discrepancies happen in practical engineering
application, leading to the necessity of elevation
fitting.
Secondly, orthometric system is one of the most
precise measurement systems that uses practical
horizontal surface as the reference for measurement.
Although this system adopts practical measurement
standard, there are still some difficulties in the
application process. In fact, GPS elevation fitting is
to fit the data gained from orthometric system and
geodetic height system, thus obtaining the most
accurate elevation data.
Finally, normal height system is actually a
compromised system out of geodetic height system
and orthometric system. As discussed above,
geodetic height system gets some theoretical
elevation data, so those data shows a low application
value to some extent. Meanwhile, orthometric
system adopts gravity measuring method, which
causes some difficulty in practical operation.
Besides, orthometric system results in too much
work of measurement. Therefore, normal height
system adopts a compromised method based on the
problems of the other two systems. Normal height
system makes use of average value for
measurement, thus ensuring the reliability of data
and largely reducing some complex and redundant
measuring steps in actual measurement. In
conclusion, normal height system is most commonly
used in elevation measurement.
3.3 Research of GPS Elevation Fitting
Above analysis shows that GPS elevation fitting has
obvious advantages including high efficiency and
precision. Besides, GPS elevation fitting can achieve
no cumulative errors of elevation measurement.
Thus, GPS elevation fitting has a very important
significance in practical application. In the
following, the work discussed how GPS elevation
fitting works out in practical application and how
GPS elevation fitting technology is achieved.
First of all, isoline map is used in GPS elevation
fitting. Isoline is a very common method for
measuring and fitting. In fact, isoline is a method to
obtain average value. So, the measurement data from
GPS can be used to figure out several contour lines
of some normal heights, and then complete the
drawing of all contour lines based on the calculation.
Here, the concept of contour line is consistent with
the concept of isoline.
Secondly, curve fitting method is also used in
GPS fitting. In fact, curve fitting method is the most
mainstream calculation method out of all
mathematical methods for GPS elevation fitting.
This method carries out corresponding data
calculation through one curve, thus geting relative
average value. Curve fitting is a relatively
commonly used GPS elevation fitting method.
Generally, accurate method is needed in curve fitting
to ensure the accuracy after the fitting.
Thirdly, multiplication curve fitting is another
method applied in GPS fitting. Based on the
extension way and advantages of curve fitting, GPS
fitting makes use of multiplication curve fitting to
maximumly reduce fitting errors. In fact, any kind of
calculation contains the risks of errors, and the
differentials of the errors to some extent determine
the accuracy of the measurement. Multiplication
curve fitting method mainly focuses on making use
of the advantages of curve fitting. Individual curve
fitting still causes some errors, so multiplication
curve fitting method is applied to minimize errors.
Therefore, multiplication curve fitting method
becomes the most commonly used method in GPS
elevation fitting.
4 LEAST SQUARE ALGORITHM
AND LEAST-SQUARE
COLLOCATION
Least square algorithm is acutally the most
commonly used data fitting method in data
optimization. Its main role is to reduce calculation
errors and problems happened in data calculation
process. In practical application, least square
algorithm usually appears in the form of least-square
collocation, because it is widely used in practical
application especially in high-precision data
calculation. Error is inevitable in many calculations.
Using least square algorithm to avoid errors is
actually to figure out the most suitable ratio function
based on the square or differential of errors and then
conduct corresponding data calculation. In addition,
least square algorithm can be easily figured out for
unknown data calculation, greatly minimize errors in
application. This is why least square algorithm is so
commonly used and why least-square collocation
method achieves so obvious achievements on
minimizing errors.
Least-Square Collocation and Its Application in GPS Elevation Fitting
81