Innovative adaptive threshold based battery energy storage system controller using deep learning forecast for peak demand reductions
Battery-based energy storage system (BESS) can reduce daily peak demands when it is managed by an effective controller or a control strategy. However, most existing BESS controllers are implemented in simulation platforms, with limited experimental validations under real operating conditions. Even w...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Year Project / Dissertation / Thesis |
| Published: |
2025
|
| Subjects: | |
| Online Access: | http://eprints.utar.edu.my/7145/1/Dissertation_MD_Mahmudul_Hasan.pdf http://eprints.utar.edu.my/7145/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1855616541066788864 |
|---|---|
| author | MD, Mahmudul Hasan |
| author_facet | MD, Mahmudul Hasan |
| author_sort | MD, Mahmudul Hasan |
| building | UTAR Library |
| collection | Institutional Repository |
| content_provider | Universiti Tunku Abdul Rahman |
| content_source | UTAR Institutional Repository |
| continent | Asia |
| country | Malaysia |
| description | Battery-based energy storage system (BESS) can reduce daily peak demands when it is managed by an effective controller or a control strategy. However, most existing BESS controllers are implemented in simulation platforms, with limited experimental validations under real operating conditions. Even when implemented experimentally, they are often tested on limited case studies or evaluated without any evaluation metrics. Additionally, majority of the controllers are developed using paid proprietary platforms, and do not
incorporate any advanced load forecasting model. Therefore, this research aims to address these gaps by developing an innovative adaptive threshold-based BESS controller using free, open-source platforms Node-RED and Python,
integrating an advanced deep learning-based one-dimensional convolution neural network (1D-CNN) model for load forecasting. The proposed controller is initially evaluated through simulation using six-months of data, with its
performance benchmarked against four different controllers using two different evaluation metrics: daily peak reduction factor ( ), and monthly failure rate ( ). Subsequently, the controller is deployed on a 200 kW/200 kWh BESS setup at a university campus in Malaysia to evaluate its practical performance over 21 days under real operating conditions. In simulation, the proposed controller performs better than that of those benchmark controllers, achieving an average of 41.62% and of 16.55%. When tested on the actual BESS setup, the controller shows improved performance, with an average
of 49.45% and of just 4.76%. These findings highlight the potential of the proposed adaptive threshold-based controller enhanced with advanced load forecasting model for real-world grid applications and can provide significant benefits to both utilities and end customers.
Keywords: Battery energy storage system (BESS), bess controller, peak demand reduction, load forecasting, deep learning, 1D-CNN |
| format | Final Year Project / Dissertation / Thesis |
| id | my-utar-eprints.7145 |
| institution | Universiti Tunku Abdul Rahman |
| publishDate | 2025 |
| record_format | eprints |
| spelling | my-utar-eprints.71452026-01-13T08:27:14Z Innovative adaptive threshold based battery energy storage system controller using deep learning forecast for peak demand reductions MD, Mahmudul Hasan H Social Sciences (General) HD Industries. Land use. Labor ZA Information resources Battery-based energy storage system (BESS) can reduce daily peak demands when it is managed by an effective controller or a control strategy. However, most existing BESS controllers are implemented in simulation platforms, with limited experimental validations under real operating conditions. Even when implemented experimentally, they are often tested on limited case studies or evaluated without any evaluation metrics. Additionally, majority of the controllers are developed using paid proprietary platforms, and do not incorporate any advanced load forecasting model. Therefore, this research aims to address these gaps by developing an innovative adaptive threshold-based BESS controller using free, open-source platforms Node-RED and Python, integrating an advanced deep learning-based one-dimensional convolution neural network (1D-CNN) model for load forecasting. The proposed controller is initially evaluated through simulation using six-months of data, with its performance benchmarked against four different controllers using two different evaluation metrics: daily peak reduction factor ( ), and monthly failure rate ( ). Subsequently, the controller is deployed on a 200 kW/200 kWh BESS setup at a university campus in Malaysia to evaluate its practical performance over 21 days under real operating conditions. In simulation, the proposed controller performs better than that of those benchmark controllers, achieving an average of 41.62% and of 16.55%. When tested on the actual BESS setup, the controller shows improved performance, with an average of 49.45% and of just 4.76%. These findings highlight the potential of the proposed adaptive threshold-based controller enhanced with advanced load forecasting model for real-world grid applications and can provide significant benefits to both utilities and end customers. Keywords: Battery energy storage system (BESS), bess controller, peak demand reduction, load forecasting, deep learning, 1D-CNN 2025 Final Year Project / Dissertation / Thesis NonPeerReviewed application/pdf http://eprints.utar.edu.my/7145/1/Dissertation_MD_Mahmudul_Hasan.pdf MD, Mahmudul Hasan (2025) Innovative adaptive threshold based battery energy storage system controller using deep learning forecast for peak demand reductions. Master dissertation/thesis, UTAR. http://eprints.utar.edu.my/7145/ |
| spellingShingle | H Social Sciences (General) HD Industries. Land use. Labor ZA Information resources MD, Mahmudul Hasan Innovative adaptive threshold based battery energy storage system controller using deep learning forecast for peak demand reductions |
| title | Innovative adaptive threshold based battery energy storage system controller using deep learning forecast for peak demand reductions |
| title_full | Innovative adaptive threshold based battery energy storage system controller using deep learning forecast for peak demand reductions |
| title_fullStr | Innovative adaptive threshold based battery energy storage system controller using deep learning forecast for peak demand reductions |
| title_full_unstemmed | Innovative adaptive threshold based battery energy storage system controller using deep learning forecast for peak demand reductions |
| title_short | Innovative adaptive threshold based battery energy storage system controller using deep learning forecast for peak demand reductions |
| title_sort | innovative adaptive threshold based battery energy storage system controller using deep learning forecast for peak demand reductions |
| topic | H Social Sciences (General) HD Industries. Land use. Labor ZA Information resources |
| url | http://eprints.utar.edu.my/7145/1/Dissertation_MD_Mahmudul_Hasan.pdf http://eprints.utar.edu.my/7145/ |
| url_provider | http://eprints.utar.edu.my |