基于GRA-IPSO-SVM的航材携行需求预测研究

异地执行飞行任务中航材需求的准确预测是做好携行保障的主要内容之一，为此提出灰色关联度（GRA）与改进的粒子群算法（IPSO）及支持向量机（SVM）相结合的航材预测方法。首先运用GRA 对航材携行需求的影响因素进行分析；其次引入活性因子和非线性惯性系数改进粒子群算法，并通过IPSO 对SVM 参数进行寻优；最后使用优化后的SVM 模型预测航材需求。结果表明：GRA-IPSO-SVM 方法预测结果的均方根误差比PSO-SVM 方法下降0.16，平均绝对百分比误差下降2.18%，且预测时间减少了0.7 s。

GRA-IPSO-SVM based on the demand forecasting of aviation material carrying

Accurate prediction of aviation material requirements for off-site missions is one of the main elements of a good trip assurance. To this end, this paper proposes a combination of grey correlation (GRA), improved particle swarm algorithm (IPSO) and support vector machine (SVM) as a method for predicting aviation material. Firstly, GRA is applied to analyse the factors influencing the demand for carriage of aviation materials; secondly, the particle swarm algorithm (IPSO) is improved by introducing activity factors and non-linear inertia coefficients, and the SVM parameters are optimised by IPSO; finally, the optimised SVM model is used to predict the demand for aviation materials. The simulation results show that the GRA-IPSO-SVM method has a 0.16 decrease in RMSE, a 2.18% decrease in MAPE and a 0.7s decrease in prediction time compared with the PSO-SVM method.