search

triangleenergycohort / remote_hub_simulation

0 stars 0 forks source link

Create energy forecast model #1

Closed dvstorel closed 5 years ago

dvstorel commented 5 years ago

Steps:

  1. mock up input files for customer usage data (monthly)
  2. research and develop weather monitoring capability
  3. determine algorithm for recommending customer curtailment
  4. determine output format and syntax
  5. create statistical model to determine confidence of prediction
  6. determine best strategy for GUI output
dvstorel commented 5 years ago
  1. complete 1/16/19
dvstorel commented 5 years ago

cc @jamcdon3

  1. complete 1/18/19 see /forecast_model/weather data/api_interface.py for API call and conversion to csv
dvstorel commented 5 years ago

cc @vmuthuk2

Simple suggestion for 3: increase minimum daily SOC threshold to 75% when forecast suggests two or more days of low insolation. Low insolation is defined as less than 120 Wh/day (50% of average)

dvstorel commented 5 years ago

Point 5 can be accomplished using data provided in the Solcast API

dvstorel commented 5 years ago

cc @vmuthuk2

Simple suggestion for 3: increase minimum daily SOC threshold to 75% when forecast suggests two or more days of low insolation. Low insolation is defined as less than 120 Wh/day (50% of average)

Alternative approach: maximum incremental DOD of 20% with a minimum of complete discharge (~40%). Assuming a full charge initially, this would provide 48 Wh/day for three consecutive days.

dvstorel commented 5 years ago
  1. done 1/22/19. Link to working document: https://docs.google.com/document/d/15kFwMbmWXZX5pprTtKFK808XIIsLhCl8vpvjcQR1fh4/edit?usp=sharing
dvstorel commented 5 years ago

1-5 done 1/22/19. A new ticket will be opened for 6.