基于人工神经网络组合预测油田产量

油田原油产量的准确预测可以对油田的生产管理进行合理的指导。该文探讨了应用神经网络组合方法预测油田产量，对开井数、含水率、动用储量以及往年产量同未来产量之间的复杂关系建立模型。采用了两层预测系统：第一层包含两个神经网络，一个多层前馈网络和一个函数链接网络；第二层是把第一层的两个网络输出进行组合。研究了五种不同的组合算法：平均法、最小平方回归法、模糊逻辑法、自适应前馈神经网络法和自适应函数链接神经网络法。根据油品类型分为稀油、热采稠油、常规稠油和总产量四组数据，对上述方法进行了测试，结果表明应用人工神经网络的组合预测方法优于其他的预测方法，而且适用范围广。

Oilfield Output's Combined Forecast Based on Artificial Neural Networks

The exact forecast for oil output can guide the production and management in the oilfield. In this article, the question about oilfield output's forecast is discussed using the combination of artificial neural network (ANN). ANN can model the complex relationship among well-open number, ratio of containing water, exploited reserves and future output. A two-stage system is proposed with the first stage containing two ANN, a multilayer feed-forward ANN and a functional link ANN. The second stage consists of a combination module to mix the two individual ANNs produced in the first stage. Five different combination algorithms are examined, they are based on: averaging, recursive least squares, fuzzy logic, feedforward ANN, functional link ANN. The performance is tested on real data from four different oil types?The results indicate that combination strategies based on a single ANN outperform the other approaches.