Parametric Simulation Dataset for Multi-Objective Optimization of a University Classroom Envelope in Harbin (Severe Cold Region, China)

300 parametric design cases of a typical university classroom in Harbin, China (9.0 m x 7.2 m x 3.9 m, south-facing, 45 occupants, teaching schedule 8:00-21:00), generated via parametric building energy simulation with the Harbin Typical Meteorological Year as boundary conditions. Input variables: external wall (4 types), roof (3 types), glazing (3 types), window-to-wall ratio (0.20-0.60), window height-to-width ratio (0.50-2.00). Output indicators: envelope construction cost (CNY), annual heating energy (kWh/a), annual cooling energy (kWh/a), and thermal discomfort hours during occupied periods (|PMV| > 0.5). Continuous variables were sampled by Latin Hypercube Sampling and discrete variables by stratified random sampling. The dataset supports surrogate model training, multi-objective optimization benchmarking, and Pareto trade-off analysis for educational building envelope design.