An Improved Differential Evolution Algorithm for Maritime Collision Avoidance Route Planning

Table 4

Settings and operating rules of the three optimizing algorithms.

Algorithm	Parameters	Settings	Operations	Settings
GA	Population size	90	Population initialization	Encode each individual with real coding; each real number string represents a route
	Individual dimension	30	Selection	Roulette wheel selection
	Crossover rate	0.4	Crossover	A single point crossover
	Mutation rate	0.2	Mutation	Select the th gene of the th individual for mutation
	Maximum number of iterations	100	Fitness value calculation	Minimum distance + minimum threat
DE	Population size	90	Population initialization	Generate initialisation vectors randomly
	Individual dimension	30	Selection	Greedy selection
	Crossover rate	0.85	Crossover	Binomial crossover
	Scalar weight	0.6	Mutation	Disturb current solution by using differential vectors
	Maximum number of iterations	100	Fitness value calculation	Minimum distance + minimum threat
MNDE	Population size	90	Population initialization	Generate initialization vectors randomly
	Individual dimension	30	Selection	Random selection for generating the neighborhood
	Crossover rate	0.85	Crossover	Binomial crossover
	Scalar weight	0.6	Mutation	Neighborhood-based mutations
	Jittering parameter	0.0001
	Maximum number of iterations	100	Fitness value calculation	Minimum distance + minimum threat