Reliability-Economy-Renewable Energy Hosting Capacity Collaborative Planning Method for Active Distribution Networks Based on Block-Based Reliability Assessment and Improved NSGA-II

Abstract

: With the large-scale integration of stochastic distributed resources, the power-flow distribution, fault restoration mode, and renewable energy accommodation boundary of distribution networks have changed significantly. Conventional reliability assessment methods are mainly based on component-level fault enumeration, which imposes a heavy computational burden in multi-feeder, multi-scheme, and multi-iteration optimization. Meanwhile, planning models that consider only reliability improvement or only renewable-energy hosting capacity may lead to biased investment decisions. To address these issues, this paper proposes a three-element collaborative planning method for active distribution networks that jointly considers reliability, economy, and renewable energy hosting capacity. First, a block-based reliability assessment algorithm is constructed. Using switching devices as boundaries, the active distribution network is equivalently partitioned into blocks; optimized node numbering, virtual-node insertion, and block-level equivalent parameters are adopted to reduce model complexity. Second, a three-element multi-objective planning model is established by incorporating life-cycle cost, the System Average Interruption Duration Index, and renewable energy hosting capacity into a unified planning framework. Third, an improved Non-dominated Sorting Genetic Algorithm II (NSGA-II) is developed, in which mixed-variable coding, feasibility repair, normalized crowding distance, and fuzzy satisfaction decision-making are embedded into the standard non-dominated sorting genetic algorithm, making the optimized solutions more suitable for engineering implementation. Case studies on four typical 10 kV feeders in Guangdong Province, China, show that the proposed method can simultaneously identify reliability weak links and renewable energy hosting-capacity enhancement potential under different feeder conditions, and can provide differentiated investment schemes. For the 10 kV Gantang Feeder F04 with relatively poor reliability, under a budget of Chinese Yuan (CNY) 8 million, the System Average Interruption Duration Index (SAIDI) decreases from 4.9196 to 1.0702 h/customer/year, representing a reduction of approximately 78.25%. For the F3 Qianwu and F4 Jinhao feeders, which exhibit high hosting-capacity enhancement potential, the hosting capacity of F3 reaches 100.00% under a budget of CNY 16 million, whereas F4 reaches 100.00% with an actual investment of CNY 10.9919 million. In addition, the block-based reliability assessment method substantially improves algorithmic efficiency. Overall, the proposed collaborative planning method can avoid investment bias caused by single-objective planning and support budget allocation and differentiated feeder governance in practical distribution network planning.