An Empirical Model for a CO2 Thermal Compressor Based on Experimental Data

A thermal compressor uses thermal energy to increase the pressure of the working gas, while maintaining a
temperature difference, as an essential factor for its effective functioning. BoostHEA T's innovative thermal
compressor, driven by eco-friendly thermal energy, not only targets the replacement of traditional compressors in
various CO2 applications, but also aligns with the critical environmental motive to address global warming. The
design of the targeted thermal compressor is inspired by the gamma type Stirling engine, replacing the power piston
with inlet and outlet valves. In this paper, the thermal compressor (treated as a black box) is implemented in a heat
pump cycle, on which the tests were conducted. In the context of thermodynamic analysis, six principal inputs are
imposed to the compressor, and five outputs are measured. These physical variables are crucial in characterizing the
compressor, and evaluating its perfromance. An empirical model is developed where each output is represented as a
function of these inputs in a general mathematical form. The objective is to fine-tune these functions using machine
learning regression methods, based on the collected data. A sensitivity analysis is carried on each output with respect
to the inputs, in order to investigate the correlations between them and find the relevant inputs for each selected
output.