Optimization of inertial micropower generators for human walking motion

Micropower generators, which have applications in distributed sensing, have previously been classified into architectures and analyzed for sinusoidal driving motions. However, under many practical operating conditions, the driving motion will not be sinusoidal. In this paper, we present a comparison of the simulated performance of optimized configurations of the different architectures using measured acceleration data from walking motion gathered from human subjects. The sensitivity of generator performance to variations in generator parameters is investigated, with a 20% change in generator parameters causing between a 3% and 80% drop in generator power output, depending upon generator architecture and operating condition. Based on the results of this investigation, microgenerator design guidelines are provided. The Coulomb-force parametric generator is the recommended architecture for generators with internal displacement amplitude limits of less than similar to 0.5 mm and the velocity-damped resonant generator is the recommended architecture when the internal displacement amplitude can exceed similar to 0.5 mm, depending upon the exact operating conditions. Maximum power densities for human powered motion vary between 8.7 and 2100 mu W/cm(3), depending upon generator size and the location of the body on which it is mounted.