Table 1 Comparison of path planning algorithms
Algorithm classification | Algorithms | Basic theory | Advantage | Disadvantage |
|---|---|---|---|---|
Traditional algorithm | Simulated annealing algorithm | Simulating the annealing process of solid matter | Simple description, flexible use, high operation efficiency, less limit of initial condition | Slow convergence rate and randomness |
Artificial potential field algorithm | Simulating object motion under gravitation and repulsion force | Smooth and safe route, simple description | Local optimization | |
Fuzzy logic algorithm | Simulating driving experience | Bein in accordance with human mind and conducted without math modeling | Hard to conclude fuzzy rule | |
Tabu search algorithm | Simulating human intelligence behavior, and stepwise global optimization | Â | Â | |
Graphics algorithm | C space method | Expand the obstacles into polygon in motion space | Intuitive and easy to obtain the shortest route | Poor flexibility |
Grid method | Representing map with the coding trellis | Suitable for environment modeling | Hard to solve complex environmental information problem | |
Free-space method | Building free space using predefined basic shapes | High flexibility | Hard to be realized | |
Voronoi graph method | Performing space division using basic shapes called elements | Realizing effective obstacle avoidance | Taking too much time in redrawing | |
Bio-inspired intelligent algorithm | Ant colony algorithm | Iterative simulation of ants foraging behavior | Sound capacity of overall control, essential parallelism, easy to be realized by computer | Large amount of computation, and easy to fall in local optimization |
Genetic algorithm | Based on biological reproduction mechanism, chromosome selection, chromosome chiasma, and chromosome variation | Easy to be mixed with other algorithms to show iterative dominance | Insufficient operation efficiency | |
Neural network algorithm | Simulating animal neural network, and performing distributed parallel information processing | Excellent learning ability and robustness | Poor generalization ability | |
Particle swarm optimization | Simulating birds flying and feeding behavior | Simplicity, easy to be realized, sound robustness, quick convergence rate | Easy to fall in local optimization |