Accurate estimation of battery degradation cost is one of the main barriers for battery participating on the energy arbitrage market. This paper addresses this problem by using a model-free deep reinforcement learning (DRL) method to optimize the battery energy arbitrage considering an accurate battery degradation model. Firstly, the control problem is formulated as a Markov Decision Process (MDP). Then a noisy network based deep reinforcement learning approach is proposed to learn an optimized control policy for storage charging/discharging strategy. To address the uncertainty of electricity price, a hybrid Convolutional Neural Network (CNN) and Long Short Term Memory (LSTM) model is adopted to predict the price for the next day. Finally, the proposed approach is tested on the the historical UK wholesale electricity market prices. The results compared with model based Mixed Integer Linear Programming (MILP) have demonstrated the effectiveness and performance of the proposed framework.

The paper considers the optimal power flow (OPF) of a meshed AC/DC power transmission network with voltage source converter based multi-terminal DC (VSC-MTDC) networks. The OPF problem is formulated to minimize the transmission loss of the whole AC/DC network with two different VSC control strategies considered, constant DC voltage control (master-slave control) and DC voltage droop control. In addition, Grid Code compliance of wind farms is also embedded in the OPF formulation. The presented OPF is evaluated and demonstrated in the paper by two example meshed AC/DC power systems.

Frequency deviation of power systems caused by grid-connected wind power fluctuations is one of the key factors which restrains the increase of wind penetration level. This paper examines a combined wind and hybrid energy storage system (HESS, supercapacitor and battery) to smooth wind power fluctuations without wind spillage. A fuzzy based wind-HESS system (FWHS) controller is proposed to suppress the wind power fluctuations. The proposed controller takes full advantage of the complimentary characteristics of the supercapacitor and battery with the supercapacitor and battery in charge of high and middle frequency components of wind fluctuations respectively. The efficient state of change (SoC) operation is considered. A differential evolution (DE) based optimal sizing method for wind-HESS systems is introduced to evaluate the minimum capacity of HESS as being limited by grid frequency deviation. The efficiency of the proposed scheme in the paper for wind-HESS system is evaluated by an real Chinese power system.

A frequency-based approach is proposed in this paper to size a battery-supercapacitor energy storage system for maintaining power balance of an isolated system with high penetration of wind generation, thus to maintain the grid frequency stability with the stochastic wind power fluctuations being considered. The sizing method proposed makes full use of the combined technical merits of two types of energy storage systems, battery and supercapacitor, as their specific power/energy densities and the corresponding expected life cycle cost/year are different. The power imbalance between hourly dispatched wind power and the corresponding actual real-time wind power output is decomposed into different time-varying periodic components by use of the Fourier transformation. Then two different types of energy storage systems are applied to balance the low-frequency and intermediate-frequency component respectively. A computational procedure is developed in the paper for choosing the optimal cut-off frequencies in determining the range of low and intermediate frequency for sizing the hybrid energy storage system. Comparative studies of a real isolated system in China are presented in the paper to demonstrate the effectiveness of the proposed method.

This paper presents an improved corrective security constrained optimal power flow (CSCOPF) for a meshed AC/DC power transmission network with voltage source converter based multi-terminal HVDC (VSC-MTDC) grids. The security constrains of N-1 contingencies for both AC and DC networks are fully considered in the CSCOPF problem. The corrective actions employed in this paper include not only the traditional long-term rescheduling control actions of the AC system, but also the short term post-contingency fast corrective control of multi-terminal VSC-HVDC. A hybrid method is developed to solve the improved CSCOPF problem and the effectiveness of the proposed method is evaluated and demonstrated in this paper by two example meshed AC/DC power systems.