www.melanexjou.fora.pl

[loiwtjsf]

Analysis of network parameters on the fuzzy neural network prediction results of


01 No. l of Zhang Wenli et al: Analysis of network parameters on the fuzzy neural network prediction results of 83 when, with the suspension of the membership to learn the error increases,[link widoczny dla zalogowanych], forecasting a significant increase in the error sum of squares; when the value is less than 0.0001, the square prediction error and increased slowly, Take a value of 0.0000001, the prediction error sum of squares was 0.771, significant increase in learning time. Figure 9 cattle Gou tunnel suspend the membership to learn the error of Fig. 9Influenceofsubordinationdegreelearningintermissionerroronforecastresult not difficult to see from the above analysis, the network parameters on the forecasting accuracy of Modular have different degrees of impact, whether directly related to the optimal range of the forecast results. And on this basis,[link widoczny dla zalogowanych], identify the deformation resistance prediction 30MnSi the best network parameters (see Table 1) Table 1 the best network parameters and rules of prediction error squared and sample rate of a network of experts trained to learn the number of membership a few number of prediction error Learning suspension error square sum error learning suspended 5020000O. 10.00010.000i0.7118 reported accuracy of the analysis to identify the best network parameters to predict on this basis can achieve the best results. 5 Conclusions 1) a comprehensive analysis of the network parameters on the Modular algorithm program 30MnSi deformation resistance prediction of the forecast results of the situation, to find the best network parameters to improve the accuracy of Modular forecast. 2) pointed out that the fuzzy neural network to predict,[link widoczny dla zalogowanych], they should first analyze the network parameters on the prediction accuracy of the impact analysis, to find the best network parameters,[link widoczny dla zalogowanych], and then based on this forecast, in order to forecast the best results. [1] [2] [3] [4] [5] can be seen in the analysis, the use of fuzzy neural network [6] to predict the actual situation, the first thing to do network parameters on the pre-reference Dai Tiejun. Fuzzy neural network in 30MnSi deformation resistance prediction. Iron and Steel Research, 2001, (1): 33 ~ 35 Dai Tiejun. Deformation resistance of the fuzzy neural network model. Rolling, 18 (3): 6 ~ 7 Zhang Liming ed. Artificial neural network model and its application. Shanghai: Fudan University Press, 1994 Xue-bridge, Ma Li ed. Application of neural network engineering. Chongqing: Chongqing University Press, 1996, edited by Wang Shi-Tong. Neuro-fuzzy system and its application. Beijing: Beijing Aerospace University Press, 1998 Transplantation and Hemopurification ed. c programming. Beijing: Tsinghua University Press, 1993AnalysingtheinfluenceofnetworkparametersontheforecastresultsofANNZHANGWen-1i (HebeiEnergyInstituteofVocation & Technology.Tangshan063004) DAITi called un (HebeiInstiututeofScienceg> Technology, Tangshan063000) Abstract: Onthebasisoftestdata,[link widoczny dla zalogowanych], theeassyanalysestheinfluenceofnetworkparametersontheresultsthatmetal-plasticde-formationresistanceof30MnSisteelwaspredictedbyfuzzyneuralnetwork (ANN) algorithm. andseekingoptimumnetworkparameters, thusthepredictionaccuracywillberaised. Meanwhile. puttingforwardthenecessaryofthatanalysingtheinflu-enceofnetworkparametersontheforecastresult, whenadoptingtheANN. Keyword: fuzzyneuralnetwork; networkparameters; forecastresultLLLLL

相关的主题文章：


[link widoczny dla zalogowanych]

[link widoczny dla zalogowanych]

[link widoczny dla zalogowanych]