A Multi-Parameter Collaborative Dimensionless Fan Selection Method Based on Efficiency Optimization

This paper proposes an efficiency-optimized multi-parameter collaborative non-dimensional selection method for industrial fans. Based on fan similarity theory, selection parameters are transformed into non-dimensional forms. The fan’s best working area (BWA) is defined according to stall margin, flow range, total pressure rise deviation, and minimum efficiency. The initial model selection uses the boundary equations of the defined BWA as screening criteria. Decision parameters comprise Euclidean distance, design point distance, pressure deviation, and current efficiency. These collectively form a multi-objective evaluation function. The NSGA-II algorithm determines the optimal weight distribution of decision parameters, generating a Pareto-optimal solution set. The initially selected models are subsequently subjected to secondary optimization through a comprehensive evaluation function. Selection case studies demonstrate that this method preliminarily screens 7 models that meet the target parameters from 400 candidate models. Secondary screening determines the model with the optimal efficiency and best comprehensive evaluation performance. The method effectively resolves the mismatch between fan model design points and target operational parameters in selection processes. This method integrates directly into selection software platforms and validation with 100 sets of fan selection parameters demonstrates that selected models achieve 99% accuracy. Achieving the secondary optimization function for fan model selection.