Data for: Collaborative Parameter Optimization of Coal-Fired Integrated Energy System with Carbon Storage and Shiftable Electrical Load

Results were produced in Python (NumPy, Numba, pandas, openpyxl, SciPy) with a two-layer framework: an outer adapted particle swarm optimization (30 particles × 35 iterations, 30 independent runs per algorithm) sizes the storage, an inner marginal-cost iterative dispatch determines minute-level operation, and a grid search (10 MW / 10 min steps) optimizes load shifting; regressions and an independent verifier complete the analysis.The data cover one representative winter weekday at a uniform one-minute resolution (1440 records, Time_min = 1–1440); the only spatial object is the single unit in North China, with no geographic coordinates.

In each time-series sheet, rows are minute indices and columns are named variables, with units given by suffix—power in MW, storage mass in tonnes (t), energy in MWh, cost in CNY, tariff in CNY/kWh, duration in min; IsValley/IsPeak/IsFlat are 0/1 period flags. There are no missing values (input AGC range 200.44–277.81 MW). The model is deterministic apart from PSO initialization (CV = 0.027% over 30 runs); fitted rules report R² and sample size n, and displayed values are rounded to two decimals.AGC_Load.xlsx is the sole input.