基于神经网络建模和遗传算法的重油脱盐系统优化研究

A study of the optimization of the desalting system of heavy oil based on the neu-ral network modeling and the genetic algorithms

概述了重油脱盐系统的BP神经网络建模以及基于遗传算法的系统优化过程,将遗传算法与惩罚函数法相结合应用于约束优化的问题,改善了遗传算法的局限性。同时为了将不等式约束优化问题转化为单目标优化问题,对惩罚函数法进行了改进。结果表明:此方法可以有效解决静电脱盐问题。

By applying the combination of the genetic algorithms and the penalty function to the restriction optimization problem, we can improve the limitation of the genetic algorithms. This paper introduces the neural network modeling of the desalting system of heavy oil and the system optimization process based on the genetic algorithm. At the mean time, the penalty function was improved by transmuting the inequality constrained optimization problem into single target optimization one. The result shows that this method can effectively solve the problem of static electricity desalting.