This paper is concerned with differential pursuit-evasion games, in which several agents chase one. The time of capture of a target is used as the criterion. The motion of agents is simple one, the velocities are piecewise-continuous. The function that specifies the maximal time of capture of the ta...
Saved in:
| Date: | 2018 |
|---|---|
| Main Author: | |
| Format: | Article |
| Language: | Ukrainian |
| Published: |
PROBLEMS IN PROGRAMMING
2018
|
| Subjects: | |
| Online Access: | https://pp.isofts.kiev.ua/index.php/ojs1/article/view/305 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Journal Title: | Problems in programming |
| Download file: |  |
Institution
Problems in programming| Summary: | This paper is concerned with differential pursuit-evasion games, in which several agents chase one. The time of capture of a target is used as the criterion. The motion of agents is simple one, the velocities are piecewise-continuous. The function that specifies the maximal time of capture of the target for the well-known strategy of parallel approach is described. This function is used as a Lyapunov function for constructing the new chase strategy, which outperforms the strategy of parallel approach in the following sense. Maximal time of pursuit for the new strategy is not more than maximal time of pursuit for the strategy of parallel approach; at the same time there are many games, for which maximal time of pursuit for the new strategy is less than for the strategy of parallel approach. In case of pursuit-evasion game on a plane we find explicit form of Lyapunov function and calculate velocities of pursuers using the gradient of this function. Numerical examples show that such velocities of pursuers reduce the maximal time of pursuit. In case of pursuit-evasion game in a multidimensional Euclidean space, Lyapunov function is equal to an optimal value of an objective function of appropriate linear programming problem. The velocities of pursuers are calculated with using the gradient of this function.Problems in programming 2017; 3: 194-211 |
|---|