Optimized methods for morphological design of mesoscale cities based on performance analysis: Taking the residential urban blocks as examples

Abstract Urban block morphology is essential to the sustainable urban development at mesoscale. The performance targets should be considered in urban block morphology design. This study established an automatic computational process of morphology generation and simulation based on Grasshopper to evaluate the energy consumption (EC) and solar radiation access (SRA) of residential urban blocks. The process combined pattern recognition and data analysis to optimize the control and guidance for urban block morphology design. The study found that 1) SRA is more sensitive to morphology variation than EC is; 2) morphology control indexes (MCIs), such as plot ratio, density, and height, have significant but different effects on performance control indicators (PCIs); 3) multivariate nonlinear regression models can well predict the SRA_MV, with an R2 = 0.90; 4) the differences in morphological features for lower EC and higher SRA can be attributed to the number and distribution of towers; 5) the holistic performance (HP) of a block with more total buildings and fewer buildings of >15 floors is better; and 6) for HP optimal samples, the number of buildings and the distribution of height are more sensitive to MCIs. These conclusions provide references for the morphological design of sustainable cities at mesoscale.