【作者】 王宗志；

【导师】 金菊良；

【作者基本信息】 合肥工业大学 ， 水文学及水资源， 2005， 硕士

【摘要】 模糊集、人工神经网络、遗传算法是当前国际学术界三大学科前沿领域,它们突破了以传统数学模型为基础解决现实问题的思维方式,探索、模拟大千世界复杂多变的非线性特征,成为目前正蓬勃发展的新兴学科—智能科学的核心内容,是仿生学的三大重要分支。 自从模糊集、人工神经网络、遗传算法引入水资源模式识别方法以后,“模式”的内涵与外延大大拓广。本文尝试应用这三大方法解决水问题,提出了基于这三大方法的水资源智能模式识别的概念,给出了水资源智能模式识别方法解决水问题的一般步骤。详述了遗传算法存在的问题并提出一套从算法整体结构上改进的加速遗传算法;提出了基于改进加速遗传算法的水质模型参数智能识别方法、基于人工神经网络的智能模式识别方法、基于模糊集理论的智能模式识别方法、基于时间序列的智能模式识别方法,并将这些方法用于解决实际的水问题。结果表明水资源智能模式识别方法在具有广阔的应用前景。更多还原

【Abstract】 Fuzzy sets, artificial neural network, genetic algorithm are three important front fields in international academe, they break through the old thinking ways of solving practical problems based on traditional mathematical models and discover, simulate nonlinear complex problems appearing in universe usually. They are three important branches of bionics and the core components of intelligence science that is developing vigorously.Since fuzzy sets, artificial neural network, genetic algorithm are applied to the methods of water resources pattern recognition, the connotation and extension on ’pattern’ have become more general. In this paper, the three methods are attempted to solve water problems, at the same time, the definition of water resources intelligence pattern recognition and the general steps solving water problem using it are given. On the other hand, a series of methods are put forward and applied to practical problem on water resources, such as, water resources models parameters intelligence pattern recognition method based on improved genetic algorithm, water resources intelligence pattern recognition method based on artificial neural network, water resources intelligence pattern method recognition based on fuzzy sets, water resources intelligence pattern recognition method based on times series analysis, etc. the better results of all these methods show that the methods of water resources pattern recognition should has splendid further in water resources system engineering.更多还原