Existing nonlinear guidance methods are difficult to reconcile performance optimality with stability assurance. This study proposes a concept of robust incremental learning for approximate optimal ...