In specific condition, when wrapping angle of cold rolling strip covering surface of shape detecting roll dynamically changes, online radial compression of the shape detecting roll is changed too, and it seriously affects online shape detecting precision of cold strip. Based on the latest intelligent shape meter developed by Yanshan University, using PSO-BP neural network and actual working condition datum, the cold strip online dynamic wrapping angle compensation model is established, and successfully applied in 1250 mm 6-high cold mill, remarkable results are achieved. The error between calculated values and measured values of total tensions is within 3 %
A high-precision shape detecting system of cold rolling strip is developed to meet industrial application, which mainly consists of the shape detecting roller, the collecting ring, the digital signal processing （DSP） shape signal processing board and the shape control model. Based on the shape detecting principle, the shape detecting roller is designed with a new integral structure for improving the precision of shape detecting and avoiding scratching strip surface. Based on the DSP technology, the DSP shape signal processing circuit board is designed and embedded in the shape detecting system for the reliability and stability of shape signal processing. The shape detecting system was successfully used in Angang 1 250 mm HC 6-high reversible cold rolling mill. The precision of shape detecting is 0.2 I and the shape deviation is controlled within 6 1 after the close loop shape control is input.
