Reduced-order modal-domain structural control for seismic vibration control over wireless sensor networks

Semi-active control devices are promising technologies to provide economical and reliable protection for the safety of civil infrastructure assets as well as preserving the comfort of their occupants due to their low costs and power requirements. However, these devices can be limited in their capacity to provide large control forces necessitating installations in large numbers, creating additional expense and vulnerabilities to the system in the form of extensive lengths of signal cables. Wireless sensor networks are gaining popularity as a means of collecting, coordinating, and processing data from spatially distributed locations in civil structures. Their inherent computational abilities can also be harnessed to command networks of structural control actuators, however significant issues exist with regard to communication delays and computational power that must be addressed. This study demonstrates the successful application of a modal-domain state-space control algorithm for use in wireless structural control of a six-story shear building with magneto-rheological actuators. Control laws and state estimators are derived in modal coordinates for models of increasing size. Increasing the complexity of the underlying model yields a more optimal control law; however due to the limitations of the wireless actuator nodes in computational power, this increase in complexity comes at the expense of significantly increased latency which degrades performance. The trade-off between speed and model order is explored in terms of control performance. Both simulation and experimental results are presented.