Investigating the Dynamics and Control of Electromechanical Networks with Semiresonant Latches

Regents Of The University Of Michigan - Ann Arbor, Ann Arbor MI

Investigators

Abstract

Many engineering technologies comprise networks of passive components, which are subjected to unpredictable dynamic loads. Examples include engineering structures and power-electronic circuits. For such technologies, a central design challenge is to control their dynamic behavior in the face of random dynamic loads or disturbances. Often, there are restrictions on the energy available for this purpose, and the focus of this project is to investigate the use of low-power controllable electromechanical switching devices called Semi-Resonant Latches (SRLs). Their fundamental advantage is that they harness the energy of external disturbances to dramatically alter the way energy flows through an electromechanical network, while requiring very little power to operate. Target applications include: (i) Vibration control of flexible structures in both civil and aerospace systems; (ii) Power transmission optimization in power-electronic networks for renewable energy technologies; (iii) Flexible truss structures capable of adapting their shape by capturing and re-directing disturbance energy; (iv) Structures that similarly adapt their stiffness properties through disturbance-driven pre-stress. This work will investigate energy-efficient techniques for effective control in such applications. Meanwhile, research results will be integrated into a rigorous new systems-oriented graduate curriculum in Civil and Environmental Engineering at the University of Michigan, as well as a summer outreach program focused on providing undergraduate research experiences to students from underrepresented demographics. SRLs have emerged independently in several disparate engineering applications. Electrical SRLs are built around transistors, while mechanical SRLs employ clutches. They have the ability to instigate near-instantaneous jumps in the voltages and currents of an electrical network, and in the forces and velocities of a mechanical network, in response to sensor feedback. Existing control techniques for these technologies are mostly ad hoc design approaches, and there is considerable ambiguity regarding the optimal use of these devices, as well as the limits on the performance they can achieve. The primary outcome of this work will be a general and abstract theory for optimal control of SRLs, which is applicable across the wide array of applications in which they are used, and which exploits recent results in hybrid control theory. A central focus will be to cultivate a solid mathematical foundation for analysis of general SRL systems in stochastic dynamic response. At present, such a theory only exists in very special cases. In addition to general theoretical contributions, concepts from this work will result in a fundamentally new class of high-performance engineering structure, capable of disturbance-driven shape adaptation, as well as stiffness adaptation.