A new neural network for l1-norm programming

This paper presents a new neural network for solving l1-norm problems with equality and box constraints by introducing a new vector. The proposed model is proved to be Lyapunov stable and converges to an exact optimal solution of the original problem for every starting point. Compared with some existing continuous-time neural networks, the proposed model has the fewest neurons and a low complexity. The simulation results show the validity and transient behavior of the proposed model.