基于过程神经网络对原油产量预测技术的探讨

An Prediction Oil Production Technology Based on Process Neural Network

预测油田原油产量对分析油藏，制定油田合理的发展规划至关重要。该文基于神经网络，对原油产量预测技术进行了探讨，提出了一种基于过程神经网络的原油产量预测模型。该模型引入时间积累算子和神经元核函数，通过对输入集进行反复训练，得到合理的调整权重后得到输出变量。实验结果表明，该原油产量预测模型一致性较好，具有较高的精确度，对油田的原油产量预测有一定的应用前景。

Predicting crude oil output is very important to analysis petroleum reservoir performance and make oil field develop- ment planning. This paper discussed the oil production prediction technology, and one crude oil output prediction model based on process neural network were proposed. The model introduced time operator and neurons kernel function, and got the reason- able adjustment of weight and the output variable through repeated training based on the input set. The experimental result show that the consistency of the crude oil output prediction model is good and the accuracy is high, the crude oil production predic- tion has a certain application prospect.