基于神经网络和遗传算法的采油控制系统  

Oil Pumping Control System Based on Neural Network and Genetic Algorithm

作　　者：李英[1] 李元春[1] 

机构地区：[1]吉林大学通信工程学院,长春130022

出　　处：《吉林大学学报（工学版）》2006年第1期82-86,共5页Journal of Jilin University:Engineering and Technology Edition

基　　金：吉林省科技发展计划项目(20040532)

摘　　要：为了解决部分抽油机“长期相对轻载”和“空抽”的问题,采用抽油机间歇采油控制方法对采油控制系统进行了设计。利用非线性规范化方法的非线性同伦BP神经网络对采油模型进行辨识,采用遗传算法优化停机时间。该控制系统在油田中的实验结果表明,在保证了采油量的前提下,节电率达30%以上,实现了抽油机采油的智能控制。In order to solve the problems of long-term light-load and idle pumping that many oil wells faced, a control system to operate the oil pump intermittently was developed. It is based on nonlinear homotopic BP neural network with nonlinear normalization method to identify the oil pumping model, and the genetic algorithm to optimize the downtime. The spot test of the control system showed that up to 30% of the energy - saving rate was achieved under the guaranteed oil output and the intelligent control of the oil pumping was realized.

关 键 词：自动控制技术 非线性同伦 神经网络 遗传算法 采油控制系统 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]