Energy Efficiency Optimization Flow Shop Scheduling Algorithm Based On Intelligent Evolutionary Algorith

The permutation flow shop scheduling problem is an important type of flow shop scheduling problem,which widely exists in many fields such as manufacturing.The upgrade of production equipment can provide different processing speeds for jobs,leading to different processing times and energy costs.Motivated by the government’s emission peak and carbon neutrality" and intelligent manufacturing plans,it is necessary to study energy-efficient scheduling algorithms for reducing energy consumption in manufacturing processes.Focuses on the permutation flow shop scheduling problem,this thesis proposes energy-efficient scheduling algorithms based on intelligent evolutionary algorithms to produces scheduling solutions minimizing the total energy consumption.This thesis proposes an improved cuckoo search algorithm-based intelligent evolutionary scheduling algorithm that takes into account different processing speeds of production equipment.A best-cuckoo-guided movement strategy is constructed in which the current best cuckoo guides other cuckoos for updating their positions.This strategy helps the cuckoo search algorithm avoid losing useful information,and therefore,improves the quality of final scheduling solution.In addition,an improved algorithm is designed to generate the initial best solution.We further develop a nest-abandoning updating strategy based on the number of iterations to control the searching range and to balance the effectiveness and efficiency of scheduling algorithms.The results of experiments performed on benchmark testing instances show that the improved cuckoo search algorithm proposed in this work significantly outperforms the traditional cuckoo search algorithm,verifying that the best-cuckoo-guided movement strategy is better than the traditional movement strategy.The effectiveness of the proposed solution initialization strategy and the nest-abandoning updating strategy are also justified.We then study the bi-objective permutation flow shop scheduling problem for jointly minimizing the total completion time and the total energy consumption.Based on the aforementioned algorithm,we propose a hybrid cuckoo search algorithm based on the wellknown Pareto-optimality theory to obtain a set of non-dominated solutions to the bi-objective optimization problem.We establish a new strategy to generate the initial non-dominated solutions and store the generated solutions in an archive set.When evaluating the fitness value of a cuckoo individual,we randomly select a solution in the archive set to decode the cuckoo into a feasible solution.In order to ensure the diversity of the initial archive set,we construct an initial archive set generation method.This method uses the initial solution generation strategy to generate two solutions that minimize the total completion time and total energy consumption,respectively.Then,we generate multiple initial solutions by using different strategies and store the solutions into the archive set.Furthermore,we propose a two-stage local search to adjust the job sequence and processing speed,respectively,for the purpose of exploring more non-dominated solutions.Experimental results demonstrate the effectiveness and efficiency of the proposed hybrid cuckoo search algorithm by comparing it with the stateof-the-art algorithm.The proposed intelligent evolutionary algorithms are capable of solving the energyefficient permutation flow shop scheduling problem by improving the performance of the main operators in the cuckoo search algorithm.Thus,the proposed algorithms are beneficial for manufacturing enterprises in reducing energy consumption,improving productivity,and saving manufacturing costs.In our future work,we will extend the intelligent evolutionary algorithms in this thesis and apply them in a variety of intelligent manufacturing scenarios,providing manufacturing enterprises with more energy-efficient solutions.