基于数据挖掘和小波神经网络的航材消耗预测方法

运用数据挖掘技术对航材消耗的历史数据进行关联分析，筛选小对保障飞机飞行有重要作用的航材消耗数据，大大缩减了需要预测的航材数量，同时对消耗航材之间的内在影响关系进行量化。在分析人工鱼群算法原理的基础上，对算法中步长参数和视野范同参数的设置方法进行了改进。实例结果表明，运用小波神经网络预测航材消耗的方法大大降低了预测误差，说明了该方法的有效性、可行性和实用性。

Prediction Method of Air Material Consumption Based on Data Mining and WNN

In this paper, the correlation analysis on historical data of air material consumption was presented by using data mining technology, tilting out the important material consumption data on the protection of aircraft flight, greatly reducing the amount of air material needed to forecast, and the influence between consumption materials relationship was quantified. The principle of artificial fish swarm algorithm was analyzed, and the setting method of step parameter and visual field parameter was improved on the basis of it. The example results showed that the method of wavelet neural network could greatly reduce the prediction error of air material consumption, illustrated the effectiveness, feasibility and practicality of the method.