Seismic evaluation and retrofitting of existing structures areundoubtedly of most important. Too many buildings in Egypt and allover. the world were built using previous codes with no consideration ofearthquake forces. Evaluating the seismic vulnerability of such buildingsis thus very necessary to determine their earthquake resistance capacity.For buildings having insufficient seismic resistance, a decision aboutrehabilitation is to be made to mitigate seismic hazards and save livesduring future earthquakes. Deriving more evaluation procedures forseismic evaluation of buildings and developing policies for selecting themost suitable method of upgrading are thus major challenges.The present study suggests a complete, robust and integratedThe major disadvantage of the three control schemes is using torque transducer to detect the new value of load torque that has mechanical problems. In order to overcome this problem the three control schemes can operate without torque transducer. In order to detect the new value of load torque a neural network can be trained to give that new value of load torque according to the variation of parameters under the new conditions of operation such as armature current or input power to be an indicator for the new load torque. However this technique make the network to be more complicated and more expensive rather than using torque transducer.