Linear interpolation
Linear Interpolation resolves small gaps by creating estimated interval values between two valid interval reads. This routine creates readings that fill the gap with no peaks or valleys.
Linear Interpolation operates on a group of intervals that does not exceed a user-defined number of hours and minutes. For example, the defined limit is 3 hours. There are two gaps in the data: Gap A, from 8:00 to 10:00, and Gap B, from 13:15 to 17:30. The linear interpolation routine can fill Gap A. The routine cannot fill Gap B, because Gap B exceeds the maximum number of 3 hours.
The routine uses the two valid intervals on either side of the gap as reference points and draws a straight line between them. The routine marks estimated intervals with status flags of ESTLINEARINT (LIN) and ESTIMATED (EST).
|
Date Time |
Actual Read |
Status After Validation |
Read After Estimation |
Status After Estimation |
|---|---|---|---|---|
|
1/1/2012 00:15 |
75 |
PV |
75 |
PV |
|
1/1/2012 00:30 |
0 |
MI,ESN |
80 |
LIN,EST |
|
1/1/2012 00:45 |
0 |
MI,ESN |
85 |
LIN,EST |
|
1/1/2012 01:00 |
0 |
MI,ESN |
90 |
LIN,EST |
|
1/1/2012 01:15 |
0 |
MI,ESN |
95 |
LIN,EST |
|
1/1/2012 01:30 |
100 |
PV |
100 |
PV |
|
1/1/2012 01:45 |
104 |
PV |
104 |
PV |
|
1/1/2012 00:15 |
75 |
PV |
75 |
PV |
|
1/1/2012 00:30 |
0 |
MI,ESN |
80 |
LIN,EST |
|
1/1/2012 00:45 |
0 |
MI,ESN |
85 |
LIN,EST |
Linear interpolation estimates data to fill gaps found at the very beginning or very end of the day. Depending on if readings are processed through AMI Readings Import (ARI) or a data collection window (DCW) process, or if they are processed through standard readings import (SRI), which is a non-DCW process, linear interpolation uses a different method to estimate and fill gaps.
For gaps in readings that are processed through SRI, linear interpolation fetches the next good interval from the database, and fills the gap using that interval as a reference point.
For example, if the first two reads of a reading group are missing when data is processed through SRI, then linear interpolation fetches from the database the closest good interval before the beginning of the reading group. To estimate the missing intervals, it draws a straight line between the good interval that was fetched from the database, and the first good interval after the missing two intervals in the reading group. Similarly, if the last two intervals of the reading group are missing, linear interpolation looks to the database for the first good read after the reading group.
Because ARI supports high-volume readings import, returning to the database to identify reference intervals is not efficient. For readings that are processed through ARI, linear interpolation uses flat-line approximation to fill gaps found at the very beginning or very end of the reading group. During flat-line approximation, linear interpolation locates and copies the closest good interval from within the same reading group, and fills the gap with that interval value.
IEE administrators can use the During Data Collection Window, Flat-Line Interpolate parameter in the linear interpolation routine on the estimation set to configure this behavior.
For more information about configuring this estimation routine, see Configuring linear interpolation estimation routine.