In domestic field, washing machine and dishwasher loads are categorized as controllable- shiftable loads.
Nowadays, what are most known published works on this topic?
In particular, what is the most effective approach in formulating such kind of loads?
What are the most effective algorithms which are able to schedule in real time the cycle of such loads without assuming to know a priori the future energy demand?
What are is the most common objective function used by a training (optimization) algorithm?
Article Real time optimal schedule controller for home energy manage...
You can find this article
Appliance scheduling for optimal load management in smart home integrated with renewable energy by using whale optimization algorithm
Here are some articles that might be of your interest
https://venturebeat.com/2020/04/20/utilities-energy-usage-covid-19-ai-machine-learning/
Conference Paper Real-Time Recognition and Profiling of Appliances through a ...
Article Machine Learning Approach Electric Appliance Consumption and...
I want to thanks all the useful message.
Reading the papers and links I realized that in the literature most of the algorithms proposed are rule based or rather based on evolutionary optimization algorithms. Also more conventional approaches are also used which are based on the adoption of well known optimization programming such as Non-Linear Mixed Integer Programming etc. In my opinion, these kind of solutions, except for rule based systems, suffer the need of a suitable prediction algorithm. In smart home field, it can be a crucial issue since the small scale smart home systems make the aggregated load very stochastic and intermittent.
In reality, the metering infrastructure at LV level is limited and they're expensive. Therefore, any practical optimization algorithm has to work with the limited data. This could be one of the reasons why we still prefer rule based control. But prediction based optimal control such as MPC is gaining momentum at least in research fields. I don't know how long it will take for those technologies to get adapted in large scale.
Thank you for the useful comment. Do you have a publication to suggest?
One of the interesting references I found recently was the Ph.D. dissertation of David Fischer (2017). The study focuses on heat pumps, but the concepts can be generalized for other decentralized services such as electric vehicles.
https://kth.diva-portal.org/smash/get/diva2:1141142/FULLTEXT02.pdf
There are some conference papers on the use of machine learning. Check for example;
Preprint Advancing Renewable Electricity Consumption With Reinforceme...
Quite recently I started working on a project on electric vehicle integration where we try to use machine learning and high-performance computing. We don't have any peer-reviewed publications yet, but hopefully, we will get there soon. I don't know what your specific interest is, but the lack of very good peer-reviewed publications on this topic itself says a lot about the potential to develop some innovative ideas in this field. Feel free to keep in touch through email and personal messages. It will be quite useful for us, those who have engaged actively in this field to share ideas and collaborate.
- Badges
- Science topic