Genetic Scheduling and Reinforcement Learning in Multirobot Systems for Intelligent Warehouses

Algorithm 1

Genetic algorithm for warehouse scheduling.

(1) Initialize parameters: population size Popsize, maximal generations MaxEpoc, crossover rate Pc;

(2) Calculate the distances between different task orders;

(3) Generate an population of feasible solutions randomly;

(4) Evaluate individuals in the initial population;

(5)  for   to MaxEpoc do

(6)   Set the alterable mutation rate Pm according to the actual evolution generations;

(7)   Select individuals according to elitist strategy and roulette wheel method;

(8)  if random   then

(9)    Crossover individuals in pairs by a variation of the order crossover operator;

(10)   end if

(11)   if random   then

(12)    Mutate an individual by swap mutation;

(13)   end if

(14)  Evaluate the produced offspring;

(15) end for