| Nowadays, hot-rolled steel producers concerned more about the quality of their products and flexible production capacity of production lines. The traditional steel production is faced with challenges, and the traditional technology needs to be improved. The necessary means to solve the above-mentioned problem is using the soft-sensing model in hot rolling. So, the paper establishes a soft-sensing model based on BP algorithm and support vector machine regression algotithm for actual needs.The primary work of establishing the soft-sensing model is arranging the data to be used in network training. Aiming at the shortcomings of site-collected data that it contains noise, the paper uses the data mining theory for reference, and deals with the site-collected data so that it contains less noise to provide reliable and sufficient data samples for the model. Then, based on the traditional experience, original chemical composition and hot rolling process parameters are selected from the data samples. According to principal component analysis conducted on the data to reduce the dimension.In the soft-sensor model,Through standard BP neural network, support vector machine regression soft measuring comparative study of two software measurement to improve their advantages and disadvantages and put forward a genetic algorithm to optimize BP neural networks, support vector machine regression based parameter optimization and giving results of two programs are compared. Experimental verification of support vector machines is better for small samples, network generalization ability is strong, predicted results is stable. Support vector machine regression based parameter optimization with genetic algorithm to optimize BP neural networks, the average error reduce30.2%, error rangesample reduce52.9%。Thus determine that support vector machine regression based parameter optimization is a more excellent way for Soft-Sensing of Hot-rolled Strips Mechanical Properties. |
PDF Full Text Request