Predicting Optimal Temperature in The Transmission System
summary
Proof of technology shows two machine learning models: one forecasts optimal supply temperature 12 hours ahead (trained on under 10,000 representative timestamps) and the other predicts hydraulic limitations 24 hours ahead to prevent under-delivery. Collaboration with domain experts yields a rule-based label for “optimal,” and a 14-day spring test period indicates potential reduction of 5.4 degrees in supply temperature. Work runs from a 30-day AI Camp and targets safer, greener district heating with two models.
AI-written summary · about
chapters (10)
- 0:00Intro
- 2:56User story
- 4:30Why?
- 6:20What is "district heating"?
- 9:20The challenge with district heating
- 14:37The challenge with machine learning
- 19:16ML model 1: Predicting optimal supply temperature
- 22:04ML model 2: Predicting hydraulic limitations
- 23:48Key takeaways
- 30:20Outro
related talks
12mReducing Carbon Footprint in Smart Buildings with Data Science
17mAI Assisted Load Sharing for Improved Energy Efficiency
18mPredictive Network: How AI Forecasts System Failures
31mAI-supported transformation of the urban infrastructure
40mDeep Learning and Time Series Forecasting for Smarter Energy
31mKeeping up with the renewable energy transition
23mDataEngBytes 2023 - MEL-T3-01 - Rajath Vanikul
35mIs it a predictive model? Is it causal inference? Well... It is running a greenhouse.