Gravitational search algorithm based optimization technique for enhancing the performance of self excited induction generator

Abstract

In wind based micro generation schemes, 3-phase self excited induction generators are prominently used in order to fulfil single phase load requirement. Hence in this context, this paper presents a 3-phase, 5.5 kW, 415 V, 50 Hz short shunt self excited induction generator for improving the voltage regulation and optimum performance of induction machine by using heuristic approach named as gravitational search algorithm. It is used in order to get the optimum capacitance values at specified speed for optimized voltage regulation and performance characteristics in terms of root mean square error and mean absolute error and mean square error. This optimization technique works on Newton law of gravity and it provides average best results for validating the performance in order to enhance machine parameters used in wind energy conversion system.

Appendices

Appendix 1: Specification of self excited induction generator

5.5 kW, 415 V, 10.1 A (line), Δ connected, 4 Pole, 50 Hz

R_s = 0.0632pu, R_r = 0.0247pu, X_s = 0.0633pu, X_r = 0.0632pu, X ^uns_m  = 3.48 pu.

Appendix 2: Coefficients of Z_total

The coefficients P and Q are defined as:

$$P_{1} = - x_{s} \cdot r_{L} (x_{r} + x_{m} ) - x_{r} \cdot r_{L} \cdot x_{m}$$

$$P_{2} = x_{s} \cdot r_{L} \cdot v \cdot (x_{r} + x_{m} ) + x_{r} \cdot r_{L} \cdot x_{m} \cdot v$$

$$P_{3} = r_{s} \cdot r_{r} \cdot r_{L} + x_{s} \cdot x_{c} \cdot x_{r} + (r_{s} \cdot x_{c} + r_{L} \cdot x_{c} ) \cdot (x_{r} + x_{m} ) + r_{r} \cdot x_{m} \cdot x_{c}$$

$$P_{4} = ( - v \cdot r_{s} \cdot x_{c} - v \cdot r_{L} \cdot x_{c} )(x_{r} + x_{m} )$$

$$P_{5} = 0$$

$$Q_{1} = 0$$

$$Q_{2} = x_{s} \cdot r_{L} \cdot r_{r} + (x_{r} + x_{m} ) \cdot (r_{s} \cdot r_{L} + x_{s} \cdot x_{c} ) + x_{m} \cdot r_{r} \cdot r_{L} + x_{m} \cdot x_{c} \cdot x_{r}$$

$$Q_{3} = (x_{r} + x_{m} ) \cdot ( - r_{s} \cdot r_{L} \cdot v - x_{s} \cdot x_{c} \cdot v)x_{m} \cdot v \cdot x_{r}$$

$$Q_{4} = - r_{L} \cdot x_{c} \cdot r_{r} - r_{s} \cdot r_{r} \cdot x_{c}$$

The performance parameters are:

• vg = vg1 * freq

• zs = rs + j(freq.Xs)

• zr = freq.R_r/(freq − v) + jfreq.Xr;

• zcL = − jrL(xc/freq)/(rL − jxc/freq)

• is = vg/(zs + zcL)

• isss = abs(is)

• iL = (− j(xc/freq)/(− j(x_c/freq) + rL)).is

• iLrms = abs(iL)

• otpr = iLrms * rL

• vt = iLrms * (rL)

• var = vt * freq/xc

• ir = − vg/zr

• irrms = abs(ir)

The complete methodology adopted for optimum performance of SEIG machine has been shown in Fig. 7. The flowchart includes all the steps from machine parameters to performance characteristics.