In this paper, a multi-objective, multi-level model is proposed for active distribution system (ADS) expansion planning with high penetration renewable energy sources (RESs) and energy storage systems (ESSs). To optimize the planning of RESs, ESSs and distribution networks cooperatively, a three-level optimization method is adopted based on the leader-follower strategy of hierarchic optimizations. In this model, the upper level and the middle level serve to model the planning problems from different perspectives of multi-stakeholders; the lower level serves to model the operation aspect of ESSs. The multi-level model enables us to integrate operation optimization into planning model and achieve the collaborative optimization of them in different time-scales. The multi-scenario tools and K-means clustering are adopted to deal with the uncertainties and capture the time-variable nature of RESs and load demand. In order to balance the multiple objectives of costs reduction, reliability improvement, and RES penetration promotion, a modified Pareto-based particle swarm optimization (PSO) is employed to solve the proposed optimization problem. Finally, results obtained by case studies are presented and discussed, where the availability and the effectiveness of the proposed planning model are verified.

Due to the widespread deployment of distributed energy resources (DERs) and the liberalization of electricity market, traditional distribution networks are undergoing a transition to active distribution systems (ADSs), and the traditional deterministic planning methods have become unsuitable under the high penetration of DERs. Aiming to develop appropriate models and methodologies for the planning of ADSs, the key features of ADS planning problem are analyzed from the different perspectives, such as the allocation of DGs and ESS, coupling of operation and planning, and high-level uncertainties. Based on these analyses, this comprehensive literature review summarizes the latest research and development associated with ADS planning. The planning models and methods proposed in these research works are analyzed and categorized from different perspectives including objectives, decision variables, constraint conditions, and solving algorithms. The key theoretical issues and challenges of ADS planning are extracted and discussed. Meanwhile, emphasis is also given to the suitable suggestions to deal with these abovementioned issues based on the available literature and comparisons between them. Finally, several important research prospects are recommended for further research in ADS planning field, such as planning with multiple micro-grids (MGs), collaborative planning between ADSs and information communication system (ICS), and planning from different perspectives of multi-stakeholders.

A fuzzy multi-objective bi-level optimization problem is proposed to model the planning of energy storage system (ESS) in active distribution systems (ADS). The proposed model enables us to take into account how optimal operation strategy of ESS in the lower level can affect and be affected by the optimal allocation of ESS in the upper level. The power characteristic model of micro-grid (MG) and typical daily scenarios are established to take full consideration of time-variable nature of renewable energy generations (REGs) and load demand while easing the burden of computation. To solve the bi-level mixed integer problem, a multi-subgroup hierarchical chaos hybrid algorithm is introduced based on differential evolution (DE) and particle swarm optimization (PSO). The modified IEEE-33 bus benchmark distribution system is utilized to investigate the availability and effectiveness of the proposed model and the hybrid algorithm. Results indicate that the planning model gives an adequate consideration to the optimal operation and different roles of ESS, and has the advantages of objectiveness and reasonableness.

The increasing integration of renewable energy generation (REG) makes the capacity value (CV) estimation for the REG become more important in the process of generation expansion planning. However, most researches on the CV ignore the hybrid generation system based on multiple complementary energy resources, as well as the operation strategy of energy storage system (ESS). To solve these problems, the models of available capacity of ESS are developed based on three different strategies. Then, using the sequential Monte-Carlo simulation method, the models and methods are developed for reliability assessment of generation system and capacity value estimation for the hybrid power generation system based on WT/PV/ES (HGS-WPE). By the application in IEEE-RTS79, the feasibility of models and estimation methods are verified. Finally the impacts of the complementation of HGS-WPE, the capacity and operation strategy of ESS on the reliability assessment and capacity value estimation are analyzed.

In this paper, a convolutional neural network (CNN) detection analysis is performed for live face detection by binocular cameras, and a binocular stereo matching network with fused edge detection is designed and implemented to target the quality of image details and parallax prediction at edges. Experiments show that the random sample pair confusion loss function can effectively improve the accuracy and generalization of the face live detection algorithm; the multi-task training approach can improve the performance of both faces live detection and face recognition; the algorithm shows excellent performance in both faces live detection and face recognition, especially the generalization of face live detection is greatly improved. A pre-trained convolutional neural network is used to extract features, and a content loss function and a domain loss function are designed to measure the feature distance between two images, and a feedforward neural network is trained as an image transformation network to migrate samples to the same domain. Experiments show that the algorithm can reduce the feature differences between the face live detection data of the two domains and can be used to improve the generalization of the face live detection algorithm.