In this paper, artificial neural network (ANN) based Levenberg-Marquardt (LM), Bayesian Regularization (BR) and Scaled Conjugate Gradient (SCG) algorithms are deployed in maximum power point tracking (MPPT) energy harvesting in solar photovoltaic (PV) system to forge a comparative performance analysis of the three different algorithms. A comparative analysis among the algorithms in terms of the performance of handling the trained dataset is presented. The MATLAB/Simulink environment is used to design the maximum power point tracking energy harvesting system and the artificial neural network toolbox is utilized to analyze the developed model. The proposed model is trained with 1000 dataset of solar irradiance, temperature, and voltages. Seventy percent data is used for training, while 15% data is employed for validation, and 15% data is utilized for testing. The trained datasets error histogram represents zero error in the training, validation, and test phase of data matching. The best validation performance is attained at 1000 epochs with nearly zero mean squared error where the trained data set is converged to the best training results. According to the results, the regression and gradient are 1, 1, 0.99 and 0.000078, 0.0000015739 and 0.26139 for Levenberg-Marquardt, Bayesian Regularization and Scaled Conjugate Gradient algorithms, respectively. The momentum parameters are 0.0000001 and 50000 for Levenberg-Marquardt and Bayesian Regularization algorithms, respectively, while the Scaled Conjugate Gradient algorithm does not have any momentum parameter. The Scaled Conjugate Gradient algorithm exhibit better performance compared to Levenberg-Marquardt and Bayesian Regularization algorithms. However, considering the dataset training, the correlation between input-output and error, the Levenberg-Marquardt algorithm performs better.
4 November 2019 in IEEE Internet of Things Journal
With the development of the Internet of Things (IoT) technologies, implementing intelligent controls in buildings to reduce energy consumption is becoming increasingly popular. Building climate control is of strong research interest due to high energy savings potentials associated with heating, ventilation and air-conditioning (HVAC) controls. Because the operation of HVAC systems directly influences occupant thermal comfort, building-specific thermal models are needed for proper control. Such models describe how indoor temperature changes under different environmental conditions and HVAC status. Previous studies mainly developed thermal models for accurate building thermal simulation which are not practical for use in real-world applications as they require domain knowledge and lengthy building-by-building configuration. In this paper, taking advantage of IoT technologies, a plug & play learning framework is proposed to automatically identify the thermal model of each thermal zone in a building without manual configuration. In contrast to the existing methods, a thermal model is learnt using low-resolution temperature readings from IoT-based smart thermostats. The learning framework has been validated using data collected from an IoT platform installed in a building over a summer. The performance of the learnt model and the error of indoor temperature prediction based on this model have been evaluated and quantified. Findings demonstrate the learning process can be automated with either edge-computing or cloud-computing. This study shows the validity of the proposed IoT-based thermal model learning framework, and offers a pragmatic solution for providing reliable thermal models to future smart building climate controls.
Buildings, as major energy consumers, can provide great untapped demand response (DR) resources for grid services. However, their participation remains low in real-life. One major impediment for popularizing DR in buildings is the lack of cost-effective automation systems that can be widely adopted. Existing optimization-based smart building control algorithms suffer from high costs on both building-specific modeling and on-demand computing resources. To tackle these issues, this paper proposes a cost-effective edge-cloud integrated solution using reinforcement learning (RL). Beside RL’s ability to solve sequential optimal decision-making problems, its adaptability to easy-to-obtain building models and the off-line learning feature are likely to reduce the controller’s implementation cost. Using a surrogate building model learned automatically from building operation data, an RL agent learns an optimal control policy on cloud infrastructure, and the policy is then distributed to edge devices for execution. Simulation results demonstrate the control efficacy and the learning efficiency in buildings of different sizes. A preliminary cost analysis on a 4-zone commercial building shows the annual cost for optimal policy training is only 2.25% of the DR incentive received. Results of this study show a possible approach with higher return on investment for buildings to participate in DR programs.
Reinforcement learning based energy management strategy has been an active research subject in the past few years. Different from the baseline reward function (BRF), the work proposes and investigates a multi-stage reward mechanism (MSRM) that scores the agent’s step and final performance during training and returns it to the agent in real time as a reward. MSRM will also improve the agent’s training through expert intervention which aims to prevent the agent from being trapped in sub-optimal strategies. The energy management performance considered by MSRM-based algorithm includes the energy balance, economic cost, and reliability. The reward function is assessed in conjunction with two deep reinforcement learning algorithms: double deep Q-learning network (DDQN) and policy gradient (PG). Upon benchmarking with BRF, the numerical simulation shows that MSRM tends to improve the convergence characteristic, reduce the explained variance, and reduce the tendency of the agent being trapped in suboptimal strategies. In addition, the methods have been assessed with MPC-based energy management strategies in terms of relative cost, self-balancing rate, and computational time. The assessment concludes that, in the given context, PG-MSRM has the best overall performance.
1 June 2007 in 2007 IEEE Power Engineering Society General Meeting
The electric power system is an enabling infrastructure that supports the operation of other critical infrastructures and thus the economic well-being of a nation. It is, therefore, very important to design for resiliency and autonomous reconfigurability in the electric power grid to guard against man-made and natural disasters. One way to assure such self-healing characteristics in an electric power system is to design for small and autonomous subsets of the larger grid. This paper presents the concept of a specialized microgrid called an intelligent distributed autonomous power system (IDAPS). The IDAPS microgrid aims at intelligently managing customer-owned distributed energy resources such that these assets can be shared in an autonomous grid both during normal and outage operations. The proposed concept is expected to make significant contributions during emergency conditions, as well as creating a new market for electricity transaction among customers.
Under-frequency load shedding (UFLS) is an important measure for tackling low-frequency events caused by load-generation imbalance. However, the uncertainty of wind power amplifies power imbalances and can potentially impair frequency stability. Electric vehicles (EVs) present a more effective means for addressing this issue compared to load shedding. However, EVs have several limitations such as commute randomness. To ensure frequency stability and simultaneously reduce load shedding, a bi-level confidence-interval-based optimal strategy is proposed to enable the participation of EVs in UFLS, where the uncertainties of wind power and the commute randomness of EVs are estimated using a non-parametric kernel density estimation (KDE) method. In bi-level optimization, the upper level reduces the dependency on commute randomness and the wind power uncertainty during load-shedding events. Further, the upper-level solutions are sent to EV charging stations for emergency dispatch. By contrast, at the lower level, an approximation-function-based priority is proposed to optimize the task allocation. Simulation results show the advantages of the proposed approach in maintaining a stable frequency compared with traditional and adaptive UFLS schemes.