Reinforcement learning-based optimization for power scheduling in a renewable energy connected grid

Cross-country electricity trade, renewable energy and European transmission infrastructure policy

Paths of renewable energy development in small island developing states of the South Pacific

Wind powered pumped hydro storage systems, a means of increasing the penetration of renewable energy in the Canary Islands

Increasing the share of intermittent renewable energy (RE) will challenge the capability of existing power systems to cope with increased variability in net load. This may result in frequent RE and load curtailment in the absence of adequate flexibility. To maximize RE utilization and maintain supply reliability, system flexibility must be enhanced. However, the capability of various flexibility resources to facilitate RE penetration needs to be assessed for optimal system planning. Further, to capture the impact of RE's seasonal variability, a longterm assessment of chronological short-term system operations is required. In this context, this paper proposes a computational framework of the yearly unit commitment model. An attempt has been made to maximize RE penetration at different integration levels by increasing operational flexibility and energy storage integration. Results from various cases examined on the IEEE RTS 24-bus system show that although enhanced operational flexibility is a viable solution for smooth RE integration, a high RE share would require alternative capital-intensive solutions.

Integrated energy-water systems for community-level flexibility: A hybrid deep Q-network and multi-objective optimization framework

Estimation of parameters of chaotic systems is a subject of substantial and well-developed research issue in nonlinear science. From the viewpoint of optimization, parameter estimation can be formulated as a multi-modal constrained optimization problem with multiple decision variables. This investigation makes a systematic examination of the feasibility of applying a newly proposed population-based optimization method labeled here as teaching---learning-based optimization (TLBO) to identify the unknown parameters for a class of chaotic system. The preliminary test demonstrates that despite its global fast coarse search capability, teaching---learning-based optimization often risks getting prematurely stuck in local optima. To enhance its fine (local) searching performance of TLBO, Nelder---Mead simplex algorithm-based local improvement is incorporated into TLBO so as to continually search for the global optima through the reflection, expansion, contraction, and shrink operators. Working with the well-established Lorenz system, we assess the effectiveness and efficiency of the proposed improved TLBO strategy. The empirical results indicate the success of the proposed hybrid approach in which the global exploration and the local exploitation are well balanced, providing the best solutions for all instances used over other state-of-the-art metaheuristics for chaotic identification in literature, including particle swarm optimization, genetic algorithm, and quantum-inspired evolutionary algorithm.

Wireless sensor networks (WSNs) are composed of sensor nodes, having limited energy resources and low processing capability. Accordingly, major challenges are involved in WSNs Routing. Thus, in many use cases, routing is considered as an NP-hard optimization problem. Many routing protocols are based on metaheuristics, such as Ant Colony Optimization (ACO) and Particle Swarm Optimization (PSO). Despite the fact that metaheuristics have provided elegant solutions, they still suffer from complexity concerns and difficulty of parameter tuning. In this paper, we propose a new routing approach based on Teaching Learning Based Optimization (TLBO) which is a recent and robust method, consisting on two essential phases: Teacher and Learner. As TLBO was proposed for continuous optimization problems, this work presents the first use of TLBO for the discrete problem of WSN routing. The approach is well founded theoretically as well as detailed algorithmically. Experimental results show that our approach allows obtaining lower energy consumption which leads to a better WSN lifetime. Our method is also compared to some typical routing methods; PSO approach, advanced ACO approach, Improved Harmony based approach (IHSBEER) and Ad-hoc On-demand Distance Vector (AODV) routing protocol, to illustrate TLBO's routing efficiency.

Lately, the world is looking forward to replacing fossil fuels with renewable energy resources. Photovoltaic (PV) is benefited among renewable energies by its simple design, easy fitting, and fast expansion. For proper selection of controller gains for the PV system's control unit, this paper introduces a new hybrid optimization algorithm. The proposed optimization technique is applied to a PV system to enhance its performance under dynamic irradiance. The introduced hybrid optimization approach includes two optimization algorithms which are Teaching Learning Based Optimization (TLBO) and Equilibrium Optimizer (EO). Moreover, this paper presents a comparison with different optimization algorithms (Harmony Search (HS), Teaching Learning Based Optimization (TLBO), Cuttlefish Algorithm (CFA), and Emperor Penguin Optimizer (EPO)) to ensure the superiority of the newly proposed hybrid technique. The results display that the introduced TLBO-EO reduced the maximum overshoot by 30% and the steady-state error by 93% when compared to other techniques.

Large‐scale renewable energy (RE) integration in the power systems is displacing conventional sources of generation, thus reducing the system's inherent inertia and primary frequency response (PFR) capabilities. This reduced response may decline the system's ability to stabilise frequency after inevitable events such as loss of generation or load and thus requires additional PFR. To meet the PFR requirements, fast‐acting resources like energy storage systems (ESS) and demand response (DR) (transferrable and interruptible loads) need to be integrated within the system. In this regard, a novel stochastic security‐constrained unit commitment framework is proposed to assess the potential of ESS and DR to ensure additional PFR requirements and handle post contingency frequency dynamics. Moreover, the proposed methodology analyses the impact of uncertain RE integration on system frequency dynamics using Bayesian inference statistics. Case studies are carried out on 39 bus New England and 118 bus IEEE test systems, incorporating post fault frequency security parameters. Simulation results analyse the PFR contribution of ESS and DR and indicate a substantial reduction in system operating cost and RE curtailment with their integration. The proposed analysis would help system operators to maintain required PFR under high RE penetration, and handle unavoidable contingencies.

Strategy for Renewable Energy Storage in a Dynamic Distributed Generation System

Performance evaluation and extraction of global power maxima under shading scenarios and load Variation: An experimental study

The challenges of achieving a 100% renewable electricity system in the United States

With the rapid development of electric vehicles (EVs) and renewable energy, the charging scheduling of EVs with renewable energy becomes a hot issue. Nevertheless, as both the renewable energy and charging demand of EVs are hard to predict, it is difficult to get the optimal charging strategy of the EVs charging problem with renewable energy. In this paper, we consider a multi-EVs charging problem with a discontinuous charging process, where the charging decision for each EV is constrained. We model the problem as Markov decision processes (MDPs) and propose a novel Deep Reinforcement Learning (DRL) method called Constrained Double Deep Q-learning Network (CDDQN) to solve the MDP problem with large state space and decision constraints. Compared with the most of existing DRL methods, the CDDQN method embeds an action constraint model to a double deep Q-learning network (DDQN), which decreases the error of Q-value estimation and improves the accuracy of the charging policy by generating more effective training data. We conduct experiments on a simulation study and compare the proposed method with other DRL methods and experience charging policies. The quantitative results show that the proposed method achieves superior performance on getting the charging policy and outperforms the other methods in getting the minimum charging cost obviously.

Power Flow Management in Multi-Source Electric Vehicle Charging Station

To promote the use of renewable energy (RE), several types of RE policies have been implemented globally. Among these, feed-in tariff (FiT) is one of the most accepted mechanisms of pricing policy. However, choosing the best policy for a market with a high RE penetration is a challenge. A case of Bhutan is considered in this paper as the source of electrical energy is predominantly hydropower. Additionally, a generous subsidy is provided by the government to keep the electricity tariff to a minimum. Recently, there has been an increase in the advocacy and motivation for other forms of RE sources (RES) to supplement hydropower and increase energy security in Bhutan. Bhutan aims to achieve a total of 20 MW of non-hydro RES by 2025 as per the RE policy of Bhutan. However, Bhutan still does not have an RE pricing policy, and therefore, there is a need to institute a suitable pricing mechanism to accommodate the penetration of the planned non-hydro RES. This paper discusses the challenges in introducing FiT for non-hydro RES in an electricity market dominated by hydropower in Bhutan. Subsequently, recommendations are made in the wake of subsidised electricity tariff, which is the lowest in the region at 0.0171 USD/kWh for low voltage customers. FiTs for solar photovoltaic based on different categories of customers have been computed and proposed.