A rough set data prediction method based on neural network evaluation and least squares fusion

To improve precision of heat insulation performance analysis algorithm of composite wall in hot summer and cold winter zone, a kind of rough-set predicative analysis method for heat insulation performance of composite wall in hot summer and cold winter zone is put forward. Firstly, structure and parameter for three kinds of walls adopted in research of this Paper are provided and corresponding heat-insulation physical model and problems are described; secondly, introducing rough set, combining with its advantages in disposing uncertain and inaccurate and incomplete data, evaluation analysis for insulation performance of composite wall has been realized; finally, the validity of proposed method is verified by simulation experiment.