讨论了运用自组织方法建立预测数学模型的智能性算法。自组织方法是基于功能系统理论设计的新一代智能控制系统中的核心算法,用于建立控制对象预测模型。为克服传统自组织方法所需计算资源庞大的局限,提出了由DEMARK趋势项和自组织方法相结合的改进型算法,并将该算法用于建立导航系统误差参数的预测模型。通过对导航系统误差预测的仿真研究,验证了改进型自组织算法在飞行器智能控制系统中建立预测模型的有效性。

The intelligent algorithm for building prediction mathematical models using self-organization method was discussed in this paper. This method is one of the key points in modern intelligent system based on theory of functional systems for establishing pre-diction model of control objects. For avoiding shortcoming of traditional self-organization method requiring enormous computing re-source, the modified algorithm combining DEMARK trend and self-organization method was proposed, based on which the predic-tion model of error parameters of navigation system was constructed. Simulation research on application of the given algorithm in prediction of errors of navigation systems shows the effectiveness of this algorithm on building prediction model in aerocraft intelli-gent control system.