Abstract: In view of the array mismatch caused by the severe tremor of the aircraft at high speed, a fast channel prediction and calibration algorithm are presented to improve the array beamforming robustness. In this study, we first analyzed the degree of distortion of the airborne conformal arc array beam pattern under certain error perturbations. And assumed random channel amplitude and phase errors, the prediction and calibration algorithms for both types of errors were given. The prediction and calibration algorithms need to be combined with the received array signal model in the context of the conformal array, and specific array robustness determination basis was given to analyze the performance in terms of array element calibration error and amplitude and phase error. Then a conformal array error prediction algorithm based on the standard calibration element technique is proposed, which can be used not only for array calibration at the beginning of array design, but also for channel correction in the actual use. Finally the algorithm can perform error prediction for spatially distributed arrays based on three calibration elements, and the error prediction values can be directly deployed in the baseband section for real-time calibration. In an array with a spacing of 10 elements, there is a good prediction for random jitter of 5, among which a single prediction deviates most from the actual error. And the use of error prediction values can be directly deployed in the baseband section for real-time calibration, with a calibration accuracy of within 2 accuracy values and high real-time error correction ability.