针对单矢量水听器估计信号多维参数问题,提出了一种利用状态空间模型联合估计方位和频率的新方法.利用单矢量水听器接收到的数据,构造一个特殊的状态空间模型,由此可得到包含了信号水平方位角和频率信息的状态过渡矩阵和观测矩阵.方位角的估计值由状态过渡矩阵的特征值获得,频率的估计值由观测矩阵经过一定的数学变换获得,无需搜索过程并且能够实现参数的自动配对,不存在角度兼并的问题.由于状态空间方法对模型误差不敏感,因而具有更好的数值稳定性.仿真结果表明,该方法具有良好的估计性能.
A new method has been proposed to solve the problem of whether using a single vector hydrophone can estimate multi-dimensional parameter of a signal, which utilizing a state-space model joint estimate frequency and azimuth approach. The state transition matrix obtained information that included information of signal azimuth along with an observation matrix that included information of signal frequency. The estimates of azimuth were obtained by eigenvalues of the state transition matrix, while the estimates of frequency were obtained by the observation matrix, which needed mathematical transformation, not at anytime was a search process needed, and the parameters were matched automatically that allowed the avoidance of the angle merger. As the state-space model was not sensitive to model errors, it had better numerical stability. The simulation results indicate that the approach allows for a good estimation performance.