Abstract: The neural network ensembles based on lateral inhibition of cortical columns are built to simulate the neuron firing phenomenon in cortical columns and to be used for position tracking.The neural network ensembles are composed of many Stein's neurons.Lateral inhibition of cortical columns is applied to the coding and decoding of neural network ensembles to realize the negatively correlated firing,and the advantages of the negatively correlated firing are proved theoretically.Simulation results show that the application of lateral inhibition with cortical columns can greatly improve the position tracking accuracy and system stability.