In this work, theoretical and experimental results on the properties of such multi-coil resonant WPT systems are presented. Theoretical results that generalize the well-known phenomenon of frequency-splitting to a system with multiple transmitters are analyzed. A simple adaptive approach to optimizing such a system and experimental results to illustrate the benefits of multi-coil resonance in the total power delivered are presented. A simple heuristic method of performing phase estimation from power-only measurements is discussed, and various closed-form solutions to optimization problems that are of interest when controlling WPT systems are derived.

This research specifically focuses on smart WPT systems that combine the use of multiple transmitting coils and the use of high-Q resonance to achieve enhanced range and performance. We consider an inductive WPT system where a number of transmitting coils seek to transfer the maximum possible amount of power to a receiver. This system consists of the transmitting and receiving coils which are designed to self-resonate with high-Q at a common frequency. We consider closed optimization for constraints that may be of interest when controlling WPT systems i.e., drive voltage limitations, loss minimization, and nullforming.

Lastly, an experimental setup that can automatically perform various optimization procedures is described in detail. The system implements receiver feedback to automatically estimate the current context of WPT system's geometry and circuit parameters. This WPT system is used as a demonstration of a multiple-input multiple-output system with resonant coils that can be controlled in real-time to precisely maximize or null power to selected receivers. We also show through heuristic methods that such a system can adaptively maximize power to a changing target.

While substantial literature exists on WPT systems using both multi-coil transmitters and coupled resonance, previous work has been limited to small-scale and simple transmitters. This limitation can be attributed to the difficulty of constructing models which adequately predict the behavior of these more complex physical systems. The ability to construct physically accurate models is needed in analyzing and optimizing such systems.

Our key novelty is to use receiver feedback to avoid the need for such models. This simple and powerful idea opens up the possibility of WPT systems consisting of a large number of resonant coils that may be diversely coupled with one another.

This thesis considers smart wireless power transfer (WPT) systems. This type of WPT system is defined by its ability to adapt to a wide range of arbitary operating conditions, and its ability to perform complex proceedures autonomously.