← Back to Portfolio

04 — Hydraulic Regime Detection

Differential pressure analysis and Gaussian Mixture Model clustering to identify two distinct operational regimes.

chiller-plant clustering gmm hydraulics regime-detection
Pythonscikit-learnmatplotlib

Key Findings

  • Bimodal delta-P distribution reveals two hydraulic configurations
  • GMM clustering identifies 422 hours in regime 0 vs 922 hours in regime 1
  • Q² vs delta-P scatter confirms two distinct flow-pressure relationships

Investigates hydraulic operating regimes using differential pressure histograms, hydraulic coefficient analysis (k_hyd = ΔP/Q²), and unsupervised Gaussian Mixture Model clustering on pressure and fan power features.

Notebook