Flatness Pattern Recognition Based On Wavelet Denosing And Intelligent Optimizition Algorithm

In recent years,as the core of the fourth industrial revolution,artificial intelligence plays an increasingly important role in the field of pattern recognition,which is more and more widely used in the speech,natural language processing,computer vision,machine learning,pattern recognition,knowledge classification,tracking and other aspects.Every aspect of human society is gradually changed by artificial intelligence.As the supporting industry of the national economy,the iron and steel industry has always been a great concern.With the progress of science and technology,the requirements from all walks of life of the quality of strip are becoming more and more precise.The key of rolling high quality strip is to improve the accuracy of flatness pattern recognition.The recognition method of artificial intelligence is applied to flatness pattern recognition,to improve the accuracy of flatness pattern recognition and the speed of recognition.Based on this,the main research of this paper is as followed:Among the flatness signal there is noise signal.Firstly,bivariate wavelet threshold method is used to eliminate noise which can overcome the disadvantages of the soft and hard threshold function method in dealing with wavelet coefficients,and better de-noising effect can be got.In order to improve the accuracy of pattern recognition,a pattern recognition model based on support vector machine is proposed.Support vector machine has strong learning ability and generalization ability in tackling small sample,and dealing with nonlinear and high dimension modeling problem.The flatness signal is discretized after de-noising,and as learning samples of support vector machine.The cuckoo algorithm is introduced to optimize the parameters of support vector machine.The simulation results show that compared with the particle swarm,genetics and grid search algorithm,the cuckoo optimization algorithm need less matching parameters,and the better optimal solution is obtained.Considering the fact that the singularity of signal is also important while removing the noise in the actual signal,we propose adaptive threshold wavelet function to remove noise.In view of the problem that support vector machines are only suitable for single input and single output systems,a flatness pattern recognition model of terminal slidingmode fuzzy neural network is proposed.Using the weight regulation law based on terminal sliding mode to replace the weight adjustment law based on the gradient descent method,which can improve the accuracy of the network.In order to improve the accuracy and speed of convergence further,cuckoo algorithm is introduced to optimize the structure parameters of the fuzzy neural network.The simulation results show that the proposed recognition model can achieve the better recognition results.