Multi-source Harvesting Systems for Electric Energy Generation on Smart Hip Prostheses - Biomedical Engineering Systems and Technologies - page 70

Fig. 10. Experimental average power and peak-to-peak voltage results harvested from the PEH

Fig. 11. Validation of the average power harvested from the PEH generators (dash-dot line refers
to the network output)

4.3 PEH Results

External sinusoidal forces, with amplitudes in the range 100 Nto 250 N and frequencies
in the range 0 . 5 Hz to 4 Hz, were applied to the generator. A load of 1 M Ω was used
to enable the energy transfer. Figure 10 reports the experimental average power and
peak-to-peak voltage results, whereas Fig. 11 introduces the validation results of the
“feed-forward” ANN, using only data not used in the training process. Tables 9 to 10
report the modulation errors. The maximum energy harvested was 0 . 6 μ J/s.

5 Discussion

5.1 Need for Multi-source Harvester Systems

The high reliability of the electric energy generation system is a necessary condition
to ensure high reliability of the therapy based on mechanical stimulation. It is tech-
nologically possible to implement active implants with the ability to monitor failures,
to communicate the states of its surrounding physiological environment to the medi-
cal specialist and carry out mechanical-based therapeutic prescriptions ordered by the
specialist. These operations demand a full availability of electric energy.