Distribution Network Target Framework Planning Algorithm Based on Fuzzy Optimization and Grey System Theory

INTRODUCTION: The global energy transition, driven by the rapid growth of distributed renewable energy, stochastic load profiles (e.g., EV charging spikes), and conflicting stakeholder objectives, has brought unprecedented complexities to distribution network planning. Traditional deterministic methods fail to handle qualitative fuzziness (e.g., subjective reliability thresholds) and quantitative uncertainty (e.g., sparse historical data), leading to inflexible and inefficient solutions. This study addresses these challenges by developing a hybrid planning framework. OBJECTIVES: This paper aims to solve the dual challenges of qualitative fuzziness and quantitative uncertainty in distribution network planning, providing a systematic solution to accommodate distributed renewable energy, handle load uncertainty, and balance conflicting stakeholder preferences through integrating fuzzy optimization theory and grey system theory. METHODS: The hybrid algorithm combines fuzzy optimization and grey system theory. Fuzzy optimization uses triangular fuzzy numbers for load growth rates ([3%, 5%, 8%]) and trapezoidal fuzzy intervals for voltage constraints ([−10%, −5%, 5%, 10%]) with membership functions (threshold λ≥0.8) to convert qualitative requirements into solvable constraints. Grey system theory applies the GM(1,1) model for load forecasting (achieving 4.2% MAPE with 15-month data) and grey relational analysis (GRA) for data-driven objective weighting to eliminate expert bias. An improved particle swarm optimization (IPSO) algorithm is used for optimization, validated in a 33-node network with 8.5 MW PV and 6 MW wind capacity. RESULTS:  In the 33-node case study, compared to the deterministic genetic algorithm (D-GA), the hybrid algorithm reduces lifecycle costs by 19% (from $8.91M to $7.23M), increases renewable energy accommodation by 24% (from 9.8 MW to 12.3 MW), and improves the system average supply availability index (ASAI) from 99.92% to 99.95%. Under extreme uncertainties (±40% renewable output, ±30% load shifts), cost deviations remain within 6% and reliability metrics within 5%, demonstrating strong robustness. CONCLUSION: This research presents a robust hybrid framework that bridges fuzzy qualitative reasoning and grey data efficiency, effectively addressing both qualitative fuzziness and quantitative uncertainty in distribution network planning. It provides a science-based tool for resilient grid design, with potential for extension to multi-energy system integration and real-time optimization in future work.