commit 0a35c59d9cc510d0c508ba354a6d51e2093fe5bd
parent 09723dfd14831262aa7f8527b6ca56108090fcad
Author: Christophe Coustet <christophe.coustet@meso-star.com>
Date: Thu, 18 May 2017 17:23:47 +0200
Add man for simulation output.
Diffstat:
1 file changed, 58 insertions(+), 21 deletions(-)
diff --git a/doc/solstice-output.5.ronn b/doc/solstice-output.5.ronn
@@ -67,7 +67,7 @@ the same line while a closed quote mark is not defined.
<#samples> ::= INTEGER # in [0, INF)
<#failed> ::= INTEGER # in [0, INF)
- <global> ::= <irradiance>
+ <global> ::= <potential-irradiance>
<absorbed-irradiance>
<cos-factor>
<shadow-loss>
@@ -80,7 +80,7 @@ the same line while a closed quote mark is not defined.
<receivers-list> ::= <receiver>
[ <receiver> ... ]
- <receiver> solstioce::= "<receiver-name> <receiver-id> <area>
+ <receiver> ::= "<receiver-name> <receiver-id> <area>
<front> <back>"
<receiver-name> ::= <entity-identifier>
@@ -138,15 +138,16 @@ the same line while a closed quote mark is not defined.
<azimuth> ::= REAL # Degrees in [0, 360[
<elevation> ::= REAL # Degrees in [0, 90]
- <absorbed-irradiance> ::= <estimation>
- <absorptivity-loss> ::= <estimation>
- <cos-factor> ::= <estimation>
- <irradiance> ::= <estimation>
- <missing-loss> ::= <estimation>
- <reflectivity-loss> ::= <estimation>
- <shadow-loss> ::= <estimation>
+ <potential-irradiance>::= <estimate>
+ <absorbed-irradiance> ::= <estimate>
+ <absorptivity-loss> ::= <estimate>
+ <cos-factor> ::= <estimate>
+ <irradiance> ::= <estimate>
+ <missing-loss> ::= <estimate>
+ <reflectivity-loss> ::= <estimate>
+ <shadow-loss> ::= <estimate>
- <estimation> ::= <expected-value> <standard-error>
+ <estimate> ::= <expected-value> <standard-error>
<expected-value> ::= REAL
<standard-error> ::= REAL # in [0, INF)
@@ -154,30 +155,66 @@ the same line while a closed quote mark is not defined.
## SIMULATION
-A `simulation-output` begins by a line that stores the sun direction used by the
-simulation. The next line lists the numbers of global, per receiver and per
-primary results as well as the overall number of Monte-Carlo experiments used
-by the simulation and the number of experiments that failed due to unforeseen
-errors as numerical imprecisions.
+A `simulation-output` begins with two header lines. The first one reports the
+sun direction used in the simulation, and the second one lists the numbers of
+global, per receiver and per primary results as well as the overall number of
+Monte-Carlo experiments used by the simulation and the number of experiments
+that failed due to unforeseen errors as numerical imprecisions.
### Global results
+After the 2 header lines, the output includes various global-result lines, the
+exact number of lines being part of the headers. Currently this number is 7.
+Each global result is a pair of real numbers: the estimate of the value and
+its standard deviation. The global results are, in this order:
+
+* `potential irradiance`: the maximum irradiance that all the primary
+ geometries could intercept if properly oriented.
+* `absorbed irradiance`: the absorbed part of the irradiance hitting any
+ receiver geometry. At most equal to potential irradiance.
+* `cos factor`: the fraction of incoming irradiance not intercepted by the
+ primary geometries due to their orientation.
+* `shadow loss`: the amount of potential irradiance lost before hitting primary
+ geometries due to another geometry's shadow.
+* `missing loss`: the amount of potential irradiance having hit primary
+ geometries, but not absorbed by a receiver; this irradiance could have been
+ blocked along its path, can have missed the receivers, or can have hit a
+ receiver but without being absorbed.
+* `absorptivity loss`: the amount of potential irradiance that could have been
+ absorbed by receivers if atmospheric absorption was not taken into account.
+* `reflectivity loss`: the amount of potential irradiance that could have been
+ absorbed by receivers if reflections occured on materials with reflectivity
+ 1.0.
+
### Per receiver results
+After the global results, the output includes various per-receiver result
+lines, one line per receiver, the exact number of lines being part of the
+headers.
+
### Per primary results
+After the per-receiver results, the output includes various per-primary result
+lines, one line per primary geometry, the exact number of lines being part of
+the headers.
+
### Per receiver and per primary results
+After the per-primary results, the output includes various result lines, each
+describing the contribution of a primary geometry to a given receiver. The
+total number of such lines is the number of receivers times the number of
+primary geometries.
+
### Receiver map
A receiver defined in the submitted `solstice-receiver`(5) file, can have a
-per-primitive irradiance estimation if its `per_primitive` flag is active. In
+per-primitive irradiance estimate if its `per_primitive` flag is active. In
this case, `solstice`(1) generates a `receiver-map` that is actually an ASCII
VTK file [2] that stores the triangular mesh of the receiver and, for each
-triangle, the estimation of its associated irradiance. The "definition" of the
+triangle, the estimate of its associated irradiance. The "definition" of the
receiver map is thus controlled by the discretisation of the receiver's shape,
as described in the `solstice-input`(5) file. Note that to obtain a good
-estimation of the per-triangle irradiance, one have to ensure that the number
+estimate of the per-triangle irradiance, one have to ensure that the number
of per-triangle experiments is sufficient regarding the targeted accuracy.
Since only a small fraction of the overall sampled radiative paths reach a
given triangle, the total number of experiments required through the `-n`
@@ -185,9 +222,9 @@ option of `solstice`(1) should be increased significantly, as 1 or 2 order of
magnitude.
If only the front or the back side of the receiver is active, then only one set
-of per triangle irradiance estimation is written. If the `side` attribute of
-the receiver is set to `FRONT_AND_BACK`, the irradiance estimation for the
-front facing triangles are written before to the estimation of the back facing
+of per triangle irradiance estimate is written. If the `side` attribute of
+the receiver is set to `FRONT_AND_BACK`, the irradiance estimate for the
+front facing triangles are written before to the estimate of the back facing
ones. The following grammar gives a brief description of the formatting of a
`VTK-RECEIVER-MAP`. Please refer to the VTK format specification [2] for more
informations on the VTK file format.
