Grid synchronization is a critical concern for proper control of energy transfer between the Distributed Power Generation Systems (DPGS) and the utility power grid. Nonlinear estimation techniques are proposed to track the voltage magnitude, phase angle, and frequency of the utility grid. Instead of directly analyzing in abc coordinate frame, the symmetrical component is employed to separate the positive, negative, and zero sequences in the transformed AlphaBeta stationary coordinate frame. By using the Fortescue's Transformations and Clarke's Transformation, the number of system state variables is reduced to five. The results show that our proposed nonlinear estimation technique is efficient in smart power system synchronization. The MATLAB simulation studies have been conducted to compare the performance of the Extended Kalman Filter (EKF), the Particle Filter (PF), and the Unscented Kalman Filter (UKF). Computer simulations have shown that the efficacy of our proposed nonlinear estimation methods. It also shows that the Unscented Kalman Filter, and the Particle Filter are better estimators, because voltage synchronization problem is nonlinear, and linearization process which the Extended Kalman Filter is based on is not very accurate. The number of particles in Particle Filter can be increased to improve the accuracy, but there exists a trade off between computational effort and estimation accuracy. In our research, considering the same amount of computational complexity, we calculate the Mean Square Error (MSE) to examine the performances of different nonlinear estimation approaches. By comparing the MSE of different estimators, we prove that the Unscented Kalman Filter shows the most accurate performance in voltage synchronization for three phase unbalanced voltage. Our results have shown the potential applications of the nonlinear estimation techniques in the future smart power grid synchronization.