【摘要】 为了高效准确地实现多输入多输出系统的数据建模 ,本文提出了数据分层建模算法 .将多输入多输出数据建模问题分解为一系列单输入多输出的建模问题 ,同时改进标准遗传编程算法 ,将单树表示的的个体结构扩展为向量树的进化个体 ,实现多输出系统的自动建模 .通过大量的计算实例表明 ,这种算法可以实现复杂的多输入多输出系统的建模 ,提高数据建模的效率和精度更多还原

【Abstract】 Genetic programming is applied to acquire models automatically from a mass of data. To improve data modeling for multi input multi output system, a data hierarchical modeling algorithm is developed to make significant discovery of multiple dimensional models efficiently and automatically. Multi input multi output modeling problem is decomposed into a series of single input multi output ones. And a single input multi output data set is modeled via the improved genetic programming algorithm, in which single tree expressed as each individual is extended to vector one. Many computational results show that the data hierarchical modeling algorithm discovers more complex mathematical models with high accuracy for multi input multi output systems and enhances efficiency and validity of multiple dimensional data modeling.更多还原