Papers by Keyword: BP Neural Network

Abstract: In order to improve the working efficiency of a manufacturing system, tool life estimation is very essential. In this paper, the dominant factors affecting tool life are analyzed by theoretical analysis. According to the nonlinear relationship between affecting factors and tool life, a tool life prediction model based on BP neural network, which is optimized by genetic algorithm (GA), is built up. 15 network patterns are trained to get the best network structure. The accuracy of GA-BP model is verified through computing and compared with the standard BP model. The results show that GA-BP model prediction value is exactly closed to the expected value of tool life and the prediction accuracy can be improved more than 5% compared than the standard BP model. The model is proved to be accuracy and it can be used as an effective method of tool selection decision.

Abstract: In order to separate wool from cashmere efficiently, an identification method based on texture analysis was proposed in this paper. The microscopic images captured by CCD digital camera were preprocessed as the texture image. Improved Tamura texture feature were employed to analyzing the final texture images and to attaining the texture parameters. Through a large number of samples, the mathematical modeling was completed by using neural network. Experiment results indicate that texture analysis can be a feasible method to identify cashmere and wool.

Abstract: Abstract. How to improve the precision of the stamping forming has been one of the stamping researchers concern in the stamping technology.This paper analysis the material stamping performance of AL5052 aluminum alloy sheet,study the influencing factors of forming precision,and find the the mapping relationship between influence factors and stamping forming to effectively predict the forming error .The first part is to analyze the factors that affect the forming process, then get the main factors that influence the part of stamping forming. And then establish the main influence factors^’ mathematical model of stamping forming error based on BP neural network, and through the training of BP neural network to prove the model^’s practicability.

Abstract: The plating parameters for optimizing the wear and corrosion resistance of Ni-TiC composite coatings were selected by orthogonal test, mainly including the TiC particles concentration, current density, duty cycle, frequency and stirring rate. A three-layer BP (Back Propagation) neural network with Lavenberg-Marquardt algorithm was established by MATLAB, which was used to train the network and predicted orthogonal experimental data. In addition, the best parameters combination of the composite coating were predicted and verified by experiments. The results predicted through the proposed BP model are in good agreement with the experimental values, the relative error is small, and the maximum error is less than 3% and the coefficient of determination value is 0.9997.

Abstract: Traditional risk assessment methods, mainly static analysis, can’t anticipate the development trend and consequences of accidents. This paper firstly presents 5M safety model, combined with the characteristics of the rail transit safety assessment, including complexity, dynamic, ambiguity, etc. By using neural network algorithm, this paper proposes a method of railway safety dynamic assessment. Finally, the method is validated by comparing the output value of the model and the veritable value. The result indicated that it has advantages of real time and predictability.

Abstract: In the field of object recognition, the SIFT feature is known to be a very successful local invariant descriptor and has wide application in different domains. However it also has some limitations, for example, in the case of facial illumination variation or under large tilt angle, the identification rate of the SIFT algorithm drops quickly. In order to reduce the probability of mismatching pairs, and improve the matching efficiency of SIFT algorithm, this paper proposes a novel feature matching algorithm. The basic idea is taking the successful-matched SIFT feature points as the training samples to establish a space mapping model based on BP neural network. Then, with the help of this model, the estimated coordinate of the corresponding SIFT feature point in the candidate image is predicted. Finally search the possible matching points around the coordinate. The experiment results show that using the prediction model, the number of mismatching points can be reduced effectively and the number of correct matching pairs increases at the same time

Abstract: According to the nonlinear dynamic characteristic of coal seam floor water inrush, coal seam floor water inrush risk evaluation which includes 4 first level indicators,14 level two indexes was built based BP neural network. According to the test collection of engineering data, coal seam floor water inrush risk evaluation system based VB and MATLAB is reliable. Application to a mine coal seam No.2 working face was verified. The results show that, the evaluation method in water inrush is feasible, reasonable.

Abstract: To solve the problems of high dimension, multicollinearity, easily over-fitting, this paper further studies characteristic wavelength selection method, and proposes the improved successive projections algorithm based on mean impact value algorithm (SPA-MIV). The results show that the proposed algorithm reduces data dimensions and improves data quality effectively. After being processed by improved successive projections algorithm, the determination coefficient of of testing set of moisture, ash and volatile PLS calibration models are increased to 0.9318, 0.9127, 0.9389, and the of testing set determination coefficient of moisture, ash and volatile BP neural network calibration models are 0.9645,0.9432,0.9536.

Abstract: The Ordovician fissure-cavern carbonate reservoirs of Tarim Basin have relatively large production thickness, long completion intervals and serious heterogeneity problems, former studies about physical parameters for fractured intervals of acid fracturing cannot be well presented through mean algorithm method. This paper uses mean square deviation method to acquire the values of the main influencing factors of acid fracturing, then optimizes the physical characteristics of the fractured intervals. The BP neural network simulation has been adopted to get the optimal BP neural network structure model. In Tarim, this study has simulated 20 wells before acid fracturing using the decision software which compiled by the optimum neural network construction. Comparative analysis has been made through application examples, the coincident rate of mean algorithm is 75%, compared with the 90% of mean square deviation method. Therefore, during the target selection for acid fracturing of heterogeneous carbonate reservoirs with long open hole intervals, it’s significant to use mean square deviation method to optimize physical parameters of reservoirs.

Abstract: Directing against the problems of too large size of the neural network structure due to the existence of a complex relationship between the input coupling factor and too many input factors in establishing model for predicting temperature of sunlight greenhouse. This article chose the environmental factors that affect the sunlight greenhouse temperature as data sample. Through the principal component analysis of data samples, three main factors were extracted. These selected principal component values were taken as the input variables of BP neural network model. Use the Bayesian regularization algorithm to improve the BP neural network. The empirical results show that this method is utilized modify BP neural network, which can simplify network structure and smooth fitting curve, has good generalization capability.