Adaptive optimization for high dimensional interference array jet distribution: thermal-field-feedback and multi-hole-linkage

To address engineering challenges of high-dimensional interference array jets, including the curse of dimensionality in the parameter space, high optimization cost, and difficulty in obtaining the optimum, a thermal-field-feedback-based multi-hole-linkage adaptive optimization method is proposed. This strategy transforms the high-dimensional independent search into a physics-guided adjustment and coordinates the interactions among the jet holes for hotspot suppression by adaptively updating jet-hole positions, significantly alleviating the curse of dimensionality and thus achieving an optimized engineering solution at an acceptable computational cost. The optimized configuration evolves from a uniform square to a hotspot-converging, non-uniform diamond distribution. For DJCP-16, under all inlet velocities tested (0.5 to 4 m/s), the optimized jet cooling plate reduces the maximum temperature T max by an average of 11.03 °C (21.69%) and temperature dispersion T σ by an average of 1.26 °C (32.82%), with only a 2.36% average increase in energy consumption. As the hole number increases, marginal performance gain diminishes under conventional heat flux. For DJCP-100 (250000 W/m 2 ), the reduction in T max is only 2.47 °C (6.85%). However, under a higher heat flux (1250000 W/m 2 ), DJCP-100 achieves a substantial T max reduction of 12.38 °C (12.31%). For the given environmental parameters, this method requires only one meshed model and one adaptive simulation to obtain an optimized distribution, greatly shortening the design cycle. This feedback-driven adaptive adjustment of the jet positions precisely matches the cooling to the hotspots, achieving efficient energy utilization at low energy cost.