Functions Reference

These helper functions live in FUNCTIONS_IN_MODELS.f90 and are available inside any CODEGEN model without being declared.

---

The following functions can be used in any mathematical expression within CODEGEN model definitions.

equal

Signature: double precision function equal(var1, var2)

Arguments: double precision :: var1, var2

Description: Returns 1.0 if var1 differs from var2 by less than $10^{-6}$, and 0.0 otherwise.

Notes: General-purpose; available in all model types.

---

equalstr

Signature: double precision function equalstr(var1, var2)

Arguments: character :: var1, var2

Description: Returns 1.0 if the non-blank portions of var1 and var2 are identical strings, and 0.0 otherwise.

Notes: General-purpose; available in all model types.

---

ppower

Signature: double precision ppower([vx], [vy], [ix], [iy])

Description: Returns the active power injected into the network.

Model type: inj only.

Computation:

$P = v_x \, i_x + v_y \, i_y$

---

qpower

Signature: double precision qpower([vx], [vy], [ix], [iy])

Description: Returns the reactive power injected into the network.

Model type: inj only.

Computation:

$Q = v_y \, i_x - v_x \, i_y$

---

vrectif

Signature: double precision function vrectif([if], [vin], {kc})

Arguments: double precision :: kc

Description: Models the voltage drop in a rectifier. Returns the output voltage $vrectif$ as a function of output current $if$ and input voltage $vin$. Used in IEEE standard excitation models with $vrectif = f_{ex} \, E_{FD}$.

Model type: exc only.

Computation:

in = kc * if / max(vin, 1e-3)

if in <= 0:
  vrectif = vin
else if in <= 0.433:
  vrectif = vin - 0.577 * kc * if
else if in <= 0.75:
  vrectif = sqrt(0.75 * vin^2 - (kc * if)^2)
else if in <= 1.00:
  vrectif = 1.732 * (vin - kc * if)
else:
  vrectif = 0

---

vinrectif

Signature: double precision function vinrectif([if], [vrectif], {kc})

Arguments: double precision :: kc

Description: Inverse of vrectif. Given the field current $if$ and rectifier output voltage $vrectif$, returns the rectifier input voltage $vinrectif$. Intended for use at initialization. The computation is iterative because the operating segment of the piecewise characteristic is not known a priori. Does not work when $vrectif = 0$ (indeterminate input).

Model type: exc only.

Computation:

nbtries = 1
vinest = vrectif + 0.577 * kc * if

loop:
  in = kc * if / max(vinest, 1e-3)

  if in <= 0:
    vinrectif = vrectif
  else if in <= 0.433:
    vinrectif = vrectif + 0.577 * kc * if
  else if in <= 0.75:
    vinrectif = sqrt(vrectif^2 + (kc * if)^2) / 0.75
  else:
    vinrectif = (vrectif / 1.732) + kc * if

  if vinrectif == vinest or nbtries > 5:
    exit loop
  nbtries = nbtries + 1
  vinest = vinrectif

---

vcomp

Signature: double precision function vcomp([v], [p], [q], {Kv}, {Rc}, {Xc})

Arguments: double precision :: Kv, Rc, Xc

Description: Returns the magnitude of a combination of terminal voltage and current of a synchronous machine. Appears in IEEE standard excitation models.

Model type: exc only.

Computation:

$$\begin{aligned} V_{comp} &= \left| K_v \bar{V} + (R_c + j X_c) \bar{I} \right| \\ &= \left| K_v V + (R_c + j X_c)\!\left(\frac{P}{V} - j\frac{Q}{V}\right) \right| \\ &= \frac{1}{V} \sqrt{\left(K_v V^2 + R_c P + X_c Q\right)^2 + \left(X_c P - R_c Q\right)^2} \end{aligned}$$

---

satur

Signature: double precision function satur([ve], {ve1}, {se1}, {ve2}, {se2})

Arguments: double precision :: ve1, se1, ve2, se2

Description: Returns the increment of field current needed to obtain a given output voltage, accounting for magnetic saturation. Appears in IEEE excitation system models.

Model type: exc only.

Computation:

$satur = m \, ve^n$

where:

$$\begin{aligned} n &= \frac{\log_{10}(se1/se2)}{\log_{10}(ve1/ve2)} \\[4pt] m &= \frac{se1}{ve1^n} \end{aligned}$$

Exception. Returns $satur = 0$ if any of the following conditions holds:

$ve \le 0 \quad \text{or} \quad ve1 \le 0 \quad \text{or} \quad ve2 \le 0 \quad \text{or} \quad ve1 = ve2 \quad \text{or} \quad se1 = 0 \quad \text{or} \quad se2 = 0$

See Also

• Block Index, all CODEGEN blocks by category
• URAMSES, building models against the RAMSES source