Interharmonic estimation based on complex valued independent component analysis

To improve the resolution and accuracy of interharmonic estimation, we propose a new interharmonic estimation algorithm. The new algorithm is based on complex valued Independent Component Analysis (ICA) technique that has been used widely in solving blind source separation of complex valued signal. Due to the problem that the interharmonic model is different from complex valued ICA model, so we fist establish the complex valued ICA model of interharmonics in complex domain by setting time delay. Then we change the interharmonic estimation problem into the blind source separation of complex valued signal and apply complex valued ICA technique to solve the problem and estimate the interharmonic components. Simulation results show that the new algorithm has higher resolution and more accurate than traditional interharmonic estimation algorithms. Especially, it does not require singular value decomposition and eigenvalue decomposition, so its computational complexity is lower than the interharmonic estimation algorithm based on them.