Abstract
Aiming at the weakness of the existing cloud neural network on training and practicality, a new improved structure of cloud neural network is designed. A hidden layer is added prior to the inverse cloud layer. Threshold level is set to zero and a simple training method is designed. In addition, considering the ignorance of signal randomness and fuzziness in the existing method of the flatness signal recognition, the cloud neural network combines the advantages of the fuzziness and randomness of cloud model and the learning and memory ability of neural network. Thus it is applied in the flatness signal recognition. The simulation contrast results demonstrate that the improved structure is able to identify common defects in shape with higher identity precision.
Similar content being viewed by others
References
Bordignon F, Gomide F (2014) Uninorm based evolving neural networks and approximation capabilities. Neurocomputing 127:13–20
Jia CY, Shan XY, Liu HM (2008) Fuzzy neural model for flatness pattern recognition. J Iron Steel Res 15(6):33–38
Jung JY, Im YT (1997) Simulation of fuzzy shape control for cold-rolled strip with randomly irregular strip shape. Mater Process Technol 63(1–3):248–254
Li DY, Liu CY, Liu LY (2004) Study on the universality of the normal cloud model. Eng Sci 6(8):28–34
Liu CY, Feng M, Dai XJ (2004) A new algorithm of inverse cloud. J System Simul 16(11):2417–2420
Liu J, Wang YQ, Sun F (2008) Fuzzy pattern recognition method of flatness based on particle swarm theory. Chin J Mech Eng 44(1):173–178
Lu XW, Peng Y, Liu HM (2011) Gauge and shape control strategy for DC mill. J Central South Univ 42(8):2309–2317
Peng Y, Liu HM, Wang DC (2007) Simulation of type selection for 6-high cold tandem mill based on shape control ability. J Central South Univ Technol 14(2):278–284
Ren HP, Liu D, Zheng G (2002) A genetic algorithm approach to shape pattern recognition. Heavy Machin 3:9–12
Rubio JJ (2014a) Analytic neural network model of a wind turbine. Soft Computing. doi:10.1007/s00500-014-1290-0
Rubio JJ (2014b) Evolving intelligent algorithms for the modelling of brain and eye signals. Appl Soft Comput 14:259–268
Shan XY, Liu HM, Jia CY (2010) A recognition method of new flatness pattern containing the cubic flatness. Iron Steel Res 45(8):56–60
Silva AM, Caminhas W, Lemos A, Gomide F (2014) A fast learning algorithm for evolving neo-fuzzy neuron. Appl Soft Comput 14(8):194–209
Su YH (2007) Cold-rolled strip production and prospects of China. Metall Inf Rev 5:44–48
Tian YQ (2003) Research on theory and application of data mining based on cloud model. Shanghai Jiaotong University, Shanghai
Xu LJ (2007) Flatness control in cold strip rolling and mill type selection. Metallurgical Industry Press, Beijing
Xu ZB, Fan ZZ (2009) Cloud neural network based fault detection and diagnosis of space propulsion system. Acta Armamentarii 30(6):727–732
Zhang XL, Liu HM (2003) GA-BP model of flatness pattern recognition and improved least-squares method. Iron Steel Res 38(10):29–34
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by V. Loia.
Rights and permissions
About this article
Cite this article
Zhang, Xl., Zhao, L., Zhao, Wb. et al. Novel method of flatness pattern recognition via cloud neural network. Soft Comput 19, 2837–2843 (2015). https://doi.org/10.1007/s00500-014-1445-z
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00500-014-1445-z