论文部分内容阅读
从线性多变量系统状态空间规范型导出了系统的新输入输出辨识规范型，该模型直接与系统的输入输出数据相联系，且待辨识的参数较少，与现在的Ｇｕｉｄｏｒｚｉ辨识模型相比，大大减少了各子系统输出间的重迭（ｏｖｅｒｌａｐｐｉｎｇ）数目。给出了确定其结构指标的算法。数值实例表明了算法的有效性。
The new I / O identification norms of the system are derived from the linear multivariable state space specification model. The model is directly related to the input and output data of the system and has fewer parameters to be identified. Compared with the current Guidorzi model, Reduce the number of overlaps between each subsystem output. The algorithm to determine its structure index is given. Numerical examples show the effectiveness of the algorithm.