The spatiotemporal energy-shifting and moving flexibility of mobile energy storage (MES) can be explored to effectively support the operation security and resilience of distribution network. With this regard, a discrete multilayer spatiotemporal coupled network is first established to model the power charging/discharging, energy-shifting and moving-trajectory coupling of MESs; then a two-stage optimal operation strategy of distribution network simutaneously considering the operation flexibility of MESs and distributed generators (DGs) is proposed. In the first stage, the optimized connecting points of MES and optimal power output of DGs are optimized to achieve the minimum distribution network load shedding cost during fault; in the second stage, MES moving trajectory and charging/discharging power, as well as the power output of DGs are collaboratively optimized to minimize the comprehensive operation cost of the distribution network in post-fault. Finally, the effectiveness of the proposed two-stage optimal operation strategy considering MES flexibility was verified by the IEEE 33-bus network. Compared to scheme 2 with location-fixed energy storage and scheme 3 with DGs, the proposed approach reduced the load shedding costs by 4.5% and 2.2%, with the distribution network comprehensive operation costs decreased by 3.7% and 10%, respectively.

Electricity distribution network system is considered one of the key component of the modern electrical power system. Due to increase in the energy demand, penetration of renewable energy resources into the power system has been extensively increasing in recent years. More and more distributed generations (DGs) are joining the distribution network to create balance in the power system and meet the supply and demand of consumers. Today, large amount of DGs inclusion in the distribution network system has completely modernized power system resulting in a decentralize electricity market. Hence, Government of UK is pressurizing 14 distribution network operators (DNOs) to include more DGs into their distribution network system. DGs inclusion in the network system might be helpful due to many factors, but it creates many challenges for distribution network system in the long term. The network security is realized to be one of the challenge that impact the efficiency of accurate calculation and distribution of network pricing among consumers. To address the aforementioned issue, this research analysed the network security on the basis of Long run incremental cost (LRIC) pricing to balance and reduce the network pricing for the DNOs in UK. However, this study presented an approach of Deep reinforcement learning (DRL) also called deep reinforcement learning algorithm (DQN) to optimize the reactive power values in the network to balance and reduce the network pricing while keeping the network security. The method considers IEEE14 bus as its mathematical model and practically simulates the method in MATLAB using DQN algorithm pseudo codes. The network security has been analysed with and without security factor before and after the nodal injection into the network.

Building energy consumption prediction is of great significance to realize intelligent decision-making of energy system and improve energy efficiency.A random forest (RF) prediction model optimized via the particle swarm optimization (PSO) algorithm is established to forecast the hourly electricity consumption of the building cluster consisting of interconnected multiple buildings.The accuracy, generalization and robustness are taken as evaluation indexes.In the case study, the building cluster located in Austin is adopted as an example to explore the predicted performance of the proposed PSO-RF model in different seasons.The results show that the hourly electricity consumption PSO-RF model of the building cluster can achieve highest accuracy, strongest generalization, and best robustness, compared with RF, decision tree (DT), XGBoost, and k-Nearest Neighbor (KNN) prediction models.Therefore, the proposed hybrid model can be used as a reliable tool for building cluster electricity consumption prediction and energy management.

This paper focuses on the application of demand response (DR) in electricity customer classification and management, puts forward universal DR level index and load curve level index, constructing the customer classification index system. Principal component analysis (PCA) is used to reduce the original data dimension. Then the two methods for clustering are adopted and compared with the results. An example will be used to illustrate the effectiveness of the advantageous algorithm and the feasibility of the model. Based on the result of customer classification, discussing on the problems of customer management in China's electricity market, including market access and different types of customer incentive schemes. The work will be conducive to the promotion of DR and the development of the electricity market in China.