Browsing by Author "Nerves, Allan C."
Now showing 1 - 1 of 1
Results Per Page
Sort Options
- Regenerative electric actuators for active control of civil structuresNerves, Allan C. (Virginia Tech, 1996)A novel technique is investigated for utilizing the motion caused by environmental forces on a civil structure to generate electrical energy when the structure's response is within safety limits. When strong winds and earthquakes occur, the utility power source which supply energy to the actuator in an active control system is usually not reliable. With a regenerative electric actuator, recovered energy can be used to reduce the peak oscillations of the structure by applying forces (through actuators which use the recovered energy) counter to the environmental forces even if the utility power is not available. The use of a regenerative electric actuator allows a precise control of the amount of damping being provided by the actuator. Another advantage of using regenerative electric actuators is the reduction of the required energy capacity of the electrical source. This translates into lower energy ratings for the electrical source, and lower equipment and maintenance costs. This study is the first of its kind to propose and investigate active control of civil structures using regenerative electric actuators. This study is also the first of its kind to investigate the applicability of sliding mode control to civil structures using regenerative electric actuators. Sliding mode control provides a natural synthesis of the on-off nature of pulse width modulation control and control force saturation, and guarantees stability for the control law. It is also invariant to parameter changes and external disturbances. New direct-control schemes for neural network and adaptive fuzzy control of civil structures using regenerative electric actuators are proposed and investigated. These allow on-line control of the structure without the need for either an accurate model of the system or a specific learning stage. Since the error at the output of these controllers will be unknown in the direct-control scheme, the error at the system output is used to train or update the controller parameters. Simulations are conducted for wind and earthquake excitations using linear and nonlinear models. It is shown that the use of regenerative electric actuators is a viable and a reliable alternative for active control of civil structures.