Improving the fault tolerance of multi-robot networks through a combined control law strategy