Azionamenti elettrici - Controllo di corrente

The transition of the transistors occurs when the error

changes its sign and exceeds 0 of a value Δ.

If output at the controller is a signal high (1), are controlled in

closing T1 and T3, and T2 and T4 in openings.

Vice versa if the signal is low (0), are controlled in closing T2

and T4, and in the opening T1 and T3.

Consider the case in which the current signal I is an ideal sinusoidal

o

and the load to be supplied both RL with a time constant very long.

Trend of the

sinusoidal current

into an RL load

Consider now the case in which you want to power the load RL with a

constant current I o Trend of the

constant current

into an RL load

CONTROL DIAGRAM – Three Phase Loads

Suppose we have a three-phase load predominantly inductive and with

the neutral connected directly to the central point of the battery

Fig.4

CONTROL DIAGRAM – Three Phase Loads

Fig.5 Representation of

bandwidth HB (i )

a

Limit of the vector Is*

Supposing to carry out the control, no longer in the reference system a,

b, c, but in the fixed system α, β, we introduce the space vector of the

current Is.

This vector also present limits of variation, which are presented in fig. 6

hexagon marked. Fig.6

Problem of this method

The control scheme in the Laplace domain is:

Block diagram of the Adjusting system with hysteresis in the d, q

δd and δq can be alone or > 0 or < 0; analysis of the sign

represents the method for choosing the voltage to be

generated.

Choice of configurations

The choice of the voltage vector to be generated is not

univocal; you choose the tension that involves the least

number of commutations.

With this type of control is never used the 'zero-configuration'

inverter.

To use it, the characteristic of the regulator hysteresis must

be:

In Field Oriented Control, the components i and i define

sd sq

the vector of the current I * that remains constant to regime,

s

in the reference d, q.

The current error ΔI is zero under ideal setting, ie when the

s

actual current vector Is equals the reference I .

sa

However, being an ideal condition is that: if we denote by

ΔI the maximum permissible error in the plan d, q, the I

s s

max

vector must move in a circle with center in I * and radius

s

equal to ΔI .

smax

Limit of I in d,q axes

s Each time that I meets the circle,

s

is changed status of the inverter in

order to generate a voltage vector

V to bring back inside the

sk

perimeter of limit.

Since more than one solution

satisfies this condition, you choose

that vector that maximizes the time

interval for which there is the

occurrence of another intersection,

in other words it tends to adopt the

solution that minimizes the

switching frequency.

Simplified circuits Fig.7

by placing