Solar PV Positioning Optimizer Module (ZKML - powered)

Category: ZKML with GizaTech's Orion Framework

The Problem     

Solar power production presents a promising pathway to mitigate anthropogenic climate change by reducing reliance on fossil fuels, diminishing greenhouse gas emissions from the energy sector (Masson et al., 2014). Solar photovoltaic (PV) systems are gaining popularity worldwide due to abundant solar irradiance and declining technology costs (Lane, 2022). However, solar PV efficiency depends heavily on positioning and orientation, which must account for geographic location and sun angle variabilities. Optimized solar PV alignment is therefore critical for maximizing power generation (Novergy, 2019). To tackle these challenges, the integration of machine learning (ML) and artificial intelligence (AI) is proposed, raising concerns about the verifiability of computations and inferences. Zero-knowledge machine learning (ZKML) offers a solution, enabling model inferences to be cryptographically proven without revealing constituent data or algorithms.

Proposed Solution   

This work proposes a Solar PV Positioning Optimizer Module that optimizes solar PV orientation in specific households via IoT-enabled data collection, cloud analytics, and ZKML. The system comprises hardware including sensors that measure panel angle, humidity, temperature and current output at differing alignments. Actuators manipulate panel positioning across these alignments. Collected data trains machine learning models to estimate expected power and irradiance at potential orientations. ZKML techniques then formally verify the integrity of these computations. The module thus provides households tailored insight on optimal solar PV positioning for their locale. Further, aggregated analytics offer solar companies rich insights into regional solar patterns for superior customer targeting and infrastructure planning. Solar optimization via ZKML thereby furnishes both personalized and collective value.

Value Proposition   

While solar trackers enhance electricity production, their additional costs often outweigh the benefits (Lane, 2022). The proposed Solar PV Positioning Optimizer Module (SPPOM) prioritizes optimizing solar panel positioning and orientation as a cost-effective alternative. By collecting accurate data to assess Solar PV efficiency for specific households, the SPPOM empowers stakeholders to make informed investment decisions. Beyond statistical reports and ML model development, the collected data becomes valuable for market research, aiding design teams in identifying optimal locations for solar communities. The unique contribution of Zero-knowledge Machine Learning (ZKML) ensures the verifiability and integrity of ML computations and inferences.

Technology Implementation   

The hardware component integrates essential sensors and actuators, including accelerometers, humidity and temperature sensors, real-time clocks, current sensors, and servos. Miniature solar panels attached to the servos move at various angles to collect power output data. The collected data is stored on an SD card data logger for local backup and is also transmitted to the cloud. The backup storage is crucial for areas with unreliable internet connectivity. The data is processed, analyzed, and used to train ML models and ZKMLs, determining optimal positions and orientations for Solar PVs to maximize power output and irradiance . The innovative inclusion of ZKML using Cairo and Gizatech's Orion Framework ensures the trustworthiness and transparency of ML-powered solutions in the context of solar energy optimization.

The Hardware Design     

The Schematic Diagram   

The CAD Model   

Conclusion and Results   

On this file, we conducted a detailed and thorough analysis of the data acquired during one of our sample field tests:

https://github.com/John-Embate/SPPOM-STARKNET-INFRA/tree/main/SPPOM%20Data%20Analysis%20and%20Modeling

Github Repository   

https://github.com/John-Embate/SPPOM-STARKNET-INFRA

The Team     

• John William Q. Embate
• Plutarch Romiel Degollacion

References

(2019, March 28). Retrieved from Novergy: https://www.novergysolar.com/heres-why-orientation-and-positioning-of-solar-panels-is-so-important/

Lane, C. (2022, January 3). Retrieved from Solar Reviews: https://www.solarreviews.com/blog/are-solar-axis-trackers-worth-the-additional-investment

Masson, V., Bonhomme, M., Salagnac, J.-L., Briottet, X., & Lemonsu, A. (2014, June 4). Solar panels reduce both global warming and urban heat island. doi:https://doi.org/10.3389/fenvs.2014.00014