A Novel Rapid Prediction Method for Electromagnetic Noise of Electric Motors Based on Spatiotemporal Structural Transfer Function

Accurately calculating the structural transfer function of electric motors (EMs) represents a pivotal technology for electromagnetic noise prediction. However, computing the spatiotemporal structural transfer function (SSTF) remains a significant challenge due to the strong spatiotemporal coupling of electromagnetic forces. To address this challenge, this study proposes an innovative theoretical calculation method for SSTF to achieve highly precise and efficient prediction of electromagnetic noise. The core of this method lies in decoupling the SSTF into two crucial components: a stiffness-governed term and a modal-governed term. For the former, a static mechanical model of the stator-housing coupled structure is established based on elastic mechanics. By rigorously considering the continuity and load boundary conditions, the static displacement of the coupled structure is determined. As for the modal-governed term, it is derived from the natural frequency and damping ratio of the system. The SSTF is then obtained by combining these two terms. Subsequently, the electromagnetic excitation is decoupled in the temporal and spatial domains using the two-dimensional (2D) FFT. By integrating the proposed SSTF, the semi-analytical prediction of electromagnetic noise is achieved. Experimental results reveal that the peak error of the predicted noise remains within 2.4 dBA. The proposed methodology provides valuable guidance for the efficient and accurate prediction of electromagnetic noise.