A Soft Sensing Method for Temperature of Large-Scale Aging Furnace Forgings Based on Grey System Theory

摘要

Setting proper holding time is very important to ensure the product quality during the production of large-scale aluminum forgings in the aging furnaces. The obtaining of the actual forging temperature is a prerequisite for setting the proper holding time. Due to the special furnace structure and limit of the aging technique, the forging temperature cannot be measured directly. According to the above problem, a traditional grey model GM (1, N) was proposed where the symbol GM (1, N) stands for first-order grey model with N variables, forecasting the temperature of the forgings, getting good prediction effect. But the control precision remains to be improved. Finally, an improved GM (1, N) model was established by combining the equal dimension and new information model with residual modification on the basis of the experimental data. The simulation results show that the improved model has higher prediction precision.