Most civil structures are passive and static. They are designed according to serviceability and safety requirements. When subjected to actions, they generally deflect in a passive manner. For example, they do not adapt actively to external loading. A more challenging functionality of civil structures is active adaptation to changing requirements, such as load modification, temperature variation, support settlements and eventually damage. This thesis aims to make progress in the field of adaptive civil structures. Intelligent control methodologies such as self diagnosis, multi-objective shape control, self repair and reinforcement learning are proposed and verified experimentally. Although these methodologies are validated experimentally only on a five module active tensegrity structure, there is potential for extension to more complex civil structures. The most important conclusions of this thesis are synthesized below: Self diagnosis can extend active structural control of tensegrity structures to situations where there may be partially defined loading events and damage. Multi-objective search is attractive for computing commands that control shape of an active tensegrity structure while maintaining robustness of both the structure and the active control system. These commands are particularly useful to control structures over scenarios of multiple loading events. Self repair of a damaged active tensegrity structure is possible. The active control system can be used to apply control commands that increase stiffness and decrease stresses within an active tensegrity structure that is damaged. Reinforcement learning is attractive for improving the control of an active tensegrity structure using previous loading events. Improvements involve lower command computation times and better control quality. Such interactions between learning algorithms and active control systems are attractive for control tasks. The integration of intelligent control methodologies such as self diagnosis, multi-objective shape control, self repair and reinforcement learning to an active tensegrity structure create an example of an adaptive civil structure that can be applied to a range of practical situations in the future.