Model predictive control (MPC), which works on the basis of receding horizon control has attracted the process control community due to its capability to grasp constraints on process variables, nonlinearities plus interactions among process variables, disturbances etc. An effective application of Model Predictive Control using Maglev system model is predicted by neural network presented in this thesis. The two concerns in model predictive controller are prediction and optimization of cost function. Process model plays a key role in model predictive controllers. The more accurate the model the more accurate is the controller. State space models, First Principle models, Hammerstein models, Volterra models etc were used widely to develop accurate dynamics of nonlinear processes which are effort demanding and time consuming. Then artificial neural network (ANN) models curved the concentration of MPC users suitable to their capacitytowardabsolutelydistinguishintricate nonlinear relations between needy and self-determining variables with less effort. Several researchers have approximated nonlinear models by neural networks besides its lengthy training time and requirement of large training data The other concern in model predictive controller is computational cost, as it does prediction as well as optimization oneverysample of time. Then main cost in the usage of non-linear programming as the optimization algorithm is during the estimate of the Hessian matrix (second derivatives) and its inverse which is difficult and costly. There is a class of evolutionary algorithms, which are derivative free techniques that uses some tools motivated by biological evolution. In this thesis an evolutionary algorithm, particle swarm optimization (PSO) is used to decipher the optimization problem related with the NN based model predictive control And simulation results explainunion to a first-class solution contained by minimum number of iterations and hence real time control of fast-sampling system like the Maglev system is possible.