Abstract: Addressing the error calibration of the ultrasonic phased array testing system, an optimized calibration method with integration of the intelligent learning algorithm is proposed. Based on the comparison with the limitation of the conventional calibration method, the said method innovatively employs the difference evolution algorithm to construct an entire flow process calibration system which mainly includes data collection,data preprocessing,intelligent training,parameter optimization,real-time feedback and result output. The experiment result shows that this method uses the differential evolution algorithm to train preprocessed data and optimize the error of ultrasonic phased array detection system so as to effectively learn how to adjust system parameters based on historical data and real-time feedback data to minimize the system error for the purpose of improving the accuracy and stability of the detection result.