Abstract
A novel method of solar power forecasting for individuals and small businesses is developed in this paper based on machine learning, image processing, and acoustic classification techniques. Increases in the production of solar power at the consumer level require automated forecasting systems to minimize loss, cost, and environmental impact for homes and businesses that produce and consume power (prosumers). These new participants in the energy market, prosumers, require new artificial neural network (ANN) performance tuning techniques to create accurate ANN forecasts. Input masking, an ANN tuning technique developed for acoustic signal classification and image edge detection, is applied to prosumer solar data to improve prosumer forecast accuracy over traditional macrogrid ANN performance tuning techniques. ANN inputs tailor time-of-day masking based on error clustering in the time domain. Results show an improvement in prediction to target correlation, the R2 value, lowering inaccuracy of sample predictions by 14.4%, with corresponding drops in mean average error of 5.37% and root mean squared error of 6.83%.
Highlights
Power service providers are increasing the use of solar power due to decreases in the cost of solar power production systems, increases in the cost of traditional energy sources, environmental concerns, and legislative requirements
These same forces increase the prevalence of homes and small businesses with solar panels and storage that produce solar power to use in the home or business or store in battery banks or smart appliances or sell power back to power companies in tiered or real-time pricing structures
artificial neural network (ANN) are strongly dependent on scale, resolution, and forecast variables [1]; the ANNs developed for power companies are not suitable for prosumer use
Summary
Power service providers are increasing the use of solar power due to (among many factors) decreases in the cost of solar power production systems, increases in the cost of traditional energy sources, environmental concerns, and legislative requirements. Forecasting the production of large solar arrays and wind farms allows power providers the time necessary to make changes to base load power plant production to minimize peak power plant use These forecasts often use artificial neural networks (ANNs) which access multiple and varied data sources to estimate power changes hours or days in advance. The ANN investigated is intended for use at the prosumer level, which is less likely to have access to complex weather data; this model uses measurements of variability in the irradiance measurements to assess cloud cover This model’s focus on the prosumer needs and the mitigation of the nondispatchable nature of renewable energy have dictated this model’s concentration on short-term forecasting. This paper details the use of input masking in ANNs unique to short-term solar power forecasting
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
