混沌时间序列的模糊神经网络预测

Prediction of the chaotic time series using neuro-fuzzy networks

设计一种新型混合模糊神经推理系统 ,该系统仅从期望输入输出数据集即可达到获取知识、确定模糊初始规则基的目的 .再利用神经网络学习能力便不难修改规则库中的模糊规则以及隶属函数和网络权值等参数 ,这样大大减少了规则匹配过程 ,加快了推理速度 ,从而极大程度地提高了系统的自适应能力 .用它对Mackey Glass混沌时间序列进行预测试验 ,结果表明利用该网络模型无论离线还是在线学习均能对Mackey Glass混沌时间序列进行准确的预测 。

A novel hybrid neural fuzzy inference system is presented. Only based on the desired input-output data pairs, are the knowledge acquisition and initial fuzzy rule sets available. Then, employing neural networks learning techniques, the fuzzy logic rules, input-output fuzzy membership functions and weights in networks can be easily tuned. So the rule matching is reduced, inferencing is accelerated, adaptability of the system is greatly improved. To illustrate the performance of the proposed neuro-fuzzy hybrid model, simulations on the chaotic Mackey-Glass time series prediction are performed. Combining either off-line or on-line learning with the proposed hybrid model, we can show that the chaotic Mackey-Glass time series are accurately predicted, and demonstrate the effectiveness of the model.