基于量子遗传算法和模糊神经网络的油田开发综合经济评价  

An economy evaluation method of oilfield development project based on quantum genetic algorithm and normalized fuzzy neural networks

作　　者：王宏伟[1] 夏斌[2] 

机构地区：[1]中国科学院研究生院,北京100049 [2]中国科学院广州地球化学研究所,广东广州510640

出　　处：《大庆石油学院学报》2008年第3期93-97,共5页Journal of Daqing Petroleum Institute

摘　　要：针对油田开发方案综合经济评价的复杂性和传统神经网络评价方法难于收敛的问题,提出一种基于量子遗传算法和正规模糊神经网络的评价方法.该算法使用量子遗传算法优化网络参数,完成网络的训练;能够充分利用专家制订的"if—then"规则,完善网络的推理结构,提高网络的识别能力,减少噪声因素的影响.仿真结果表明,该方法对油田开发方案综合经济评价具有良好的适应性和实用性.Aiming at the complexity of economy evaluation for oilfield development project and the convergence difficulty for the traditional neural networks evaluation, a comprehensive economy evaluation method is proposed based on quantum genetic algorithm and normalized fuzzy neural networks. In this method, the quantum genetic algorithm is applied to optimize the network parameters, and the fuzzy 'if-then' rules can be sufficiently utilized formulated by experts. Hence, this method can perfect the reasoning structure of network, improve the network recognition capacity, and reduce influence of disadvantage factor. The experiment shows that this algorithm has a better adaptability and practicability for the comprehensive economy evaluation of oilfield development project.

关 键 词：量子遗传算法 模糊神经网络 综合经济评价 油田 

分 类 号：F061[经济管理—政治经济学]