Research On Intelligent Prediction Model Of Plate Straightening Based On Artificial Neural Network

With the continuous innovation of medium and heavy plate production technology,customers’ requirements for medium and heavy plates are getting higher and higher.In the medium and heavy plate production line,TMCP(Thermo Mechanical Control Process)process is currently used in the mainstream,resulting in low temperature,high yield strength and unbalanced heating and cooling when the plate enters the leveler.In the medium and heavy plate production line,the straightening machine is an important equipment to ensure the flatness of the plate,and it also affects the performance and productivity of the product,and the old straightening machine and control system are difficult to meet the requirements of the new process.The process flow involved in the production line of plate straightening machine is multi-parameter coupling process.In the past,the parameter setting is usually set through manual experience,which will inevitably have a large error.The development of artificial neural network makes it possible for people to help us set process parameters through previous experience data.Artificial neural networks are an attempt to mimic the capabilities of the human brain by connecting artificial neurons to each other.The weights and biases of transfer functions in the network are updated iteratively to obtain specific inputs and outputs.Based on the background of the thirteen-roll straightener and its intelligent database,which have been relatively mature in the industrial field,this paper analyzes the current industrial situation and background in the field of straightening and discusses the existing relatively mature straightening models.Aiming at the problems of low precision in the current traditional straightening field,an artificial intelligence model of medium and heavy plate straightening machine is proposed.In this paper,BP neural network is used to train and verify the existing collected data sets,and on this basis,an improved cuckoo(ICS)optimization algorithm is proposed.Compared with the traditional BP neural network,it improves the accuracy of data prediction.accuracy,reducing prediction errors.On this basis,this paper also uses a onedimensional convolutional neural network for model prediction.The results show that compared with the traditional BP neural network,higher prediction accuracy can be achieved.By combining the artificial intelligence model with the traditional straightening machine,this paper predicts the straightening force in the straightening process of medium and heavy plates and provides technical support for the setting of process parameters in the actual production process.The intelligent model of the straightening machine will effectively improve the straightening efficiency and accuracy of the straightening machine and play a pivotal role in improving the yield of industrial production.