Introduction to the Software

GRAPE software has been created assembling twenty one different subroutines organised by one main routine. It has been designed to accept any number of NXF files, one for each antenna used to collect the GPS data. One of the antenna positions will be chosen as the common origin for the satellites and the rest of the antennas.

The program structure consists of many subroutines, in descending levels from the main program level. This made any modifications (such as the validation and rejection criteria as implemented in Chapter Seven) easier to perform as further subroutines could be added as required. The flowcharts of the steps involved within the GRAPE software are presented in Figures 4.4, 4.5 and 4.6. Figure 4.4 shows the flowchart of the main working principle of the GRAPE software whereas Figures 4.5 and 4.6 are the flowcharts of the subroutines used within the main flowchart, i.e. Figure 4.5 is the ambiguity ftmction routine and Figure 4.6 is the least squares routine.

C hapter Four: P rocessing S tra teg y a n d Softw are D esign o f G R A PE

Num ber of a n te n n as >= 3?

O pen NXF files for all a n te n n as R ead initialisation file

Display input information

If G P S time < start time s e t G P S time = start time

R ead first ep o ch from e ac h NXF file an d wind

to the sa m e G P S time

C reate output file

S ynchronise files & so rt out com m on healthy satellites

C om pute SV local level coordinates at all a n te n n as

-Yes N um ber of common

satellites >= 4 ?

Time += interval k -

C h o o se the highest SV

S ort out all NXF files so that they point to th e comm on satellites

B a se d on an attitude increm ent, construct a search in g s p a c e around

th e initial attitude

C om pute the trial attitudes

C om pute AFV from all LI & w eighted 12 d ata for every trial attitude and store attitudes with AFV>90% of max. AFV a s

can d id ate attitudes (refer to Figure 4.5)

Com pute residuals & perform F-test

Solve for corrections to the can d id ate attitudes by least s q u a re s com putation using all L1

& L2 p h a s e d ata and com pute th e final candidate attitudes

(refer to Figure 4.6)

C h o o se attitude with the minimum residuals a s

th e correct solution

Distinguish the correct attitude?

Write results

to output file More ep o ch ?

Chapter Four: P rocessing S tra teg y a n d S oftw are D esig n o f G R A PE

B ased on e ac h trial attitude, com pute th e nine direction cosine elem en ts

C om pute AFV from all L1 & w eighted L2 d ata for every trial attitude

S to re attitudes with AFV>90% of max. AFV

a s can d id ate attitudes C om pute th e observed

L1 & L2 double difference using the

highest SV a s the differencing satellite B ased on th e nine direction

cosine ele m en ts and the body fram e co ordinates of th e an te n n as, com pute the local level coordinates

of th e a n te n n as

B ased on th e SV and the a n te n n a s local level coordinates, com pute the

modified (com puted) L1 & L2 double difference using th e highest

SV a s th e differencing satellite

C hapter Four: P rocessing S tra teg y a n d Softw are D esign o f G R A PE

A ssign full weight matrix

C om pute the final can d id ate attitudes

C om pute th e partial derivatives of the can d id ate attitudes

(i.e. A matrix) B ased on e a c h candidate attitude, co m p u te th e nine direction c o sin e ele m en ts

Solve for corrections to th e c an d id ate attitudes using th e le a st s q u a re s

com putation C om pute th e o bserved

L1 & L2 double difference using the

highest SV a s the differencing satellite

B a se d on th e nine direction c o sin e e le m e n ts and the body fram e c o o rd in ates of th e a n te n n a s , co m p u te the local level co o rd in ates

of th e a n te n n a s

B ased on th e SV and the a n te n n a s local level coordinates, com pute the

modified (com puted) LI & L2 double difference using th e highest

SV a s th e differencing satellite

C hapter Four: P rocessing S trategy a n d Softw are D esign o f G R A PE