摘要
军品订货量的预测是规划未来国防工业生产规模的关键．本文首先分析确定了影响陕西国防工业军品订货量的主要因素，对其中的定性因素进行了量化；之后将预测过程模拟成一个“Ｆｕｚｚｙ”系统，预测依据作为系统的输入，预测量作为系统的输出，根据历史的输入、输出数据确定系统的特性，建立预测模型；以此为基础，给定输入数据，在系统特性的作用下，即可求得预测输出．由于使用了一种新的基于Ｔ－Ｓ模糊模型的自适应模糊神经网络，从而使预测模型具有很强的自适应能力．本文的军品订货预测方法具有比较强的通用性．
The forecast of military products orders is the key to planning the production scale of the future defence industry.Firstly, the main factors which influence the orders shaanxi military products are analysed and among which the qualitative factor is quantified.Then the forecast process is formulized as a 'fuzzy' system. Forecast basis is acted as the input and the predicted quantity is served as the output.Thus the characteristic of the system and the predicted model are formed according to the input output relationship.Based on this, under the given input conditions and the influence of the system characteristic, the predicted output can be obtained.The predicted model has very strong self adaptabilitry because of the use of the new adaptive fuzzy neural network based on T-S theory.The method of the forecast of military products orders can be widely used in other fields.
出处
《系统工程理论与实践》
EI
CSCD
北大核心
1999年第3期38-46,共9页
Systems Engineering-Theory & Practice
基金
国家自然科学基金
