Website Solar Monkey
Improving state-of-the-art PV system yield prediction algorithms
Duration: 10 weeks
Start: July 1st
Class size: 5 students
Supervisor: Ir. J. Donker (firstname.lastname@example.org)
Note: this fulfills the requirement for the SET MSc Internship of 15 ECTS
Solar Monkey monitors approximately 4000 systems. Based on the data from these systems, a thorough analysis will be performed on a range of improvements. The process in which these improvements will be proposed by the students is as follows:
- Use KNMI (hourly) weather reports to predict the hourly yield for days that we have monitored systems.
- Change or add to different parts of the prediction algorithm.
- Run experiments to see the effect on the change in accuracy of the predictions for a large number of systems
- Analyze the results from the experiments and present to the team in a weekly meeting
- Draw conclusions based on the experiments
- Recommend changes to Solar Monkey how to improve our calculations from a pv modelling perspective
A incomplete list of changes/additions to be made to algorithms to evaluate:
– Seasonal differences (summer vs winter over and underestimation)
– Temperature modelling
– Inverter efficiency modelling
– Irradiance modelling (use different irradiance diffuse + direct splitting models)
It is expected from the participants to come up with their own ideas for improvement. The students will work as an autonomous team, dividing responsibilities between themselves and present their progress and findings during weekly meetings. After 10 weeks, they will give a presentation to the whole Solar Monkey team, presenting their findings and recommendations.
To apply for this job email your details to email@example.com