Purpose This paper aims to propose a new ultrasonic detection method for stainless steel weld defects based on complex synergetic convolutional calculation to solve two problems in the ultrasonic detection of austenitic stainless steel weld defects. These include ignoring the nonlinear information of the imaginary part in the complex domain of the signal and the correlation information between the amplitude of the real part and phase of the imaginary part and subjective dependence of diagnosis model parameters. Design/methodology/approach An ultrasonic detection method for weld defects based on complex synergetic convolution calculation is proposed in this paper to address the above issues. By mapping low-density, 1D detection samples to a complex domain space with high representation richness, a multi-scale and multilevel complex synergetic convolution calculation model (CSCC) is designed to match and transform samples to mine amplitude changes, phase shifts, weak phase angle changes and amplitude-phase correlation features deeply from the detection signal. This study proposed an Elite Sine-Cosine Sobol Sampling Dung Beetle Optimization Algorithm, and the detection model CSCC achieves global adaptive matching of key hyperparameters in 19 dimensions. Finally, a regulative complex synergetic convolutional calculation model is constructed to achieve high-performance detection of weld defects. Findings Through experiments on a self-developed Taiyuan intelligent detection and information processing weld defect dataset, the results show that the method achieves a detection accuracy of 92% for five types of weld defects: cracks, porosity, slag inclusion and unfused and unwelded components, which represent an average improvement of 11% relative to comparable models. This method is also validated with the PhysioNet electrocardiogram public dataset, which achieves better detection performance relative to the other models. Originality/value This method provides a theoretical basis and technical reference for developing and applying intelligent, efficient and accurate ultrasonic defects detection technology.