Framework for efficient dynamic analysis applied to a tubular generator for suspension applications

This paper considers a slotless three-phase tubular permanent magnet generator located in an automotive suspension system for the application of vibration energy harvesting. A two-dimensional finite element method model of the harvester is produced and an experimental setup that contains the generator is constructed. Signal decomposition methods are applied to measured suspension displacement data and the resulting signal components are used as input for the model. Two approaches for signal decomposition are discussed, namely, the discrete Fourier transform and the continuous wavelet transform. The individual emf responses of the model are reconstructed to a single output, while a sideband prediction algorithm accounts for the non-linearities in the system. The simulation results are compared with the reference measurements conducted on the setup to determine the accuracy of each of the signal decomposition methods.