Whether you are doing your experiment practically at a lab or virtually using a simulator, you need to measure both input and output powers of the inverter to calculate the efficiency.  Regarding the harmonics contents and calculation of the quality of the AC you obtained you need to use a power analyzer to read harmonics amplitudes, or if you are using say Simpowersystems of MATLAB/SIMULINK, you need to use its FFT analysis.

Hope this helps.

Thanks  Salah Eldeen Gasim,

sir,

Currently i am doing simulation using simpowersystems. I did FFT analysis and found harmonic contents. 

you mean to say, efficiency = [vo(rms)* io(rms)*pf)/(Vdc* idc)].

where idc is the dc component obtained when do FFT of the input current... 

Welcome Mr. Sreekanth,

Yes exactly, you can measure the DC voltage and current as well at the inverter input side.

In the presence of harmonics, the correct way to measure power is to take the time average of the instantaneous product of voltage and current. It is the same as the average of instantaneous power. Measuring pf under distortion can be quite tricky since current harmonics do not contribute to average power if voltage is sinusoidal. Hence Irms may not be of use either.

However, on DC side, if the voltage is perfectly constant (without ripple), then Vdc*Iav will give the true power. Otherwise the above process is needed.

Thank you sujit K. Biswas sir,

as you said we have to consider only DC component of input current while calculating input power. My question is what about other components like 100Hz ripple and switching ripple. what is the importance or drawbacks of these components... If load is at UPF ,  there is no reactive power, then what is the use of this components? and where this components go from input side.

If you are talking of harmonics on DC side, then they are basically AC components with zero average value. Hence, multiplying them by a steady DC means average value is still zero ! Hence they cannot contribute to average power transfer.

The 100Hz ripple is created by the reflection of power flow at AC side to the DC side, At unity pf, when the current starts from zero, the power also starts from zero. When the current reaches its peak (at same time as voltage peak), the power drawn reaches peak. So on and so forth. Thus the power drawn is not steady but a double frequency component with an average value. This creates the 100Hz ripple current on DC side. Similar for switching ripples.

It is interesting to note that such multiple of power frequency ripple will not be ideally present at the DC side of a 3-phase inverter since the power flow can be shown to be steady in 3-phase systems, unlike the pulsating form in 1-phase systems ! Switching frequency components however will still be present.

Thank you sujit K. Biswas sir,  your explanation is very good.

Now my doubt is clarified.

Instead an answer I woulld like to ask a question. Namely, if one (Sreekanth Rtg, for example) uses

efficiency = [vo(rms)* io(rms)*pf)/(Vdc* idc)]

what numbers he gets?

May I have them all: efficiency,  vo(rms),  io(rms), Vdc, idc?

I would like add a question here. Since due to harmonics graphs shows fluctuations. How to take a quantative value from this graph ? average or mean constant value i.e. 78% Efficiency or 1400W power.