Abstarct: State estimation is the key operation to control and manage a power network. State variables of power system, including magnitude and angle of buses voltage of the network, represent a comprehensive knowledge across the network. On the other hand, inaccurate estimation of these variables may lead to wrong decisions and finally to heavy damages and even the destruction of the network. So, having a reliable and robust estimator is the necessary provision to possess a safe network. This research has aimed to increase the robustness of the estimator against falsifications or bad data, considering this fact that the information by measurements because of different reasons such as measuring errors, lack of simultaneous information acquisition, communication system errors, hacking of information in smart grids and …, are continuously exposing to be falsified. Firstly, the pre-filtering of measurements data is proposed using information processing techniques and applying Principle Component Analysis (PCA). Furthermore, in the case of adding a new measurement, optimized place with the aim of supporting the ability of detecting and identifying bad data, will be also proposed. Then, in order to estimate the static state and offline analysis of network, a robust estimator with using contraction map and Least Absolute Values (LAV) of measurements residuals have been represented. The results of simulation are indicating high precision and robustness of this method comparing to other static state estimation methods. Also regarding this fact that with having dynamic equations of a system, we can estimate and track the state trajectory of that system with a proper approximation, we can find a robust dynamic estimator which can be applicable as an online state estmation. At first, a pseudo dynamic model of power network is presented in this thesis. The proposed model is simpler than other dynamic models and also more efficient in large scale and wide power network modeling, because in this modeling there is no need for dynamic parameters of loads, generators and etc. Also it omitted dynamics with intervals less than one cycle in network which are not important in estimating the state. In second step, considering the extension and importance of Phasor Measurement Units, which provide precise and reliable information of voltage and current phasors of network, a method is represented in order to complete use of phasor measurements and conventional measurements and their information synchronization. Then a dynamic state estimator baxsed on Kalman filter which equipped with synchroniser data of conventional measurement and phasor measurements on the proposed pseudo dynamic model of the network has been introduced that estimate the states of the network with a proper speed and accuracy and track the network situation. Finally, with the aim of maximization posterior probability density function and applying Chapman-Kolmogorov relations and Bellman’s principal, a index to estimate the dynamic state is proposed which is simpler and without limitations that are challenges of Kalman filter application. In order to fulfill a precise and robust dynamic estimation, it is enough to optimize this criterion. In this research, this optimization is accomplished with applying Imperialist Competitive Algorithm, that with a high speed convergent, accurate outcomes are resulted.