中文摘要：

针对低信噪比情况,建立了新的接收阵列误差模型,提出了一种改进单辅助源通道阵列幅相误差校正算法。该算法利用自相关矩阵对角线部分求解通道幅度误差,用最小二乘法对上三角部分求解通道相位误差。仿真结果表明：在低信噪比时,改进校正算法性能明显优于传统校正算法,在高信噪比时,2种算法残差均非常小。实验结果表明：利用改进的校正算法对综合孔径辐射计进行校正后所得的图像边界清晰、干扰较少,与实际亮温对应较好,明显优于采用传统校正算法进行校正后所得的图像。

英文摘要：

Aiming at low SNR,a new receiving array error model is established,and an improved single source calibration method for amplitude and phase error is proposed in this paper.This method uses diagonal of self-correlation matrix to estimate channel amplitude error,and uses up-triangle part to estimate channel phase error by least square criterion.Simulation results show at low SNR,the proposed method has better performance than the conventional method;at high SNR,residual errors of both methods are very small.Experiment results show that better brightness temperature images with clear fringe and little disturbance are achieved by using our calibration method,rather than by using the conventional method.