Data and code: multi-objective day-ahead dispatch of a park microgrid with battery rainflow ageing and outage resilience (dispatch-structure search operators and multi-stakeholder benefit evaluation)

Data and code accompanying the manuscript Multi-objective day-ahead dispatch of a park microgrid with battery rainflow ageing and outage resilience: dispatch-structure search operators and multi-stakeholder benefit evaluation (H. Yao, F. Liu, T. Zhao, J. Wei and H. Su). The record contains one multi-volume 7-Zip archive park-microgrid-dispatch-package.7z (25 MB volumes; put all volumes in one folder and open the .001 volume with 7-Zip, or run 7z x park-microgrid-dispatch-package.7z.001). It holds: data/: the open input records actually used (2019 subset of the Suzhou industrial-park load dataset, CC BY 4.0; two State Grid renewable stations, CC0), the processed 15-min case-study dataset (35,040 rows: park load, PV and wind per-unit output, Jiangsu 2025 time-of-use price), data-quality tables and machine-readable records of all externally sourced parameters (tariff, carbon factors, reliability statistics, cost references) with document titles, URLs and access dates. results/: raw outputs of every experiment: 13 metaheuristic configurations and the MILP-ε reference on 8 test days (20 runs each) with FE snapshots, the MO-ISSA ablation and hyper-parameter studies, the CEC2022 benchmark runs, the 365-day operation of the six scheme chains (daily Pareto sets, schedules, energy flows, rainflow statistics), the multi-stakeholder evaluation (decision matrix, AHP / CRITIC / combined weights, TOPSIS / anchored TOPSIS / RAFSI / VIKOR / GRA, bootstrap, judgement perturbation, SMAA, leave-one-scheme-out, weight sweeps) and results.json with every number quoted in the manuscript. code/: the complete Python source (data preparation, Numba dispatch simulator with rainflow ageing and outage Monte Carlo, MO-ISSA / NSGA-II / SMS-EMOA with the physics-informed repair and dispatch-structure Pareto local search, linearised MILP warm start, year-round simulation, evaluation, statistics and figure scripts) with a requirements file. figures/: all figures of the manuscript (600 dpi PNG and vector PDF). README.md (also uploaded separately) lists the files, the original data sources with DOIs and licences, and the step-by-step reproduction commands. Software: Python 3.12, NumPy, Numba, pymoo, HiGHS. Licence: derived data and result files CC BY 4.0 (please also cite the original datasets); source code MIT.