Novel path planning of robots based on bidirectional ant colony algorithm

For solving the shortcomings of converging slowly and falling into partial delay easily, a new approach for robot path planning is introduced based on bidirectional ant colony searching the optimal solution at the same time. First of all, the two ant colonies at the original and the terminal run with different strategies. Searching on the two paths at the same time makes it more global. Secondly, the descending strategy is taken to adjust the scale of ant colony, so that the positive ants search paths mainly, the positive feedback is enhanced and the speed of the convergence is faster at the latter of the algorithm. Finally, each path can be optimized locally. According to the observations, redundant paths will be found possibly in a 3*3 grid, and it can be pulled into a straight-line distance. Experimental results show that the algorithm actualizes simply, effects stably, and the global optimal path can be found in short time after numerous running.