预测及其主要模型的实现系统

Realized system of prediction and its main models

针对已有预测模型并不完善的问题 ,在总结现有的预测模型 (线性回归模型、时间序列分析、线性随机模型、神经网络模型等 )基础上 ,采用实验分析的方法提出一种新的加权模糊推理的预测模型及基于马尔科夫性的区间预测方法 ,给出了实现系统 ,经试验验证 ,这种模型是比较准确的 ,一般误差不超过 1 0 %。

The prediction problem and models are discussed. Although there have been some prediction models widely used, they are not completely satisfied. A general introduction on these models and a new model based on fuzzy reasoning are proposed. Experiment shows that it is more accurate that other models. A method of section prediction based on Markov property is also presented.