<< CLRseg SegPal DiracSync >>
SegPal >> SegPal > DerFil

DerFil

Filtered derivative with initial condition

Block Screenshot

Contents

Description

  • The DerFil block represents a filtered derivative. The time constant Td and the gain Kd of the derivative are programmable.
  • In automatic mode (Auto), the filter initialization at time t=0 is performed automatically so that the derivative is zero at the start of the simulation, avoiding any transient regime (see DerFilC.c)
  • In normal mode, the initialization condition at the filter's input value must be specified to achieve identical behavior.

    • With the remark :

    , the derived action can be canceled by programming Kd=0, if necessary.

    Parameters

    Alternative forms

    Interfacing function

  • SegPal\macros\DerFil.sci

    • Computational function

  • SegPal\src\c\DerFilC.c
  • SCI/modules/scicos_blocks/src/c/csslti4.c (Type 4)

    • Test programs

  • SegPal\demos\DerFil.zcos
  • SegPal\demos\DerFil_Auto.zcos

    • These two programs are identical but differ in their initialization condition: Normal/Auto

    Remark
  • The DerFil block is for example used in the REGxx models of the speed governors to represent the accelerator-tachometer derivative making it possible to obtain phase advance in the adjustment of the speed loop. Depending on the state of the circuit breaker, the initialization value is 0 if the group is stopped, or the nominal value if the simulation starts with the circuit breaker closed, unit connected to the network. This device allows the simulation to be launched smoothly, and at steady state.
  • See ATOMS module HydroSim (atomsInstall("HydroSim"))

    • The example produced previously shows three possible builds of the same derivative filter.

    The state space in matrix form (A, B, C, D) is calculated in the preamble to launching the simulation (Simulationīƒ¨ Set Context):

    The state space and the derivative filter DerFil produce an identical result, with, in particular, an initialization to 0 of the derived action.

    The standard CLR function of Scilab does not allow this result to be obtained, hence a non-zero derivative at t=0.

    Also see Components with Auto mode


    Report an issue
    << CLRseg SegPal DiracSync >>