Over the last decade, the energy optimization of
resource constrained sensor nodes constitutes a major research
topic in smart environments. However, state of the art energy
optimization algorithms make strong and unrealistic assumptions
of energy models, both in simulations and during the operation of
smart systems. For instance, simplistic energy models for energy
harvesting leads to inaccurate representation and prediction of
the true dynamics of energy. Consequently, systems for smart
environments are unable to meet expected performance criteria.
In this paper, we propose innovative models to overcome the
drawbacks of simplistic energy representations in smart environments.
We provide the insights of how to generate precise
lightweight energy models. Using the physical properties of solar
and flow energy harvesting as case studies, the trade-off between
energy harvesting inference and real-time measurement of energy
generation is explored. To evaluate our proposed energy models
against the simplistic versions, we use real measured data from
our environmental micro-climate monitoring deployment in an
urban park and a 103% improvement is seen. Additionally,
to define the trade-offs between inferred and measured energy
generation, experiments are conducted utilizing solar and smart
water testbeds.