Advanced Materials Research Vol. 1037

Abstract: A design layout optimization for deformable trimaran with the basis of DTMB5415 based on BP neural network is presented. With taking CFD software calculated results for its sample values, BP neural network can predict resistances of the unit displacement under different layout and speed through trained. From above, the best layout of the ship adjusted with the speed can be drawn on the basis of ensuring the calculation accuracy. Compared to traditional CFD method, BP neural network can not only greatly reduce the computational time but also search for the optimal layout corresponding to any speed quickly and reliably. CFD method limited the layout and speed to obtain the minimum resistance value, and BP neural network method broke this inherent mode above and it really play the “deformable” advantage. It can be better applied to engineering practice and provides a new idea for computer-aided layout optimization of triamran.

Abstract: With the development of computer technology, casting simulation plays an increasingly important role in foundry production. VOF (Volume of Fluid) method which tracks free surfaces is widely used in casting simulation because it is simple to solve and easy to achieve. However, in the last stage of mold filling process, the solution speed of VOF method is very slow because the space to fill becomes smaller. To solve this problem, interpolation formula is applied to the calculation of temperature field and an interpolation method to calculate the last part of temperature field in mold filling process is presented in this paper. The solution is to calculate temperature field of unfilled grids based on the temperature filed of filled grids at the last stage of filling process. Comparison between the results of interpolation method and theoretical calculation illustrates the rationality of the interpolation temperature field in this paper. Numerical simulations of actual castings indicate that the interpolation method can greatly improve the speed of getting temperature field under a certain accuracy. This paper has important implications to the calculation of numerical simulation in mold filling process.

Abstract: In this paper, we mainly address the problem of tracking a single ship in inland waterway CCTV (Closed-Circuit Television) video sequences. Although state-of-the-art performance has been demonstrated in TLD (Tracking-Learning-Detection) visual tracking, it is still challenging to perform long-term robust ship tracking due to factors such as cluttered background, scale change, partial or full occlusion and so forth. In this work, we focus on tracking a single ship when it suffers occlusion. To accomplish this goal, an effective Kalman filter is adopted to construct a novel online model to adapt to the rapid ship appearance change caused by occlusion. Experimental results on numerous inland waterway CCTV video sequences demonstrate that the proposed algorithm outperforms the original one.

Abstract: The adaptive Kalman filtering algorithm was adopted in the online estimate of navigation state of unmanned aerial vehicle (UAV) as the simplified model often used. At the moment, the alogorithms those usually applied in this territory are not perfect. Analysed the adaptive Kalman filtering based on Maximum-Likelihood Estimation and Sage-Husa Kalman filtering, take advantage the characteristics of residue, choose the estimation windows, a simplified adaptive Kalman filtering algorithm was gived.

Abstract: Solving large-dimensional multi-objective optimization problems is one of the focus research areas of multi-objective optimization evolutionary . When using traditional multi-objective optimization algorithms to solve large-dimensional multi-objective optimization problems,we found that the unsatisfactory optimizing results often exist. To overcome this flaw, in this paper we studied scalable dominant mechanism and proposed a D dominant strategy. According to the superior theory of D strategy ,we improved the current four kinds of typical multi-objective optimization evolutionary algorithms. The numerical comparison test on DTLZ1-6 (20) questions which were solved by the improved algorithms indicated that D strategy had in varying degrees improved the algorithms for solving large-dimensional multi-objective optimization problems .Thus ,we confirmed that the D strategy for solving large-dimensional multi-objective optimization problems is effective.

Abstract: in order to fully optimize the BP neural network, and make it have better generalization performance, we improve and design a genetic BP neural network, the algorithm is given, and their crossover operator was improved. And this method is applied in identifying lithology, the experimental results show that this method increases the algorithm convergence speed and effect, and it has certain practical value.

Abstract: Mathematical morphology can be seen as a special digital image processing method and theory, which has been widely used in various fields. In this paper, the mathematical morphology is applied to the color image processing. In thespace of color image, I have simply expounded the theories and properties of color morphological changes, and defined its morphological operators. According to the application of omni-directional and multi-angle structuring elements composite morphological filter in gray image, I put forward a kind of color morphological filter with omni-directional and multi-angle structuring elements composite. This algorithm has retained its advantages in gray image, however, remaining some drawbacks. Through the optimization of results based on this algorithm, we finally get the relatively ideal denoising effects.Keywords: mathematical morphology;color model;color model; color morphological filter

Abstract: Fault diagnosis is very important to ensure the safe operation of hydraulic generator units (HGU). Because of the complexity of HGU, the vast amounts of measured data and the redundant information, the accuracy and instantaneity of fault diagnosis are severely limited. At present, feature selection technique is an effective method to break through this bottleneck. According to the specific characteristics of HGU faults, this paper puts forward a hierarchical feature selection method based on classification tree (HFSMCT). HFSMCT selects the most effective feature for each branch node through filtering evaluation criteria and heuristic search strategy, and all the selected features constitute the final feature set. Moreover, HFSMCT is easy to design and implement, and it is very prominent in computational efficiency and accuracy. The simulation results also prove that HFSMCT is very suitable for HGU fault diagnosis.

Abstract: In order to optimize the railway freight transport network, integrate the limited transport resources and overcome the current problems existing in the traditional transport organization, in this study, we propose a three-layer railway freight transport network system, analyze its hierarchical structure and describe the respective function orientation of the railway freight stations in different layers. Then we design a BP neural network model with adaptive learning algorithm and momentum BP algorithm to classify the railway freight stations into three layers. Finally, an empirical case study is presented to test the feasibility of the BP neural network. The simulation result indicates that the BP neural network model can classify the railway freight stations into three layers under relatively high training accuracy.

Abstract: The fuzzy edge detection algorithm proposed by Pal-King has some disadvantages for extracting the low gray level edge information for the infrared images, such as high computation complexity, low threshold segmentation inaccuracy and the leakage edge information. For overcoming the disadvantages, the improved image fuzzy edge detection algorithm is proposed in this paper. First, redefining membership function to simplify computation complexity, the new conversion function enable the function transform interval is [0, 1], thus the value of the low gray level edge is not to be set to 0. Second, Ostu's algorithm is used in the selection of segmentation threshold named as transit point. The traditional threshold value is improved in order to make the segmentation accurate. The experimental results show that the lower gray infrared image edge information is preserved via proposed algorithm in this paper. The detecting results are more accurate. The run time is decreased obviously than the traditional Pal - king algorithm.