基于时序分析与神经网络的能源产量预测模型

实际生产系统中存在大量时间序列问题，为了研究系统的结构和规律，我们需要建立时间序列模型，对其进行预测和分析。目前时间分析方法多采用AR或ARMA模型，但由于实际问题错综复杂，导致模型求解困难，实际中难以应用。为了解决上述问题，首先分析了生产系统时序分析的基本原理，利用BP神经网络建立了时序—神经网络模型，然后利用该模型对能源产量进行了预测。通过预测结果的分析可看出，该模型具有利用方便、动态性能好、预测准确性高等优点，在实际中具有一定的实用价值。

Forecast model of energy production based on time series analysis-artificial neural network

There are many time series problems in the actual production system. To study system structure and regularity, we need to build time series model. With the time series model, we can forecast and analyze the production system. At present, the time series analysis method often uses AR or ARMA model, this method is very complicated and difficult to apply. To overcome the above problems, this paper analyzed basic principle of the time series analysis, and the model of time series analysis-artificial neural network is founded on BP neural network. The energy production was forecast with this model. The forecast result proved that this model has many advantages of convenient use, fine dynamic function and high precision. The model has fine practical value.