01. Training For New Users » 1_02 Simplify Tutorial of Python

This software "LinearLoadEstimation_v1" is the first version of an algorithm for 

parameter estimation using trajectory sensitivity method. The model is linearized 

to use perturbance measurements provoked by a under load tap changer of 

transformer (ULTC).All files are written in Python 2.

The main routine is written on the 'LinearLoadEstimation_v1.py' file. In this 

code, the user will find the set of parameters that represent the real behaviour 

of the load used for the study, the set of initial guesses for the parameters 

and some method settings, such as tolerance, time step and parameters' variation.

The equations describing the model's output behaviour can be found on the 

'Matrix.py' file. The file was separated from the main routine to allow the use 

of the algorithm to identify parameters of different models, not only load models. 

The output behaviour is simulated using 'rk4.py', where the Runge-Kutta Method 

is implemented.

'Classification.py' ranks the set of parameters from ill to well-conditioned. 

This rank will be later used to identify the parameters in two steps, first 

the ill-conditioned and later the whole set. The function responsible to calculate 

the error between the real and modelled outputs via the Least-Square Method, is 

written on 'Error.py'. The code was written separately for the same as reason as 

'Matrix.py', allowing the user to choose the error function. 'Gamma.py' gives us 

the error sensibility and the Hessian Matrix and calculates the parameters 

increment to minimize the error.

In order to estimate the set of parameters of a different model all the user must 

do is change the set of parameters, the initial guesses and the settings on the 

'LinearLoadEstimation_v1.py' and the matrices on 'Matrix.py'.