In this paper, we give a smoothing approximation to the lower order exact penalty functions for inequality-constrained optimization problems. Error estimations are obtained among the optimal objective function values of the smoothed penalty problem, of the nonsmooth penalty problem and of the original optimization problem. An algorithm based on the smoothed penalty function is presented, which ...