Tensegrity structures are spatial, reticulate and lightweight systems composed of struts and cables. Stability is provided by a self-stress state between tensioned and compressed elements. Tensegrities have received interest among scientists and engineers in fields such as architecture, civil and aerospace engineering. Flexibility and ease of tuning make these systems attractive for controllable and adaptive structures. However, tensegrities are often prone to difficulties associated with meeting serviceability criteria and with providing adequate damage tolerance when used as civil engineering structures. This paper extends research on active control of tensegrity structures to study self-repair of a tensegrity pedestrian bridge that is damaged. Self-repair is intended to meet safety and serviceability requirements in case of cable damage in the pedestrian bridge. Intelligent control methodologies that implement stochastic search with active member grouping are proposed. Case studies for several damage scenarios are presented to show the effectiveness of the methodology. Results from simulated damage scenarios show that self-repair can be successfully performed with a minimum number of active members leading to a reduction in control complexity.