能否使用PVlib对单轴（垂直）及双轴光伏阵列进行预测？

Absolutely! PVlib is fully equipped to handle energy yield predictions for both vertical single-axis and dual-axis PV arrays—nice work getting your POA irradiance calculations sorted out already, that’s a critical first step. Here’s how you can build out the rest of the prediction workflow:

Vertical Single-Axis Arrays #

Since you’re already using pvlib.irradiance.aoi_projection to calculate your POA irradiance, you’re halfway there. The next steps mirror standard PV yield prediction:

• DC Power Calculation: Use PVlib’s component models to compute DC power. You can start with the simplified pvlib.pvsystem.pvwatts_dc if you have module specs from the PVWatts database, or use more detailed models like pvlib.pvsystem.singlediode (paired with parameter calculators like pvlib.pvsystem.calcparams_desoto or pvlib.pvsystem.calcparams_pvsyst) for higher accuracy.
• AC Power & Yield: Pass the DC power through an inverter model (e.g., pvlib.inverter.pvwatts_ac) to get AC power, then aggregate over time to calculate total energy yield.
• Bonus: Validate Tracking Logic: PVlib’s pvlib.tracking.singleaxis supports vertical-axis configurations (tracking solar azimuth). You can use this function to generate expected tracking angles, cross-check your AOI projection calculations, or even replace your current POA calculation if you want a streamlined end-to-end workflow.

Dual-Axis Arrays #

Your initial POA irradiance work sets you up perfectly for dual-axis predictions too. Here’s what to do next:

• Cross-Check POA Calculations: PVlib’s pvlib.tracking.dualaxis function computes dual-axis tracking angles, AOI, and POA irradiance automatically. Use this to validate your existing POA results—matching outputs will give you confidence in your data.
• Yield Prediction Workflow: Follow the same DC-to-AC path as vertical single-axis arrays: model module performance (don’t forget to account for cell temperature with pvlib.temperature.pvsyst_cell or similar), apply inverter efficiency, and add system losses (via pvlib.pvsystem.losses for things like soiling, shading, or wiring losses).

Key Tips for Accuracy #

• Temperature Matters: Cell temperature has a huge impact on module output. Make sure to integrate a temperature model using ambient temperature, wind speed, and your POA irradiance data.
• Losses Add Up: Don’t skip system loss calculations—even small losses (2-5% for soiling, 1-2% for wiring) can significantly affect long-term yield predictions.
• Calibrate with Real Data: If you have historical generation data from the array, use it to tune model parameters (like loss factors or module temperature coefficients) for more precise predictions.

内容的提问来源于stack exchange，提问作者Jordi Vidal de LLobatera