**5. Conclusions**

The paper has proposed a simple methodology for quantifying the annual energy production by a microinverter based on experimental mission profiles of the solar irradiance and ambient temperatures, PV module parameters identified from a datasheet using PSIM Solar module utility, and experimental efficiency interpolation based on measured values. The methodology uses a single-diode five-parameter model of a PV cell enhanced with thermally dependent model of the series resistor.

The proposed methodology was applied to a case study microinverter utilizing a buck-boost front-end converter controlled by variable and fixed dc-link voltage control methods. The experimental efficiency was found to reach over 95% in both cases, while it is increased by up to 2% on average in the tested range with the application of the variable dc-link voltage control. Experimental efficiency interpolations were used to estimate energy production by a microinverter connected to a properly sized PV module. The variable dc-link voltage control extends efficient operation from a single to several compatible PV module configurations, i.e., different number of PV cells per module.

Our one-day experimental test of PV energy harvesting by the case study microinverter showed an energy production increase by 1.8%. This value correlates well with the predicted increase of the annual energy production by 1.8% and 2% observed for a cloudy northern climate and a sunny southern climate, correspondingly. This increase could be translated in 22.3% and 26% reduction of the annual energy loss, correspondingly. Hence, the variable dc-link voltage control provides over 20% reduction of the annual energy loss regardless of the climatic conditions.

The variable dc-link voltage control is a promising solution for the two-stage microinverters with capacitive intermediate dc-link buffering 100/120 Hz power pulsations. Our future work will aim to extend the proposed methodology with a model of power loss from the MPPT, which was left out of this study.

**Author Contributions:** Conceptualization, A.C. and D.V.; methodology, A.C. and E.L.; software, A.C. and S.S.; validation, D.V. and S.S.; formal analysis, A.C. and D.V.; investigation, E.L. and F.B.; resources, E.L. and D.V.; data curation, F.B. and S.S.; writing—original draft preparation, A.C. and S.S.; writing—review and editing, E.L., D.V., and F.B.; visualization, A.C. and S.S.; supervision, E.L. and F.B.; project administration, D.V. and E.L.; funding acquisition, D.V., S.S., and A.C.

**Funding:** This research was supported in part by the Estonian Centre of Excellence in Zero Energy and Resource Efficient Smart Buildings and Districts, ZEBE, gran<sup>t</sup> 2014-2020.4.01.15-0016 funded by the European Regional Development Fund, in part by the Estonian Research Council gran<sup>t</sup> PSG206, in part by the European Regional Development Fund and the programme Mobilitas Pluss under the project MOBJD126 awarded by the Estonian Research Council, and in part by SERC Chile (CONICYT/FONDAP/15110019) and AC3E (CONICYT/BASAL/FB0008).

**Conflicts of Interest:** The authors declare no conflict of interest.
