While moving towards a low-carbon, sustainable electricity system, distribution networks are expected to host a large share of distributed generators, such as photovoltaic units and wind turbines. These inverter-based...While moving towards a low-carbon, sustainable electricity system, distribution networks are expected to host a large share of distributed generators, such as photovoltaic units and wind turbines. These inverter-based resources are intermittent, but also controllable, and are expected to amplify the role of distribution networks together with other distributed energy resources, such as storage systems and controllable loads. The available control methods for these resources are typically categorized based on the available communication network into centralized, distributed, and decentralized or local. Standard local schemes are typically inefficient, whereas centralized approaches show implementation and cost concerns. This paper focuses on optimized decentralized control of distributed generators via supervised and reinforcement learning. We present existing state-of-the-art decentralized control schemes based on supervised learning, propose a new reinforcement learning scheme based on deep deterministic policy gradient, and compare the behavior of both decentralized and centralized methods in terms of computational effort, scalability, privacy awareness, ability to consider constraints, and overall optimality. We evaluate the performance of the examined schemes on a benchmark European low voltage test system. The results show that both supervised learning and reinforcement learning schemes effectively mitigate the operational issues faced by the distribution network.展开更多
Installation of smart meters enables electricity retailers or consumers to implement individual load forecasting for demand response.An individual load forecasting model can be trained either on each consumer’s own s...Installation of smart meters enables electricity retailers or consumers to implement individual load forecasting for demand response.An individual load forecasting model can be trained either on each consumer’s own smart meter data or the smart meter data of multiple consumers.The former practice may suffer from overfitting if a complex model is trained because the dataset is limited;the latter practice cannot protect the privacy of individual consumers.This paper tackles the dilemma by proposing a personalized federated approach for individual consumer load forecasting.Specifically,a group of consumers first jointly train a federated forecasting model on the shared smart meter data pool,and then each consumer personalizes the federated forecasting model on their own data.Comprehensive case studies are conducted on an open dataset of 100 households.Results verify the proposed method can enhance forecasting accuracy by making full use of data from other consumers with privacy protection.展开更多
文摘While moving towards a low-carbon, sustainable electricity system, distribution networks are expected to host a large share of distributed generators, such as photovoltaic units and wind turbines. These inverter-based resources are intermittent, but also controllable, and are expected to amplify the role of distribution networks together with other distributed energy resources, such as storage systems and controllable loads. The available control methods for these resources are typically categorized based on the available communication network into centralized, distributed, and decentralized or local. Standard local schemes are typically inefficient, whereas centralized approaches show implementation and cost concerns. This paper focuses on optimized decentralized control of distributed generators via supervised and reinforcement learning. We present existing state-of-the-art decentralized control schemes based on supervised learning, propose a new reinforcement learning scheme based on deep deterministic policy gradient, and compare the behavior of both decentralized and centralized methods in terms of computational effort, scalability, privacy awareness, ability to consider constraints, and overall optimality. We evaluate the performance of the examined schemes on a benchmark European low voltage test system. The results show that both supervised learning and reinforcement learning schemes effectively mitigate the operational issues faced by the distribution network.
基金The work was supported by the activities of the Renewable Management and Real-Time Control Platform(ReMaP)financially supported by the Swiss Federal Office of Energy(SFOE)。
文摘Installation of smart meters enables electricity retailers or consumers to implement individual load forecasting for demand response.An individual load forecasting model can be trained either on each consumer’s own smart meter data or the smart meter data of multiple consumers.The former practice may suffer from overfitting if a complex model is trained because the dataset is limited;the latter practice cannot protect the privacy of individual consumers.This paper tackles the dilemma by proposing a personalized federated approach for individual consumer load forecasting.Specifically,a group of consumers first jointly train a federated forecasting model on the shared smart meter data pool,and then each consumer personalizes the federated forecasting model on their own data.Comprehensive case studies are conducted on an open dataset of 100 households.Results verify the proposed method can enhance forecasting accuracy by making full use of data from other consumers with privacy protection.