scieee AI-readable full text Open interactive document viewer

Derivation of the VOC method

Arghir, Catalin

Abstract

In this document, the idea behind the VOC method was first presented.

Full text

Sketch of an idea Catalin Arghir November 4, 2014 Suppose we start with the following model of a 3-phase inverter in α-βframe: ˙x1=u1−w1 ˙x2=u2−w2(1) Where xrepresents the output capacitor voltages, uthe converter side currents charging the two 1 Farad capacitors and wthe grid side currents. We plan to stabilize the set: γ=x:px2 1+x2 2=ρ0 We use the following coordinate transformation: r=px2 1+x2 2 φ= arctan(x2 x1)(2) By taking the time derivatives of (2) we get: ˙r=1 2 2x1˙x1+2x2˙x2 √x2 1+x2 2 =x1(u1−w1)+x2(u2−w2) r=1 rx1x2u1 u2−1 rp ˙ φ=1 1+ x2 2 x2 1 x1˙x2−x2˙x1 x2 1 =x1(u2−w2)−x2(u1−w1) r2=1 r2−x2x1u1 u2−1 r2q (3) Where p and q are the instantaneous active and reactive powers: p=x1w1+x2w2 q=x1w2−x2w1(4) We rewrite eq. (3) as: ˙r ˙ φ=1 r2x1r x2r −x2x1u1 u2−1 rp 1 r2q(5) We use the following feedback transformation (invoking a transverse feedback linearization argument or simply to resemble a harmonic oscillator): u1 u2=x1−x2 x2x1b ω(6) Eq. (5) becomes: 1 ˙r ˙ φ=1 r2x1r x2r −x2x1x1−x2 x2x1b ω−1 rp 1 r2q=1 r2(x2 1+x2 2)r0 0 (x2 1+x2 2)b ω−1 rp 1 r2q We end up with the following system: ˙r ˙ φ=r0 0 1b ω−1 rp 1 r2q(7) By letting ˙ φ=ω+˙ θand setting ω=const., we get the following phase dynamics: ˙ θ=− 1 r2q By letting b=b01−r2 ρ2 0, where ρ0is the radius of the circle-set to be stabilized, we ge the following amplitude dynamics: ˙r=b0r1−r2 ρ2 0− 1 rp In fact one could pick any function b(r) that is zero at ρ0, strictly positive for r < ρ0and strictly negative for r > ρ0as b is the real part of the complex poles of the harmonic oscillator: ˙x1 ˙x2=b−ω ω b x1 x2(8) The sketch shows that if instead of starting with a Lienard oscillator, one starts with a 2 state integrator, by applying a certain coordinate and feedback transformation one is able to highlight feedback that turns a given set into an attractive limit cycle. Moreover, for a particular choice of feedback, the closed loop system may take the shape of a Van der Pol oscillator. 2