Pareto_front_index = pareto_front_finding(fitness_values, index)įitness_values = fitness_values Offspring_from_local_search = local_search(pop, rate_local_search, step_size) Offspring_from_mutation = mutation(pop,rate_mutation) Offspring_from_crossover = crossover(pop,rate_crossover) Pop = random_population(nv,pop_size,lb,ub) While len(pareto_front_index) < pop_size: Pareto_front = np.ones(pop_size, dtype=bool) # initially assume all solutions are in pareto front by using "1"ĭef selection(pop, fitness_values, pop_size): Selected_pop_index = np.asarray(selected_pop_index, dtype=int) # Convert the data to integerĭef pareto_front_finding(fitness_values, pop_index):
Selected_fitness_values = np.zeros((number_solutions_needed, len(fitness_values))) Selected_pop_index = np.zeros((number_solutions_needed)) Pop_index = np.arange(fitness_values.shape)Ĭrowding_distance = crowding_calculation(fitness_values) Matrix_for_crowding = np.zeros((pop_size, fitness_value_number))Ĭrowding_distance = np.sum(matrix_for_crowding, axis=1) # crowding distance of each solutionĭef remove_using_crowding(fitness_values, number_solutions_needed): # number of offspring chromosomes generated from the local searchįitness_values = np.zeros((pop.shape, 2)) # because of 2 objective functionsįitness_values = 1 - math.exp(-sum((chromosome - 1/math.sqrt(n))**2))įitness_values = 1 - math.exp(-sum((chromosome + 1/math.sqrt(n))**2))ĭef crowding_calculation(fitness_values):įitness_value_number = len(fitness_values) Offspring = np.zeros((mutation_rate, pop.shape))Ĭutting_point = np.random.randint(0, pop.shape) Offspring = np.zeros((crossover_rate, pop.shape))Ĭutting_point = np.random.randint(1, pop.shape)
0 kommentar(er)
