In this paper an evolutionary algorithm, which can be used for non-linear optimization problems, is introduced. The non-linear system taken up for the study is conical tank system because it is used in the field of medical sciences. Its non-linearity is due to the cross sectional behaviour of the process for the evacuation of products without wastage is possible. To control the level for the tank, the closed loop performances are determined with the conventional PI controllers. It is optimized by an Optimization algorithm which is motivated by the egg laying activities of bird cuckoo. The major advantage of this method is simplicity and used as a simple optimization algorithm. Application of the projected algorithm to some standard functions and a real problem has confirmed its capability to deal with difficult optimization problems. Simulation results show that the proposed the proportional integral controller outperforms with its optimization.