Abstract
Wave energy is an enormous untapped potential for an alternative form of energy due to its reliability and consistency. To date, there have been several proposed designs of wave energy converters, with pendulum based designs as a prominent option. These devices take advantage of the relative motion between an inner oscillating mass and an outer body to harvest electricity. While normal pendulum configurations are frequently studied in literature, seldom have investigated the inverted pendulum configuration, where the center of mass is inverted above the axis of rotation. The inverted configuration is of particular interest because preliminary studies suggest the configuration could better match the impedance of the outer body, leading to higher power performance.
The research herein describes the theory and validation of a numerical model for an inverted pendulum based wave energy converter to power oceanographic buoys. Governing equations of motion are derived to obtain linear and nonlinear models of this configuration. The parameters of the numerical models are fine tuned through system identification by comparing against experimental data from a bench test. To assess the bounds for application of the linear model, a comparative analysis was conducted with varying frequencies and amplitudes as excitation input to evaluate root mean square error between the nonlinear and linear model. Additional metrics, such as total harmonic distortion, were used to quantify the nonlinearities captured in the nonlinear model. Results showed significant nonlinearities at near-zero frequencies and amplitudes, suggesting that static and Coulomb friction play a pronounced role during initial transients of the device. The linear model generally approximated the nonlinear model well, but exhibited a lower degree of accuracy at high amplitudes and frequencies near resonance.
This work provides valuable insight into the dynamics of wave energy converters that allow for design optimization and enhanced power performance. In quantifying conditions where the linear model sufficiently approximates the nonlinear model, design iterations can apply the linear model to reduce computational complexity and improve efficiency. Future work will involve improving the inverted pendulum model and investigating additional nonlinearities that may arise in a full-scale model.