forecasting:meeting_minutes_september_13_2016

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
forecasting:meeting_minutes_september_13_2016 [2016/09/14 02:49]
jobatake
forecasting:meeting_minutes_september_13_2016 [2021/09/19 21:59] (current)
Line 9: Line 9:
  
 Tuesday, September 13, 2016 Tuesday, September 13, 2016
 +  * Produce:
 +    * Well-written code
 +    * New python libraries to pull off of
 +    * Build library of functions other groups can use
 +    * Solar Industry
 +    * Calibrate Data
 +  * Decisions:
 +    * Normalization Method (Zenith Angle?)
 +    * Prediction time (1 hr?)
 +    * Solar Irradiance doesn'​t need to be sampled too quickly (1-2 mins ok)
   * Understanding PV:   * Understanding PV:
     * How much in Hawaii     * How much in Hawaii
       * Commercial and residential       * Commercial and residential
       * Projection of these       * Projection of these
-  ​* Solar Thermal +    ​* Solar Thermal 
-  * Solar Farms +    * Solar Farms 
-  * Concentrated Solar +    * Concentrated Solar 
-  * Energy Production +    * Energy Production 
-    * PV Data Sheets +      * PV Data Sheets 
-    * 10-15% efficiency w/ lower cost panels +      * 10-15% efficiency w/ lower cost panels 
-    * 40% efficiency w/ increase in efficiencies,​ cost, etc. +      * 40% efficiency w/ increase in efficiencies,​ cost, etc. 
-  * Breaks for Federal and State and HECO +    * Breaks for Federal and State and HECO 
-    * Larger contracts (PPA) +      * Larger contracts (PPA)
   * Forecasting:​   * Forecasting:​
     * Weather and time series prediction     * Weather and time series prediction
Line 29: Line 38:
       * Depends on normalization (ex. subtract mean of specific time of day/std, or zenith angle)       * Depends on normalization (ex. subtract mean of specific time of day/std, or zenith angle)
       * Take **zenith angle**: Time of day and day of year in account       * Take **zenith angle**: Time of day and day of year in account
-        ​Depends on first and second order (mean and covariance matrixes)+    ​Window ​(FIR filters) 
 +      * Straight average of data over time   
   * Online Methods:   * Online Methods:
     * Estimate of parameter weight and update estimate based on new data     * Estimate of parameter weight and update estimate based on new data
     * New linear estimate     * New linear estimate
     * Recursive Least Squares (RLS)     * Recursive Least Squares (RLS)
 +      * Depends on first and second order (mean and covariance matrixes)
     * Least Mean Squares (LMS)     * Least Mean Squares (LMS)
       * Won't perform as well       * Won't perform as well
- +  * Supervised and Unsupervised Learning: 
-Thursday, September 152016 +    * Nearest Neighbor (supervised) 
-  * +    * Support Vector Machine Kernels () 
 +    * Deep Learning () 
 +    * Clustering used for large dimensional data (unsupervisedLots of factors) 
 +  * Estimation (regression) and Detection (classification)
  
  
 ===== Reminders ===== ===== Reminders =====
-  * +  * Weekly REIS Thursday Seminar 4:30-5:30
  
  • forecasting/meeting_minutes_september_13_2016.1473821370.txt.gz
  • Last modified: 2021/09/19 21:59
  • (external edit)