Computational Drug Design is important as it reduces conventional research deadline and cost. Here we have considered structure based drug design where we have identified a protein target and performed optimization study to find a molecule or ligand that suitably fits in the target pocket. Different docking strategies are considered to make it more efficient with evolutionary algorithms. Our algorithm assumes a variable length tree structure which represents a drug molecule and arranges essential functional groups in different positions of that drug. Once the geometry of the drug is obtained, its docking energy is minimized. We have also considered several intermolecular forces for more accuracy. Results show that PSO performs better than the earlier methods. We have prepared a set of molecules having energy less than a threshold value. All these molecules are the potential drugs for a particular protein. These set of molecules will help the biologists or pharmacists to choose the best drug for a particular disease with very less effort. We have also incorporated different tool based study and do performance analysis of AutoDock & AutoDock Vina.