The detection of filling-to-packing switchover point during injection molding plays a crucial role in ensuring the quality of the molded parts.Traditional detection methods

such as using predetermined screw cushioning or checking injection time have the disadvantages of depending on human experience and failure of adaptation to process parameter variations.This study presents a filling-to-packing detection method using the theory of singularity detection based on wavelet transform modulus maxima

according to the jump feature at filling-to-packing point.It adopted the pressure profile of a previous batch to obtain wavelet decomposition scale and switching threshold etc.

and then used sliding window to conduct on-line detection of switch point.The experimental results indicated that the innovative switchover method yielded a more uniform product mass than any traditional methods.