Logistics Loading Problem Based on Particle Swarm Auto-evolving Algorithm
In order to deal with the strong cargo heterogeneity in logistics loading process nowadays and im-prove filling rate and loading efficiency,we proposed a meta-heuristic algorithm,namely particle swarm auto-evolving algorithm,for solving the three-dimensional loading problem.Firstly,we established a mathematical model satisfying multiple constraints for the three-dimensional load-ing problem which is a core academic concept in logistics and freight transportation,and constructed a fitness function for obtaining the filling rate of the cargo to provide evaluation criteria for the proposed algorithm.Said algorithm consists of two main modules,respectively the limit point construction-based heuristic algo-rithm and the particle swarm auto-evolving rule.Therein,the limit point construction-based heuristic algo-rithm is responsible for decoding the particle swarm sequence,generating the solution,and evaluating its fit-ness.The concept of limit point is introduced as the basis of the heuristic algorithm,and three cargo placement strategies are designed there upon to derive a new heuristic method to effectively solve the three-dimensional pallet loading problem.Moreover,the particle swarm auto-evolving rule is responsible for the evolution of the coding solution(particle swarm sequence)so as to provide the optimal solution for the limit point construc-tion-based heuristic algorithm.Said rule proposes the method of representing the particles in the cargo loading sequence,deduces the partial uniform crossover operator and self-mutation operator,obtains the next genera-tion through the crossover and mutation of the particles,and updates the loading sequence,which provides par-ticle swarm sequence input for the limit point construction-based heuristic algorithm thus to optimize cargo loading.The comparison result shows that the proposed algorithm significantly improves the space utilization rate in the logistics loading process(raising the average loading rate of strong heterogeneity cargo to 85%),which veri-fies the effectiveness and superiority of the algorithm for the loading of strong heterogeneity cargo,and gives a three-dimensional model of cargo loading.Due to the lack of practical test sets,we respectively established an example generator for the cabin loading model and the pallet loading model to automatically generate the cargo output meeting the real situation in or-der to test the ability of the algorithm on solving practical logistics problems.Through the generated test set,we tested the algorithm,verified the compactness,practicability and speed of the algorithm in practical applica-tion,which can satisfy the actual requirements of aviation logistics,and gave the effect diagram of the corre-sponding loading model.
NETL
NSTL
万方数据
