Advanced Materials Research Vols. 756-759

Abstract: In this paper, aiming at the characteristics of auto parts enterprises product development, the object-oriented case knowledge expression to build auto parts product development database is used, the nearest neighbor algorithm is as similarity algorithm of case retrieval, the TCP/IP network technology is as the main network transmission protocol, C/S is as the application system to build the case retrieval system, and EPS steering gear is as an example.

Abstract: The work of the gear transmission is very complex, and its failure in the form and features tend to show non-linear mapping. Fault signal is often submerged in conventional vibration signal and noise, it is not easy using traditional signal processing methods to extract fault features which in a difficult to gear fault diagnosis. This paper based on the genetic algorithm to optimize the structure of the BP neural network model for the intelligent diagnosis system which is used in gear fault diagnosis.The experimental results show that this method can be effectively used for the diagnosis and identification of the gears common fault type.

Abstract: In order to reduce the delays of train groups when suffering from railroad emergencies, we propose a rescheduling algorithm of train groups to reschedule trains, instead of previous just waiting. A heuristic algorithm is designed to search the proper path for the trains involved in the railroad emergencies. An evaluation criterion based on the extent of the disturbance towards the original timetable and limited time is designed to select the optimal path for each train. And a conflicts resolution strategy is designed to deal with the path and time conflicts among trains when rescheduling. In the end, due to the real-time needs of the railroad scheduling, we implement the parallel processing for the train-group rescheduling referring to the allocation of loads. Experimental results show that the rescheduling algorithm is efficient to reduce the delays of train groups and the parallel processing meets the real-time response of the train-group rescheduling well.

Abstract: Instead of performing the integration explicitly to calculate the magnetic field from an arbitrary shaped wire, an improved method was proposed. The magnetic field generated by a straight line segment carrying steady current was calculated based on the Biot-Savart law. The new approach is to break the wire down into a number of straight line segments in its plane. According to the principle of vector superposition and coordinate rotation, the magnetic field from the complete current wire can then be calculated by summing the contribution from each of the separate straight line pieces. The simulation results show that the proposed numerical method has high accuracy and faster computational time and is efficient and convenient to the application in projects.

Abstract: Mining association rule is an important matter in data mining, in which mining maximum frequent patterns is a key problem. Many of the previous algorithms mine maximum frequent patterns by producing candidate patterns firstly, then pruning. But the cost of producing candidate patterns is very high, especially when there exists long patterns. In this paper, the structure of a FP-tree is improved, we propose a fast algorithm based on FP-Tree for mining maximum frequent patterns, the algorithm does not produce maximum frequent candidate patterns and is more effectively than other improved algorithms. The new FP-Tree is a one-way tree and only retains pointers to point its father in each node, so at least one third of memory is saved. Experiment results show that the algorithm is efficient and saves memory space.

Abstract: Local image information is crucial for accurate segmentation of images with intensity inhomogeneity which usually occurs in medical images. However, image information in local region is not incorporated in popular region-based active contour models, such as piecewise constant models and piecewise smooth models. In this paper, a method which is able to use local information is proposed. The main point is the introduction of the local fitting information expressed by a kernel function which is crucial for segmentation. Our method is based on piecewise constant Chan-Vese model, and compare with different methods for several synthetic images and medical images.

Abstract: This paper focuses on the airborne SAR double channel data using ATI method into the Moving Target Indication process various factors affecting the discussion. Based on the actual flight appeared in the process of nonideal motion state and the engineering application of the factors such as detailed analysis, only understand the influence factors to determine the effective detection method, simplified data processing process, and improve the data processing efficiency.

Abstract: Hand tracking is a challenging research direction in computer vision field. Although gesture tracking algorithm has been widely applied to the system of human-computer interaction, it is difficult to meet the robustness and the real-time requirements because of the hand lacks sufficiently rich texture information for discrimination. In this paper, we propose a new method named Adaptive HOG-LBP features to track the palm in the unfettered color images by fusing the HOG features and LBP features. The fused features could give more texture information of the edge features and sub-structure of the palm, and are less sensitive to light variations and background clusters. Macro structural features of the palm contain the maximum amount of information that can be used for discrimination, so we can use the fusion of HOG features and LBP features to process palm detection. The fusion features are inputted into linear SVM classifier learning. Experimental results show that in our own established palm dataset, performance has been improved significantly. Particularly worthy of note is that in our own established database, most of gesture pictures are of low quality, and the palm library contains a wealth of inter-plane and out-of-plane rotated pictures.

Abstract: Prediction and optimization of quality characteristics is an important means to improve the quality of laser cutting. Kerf width (KW) and material removal rate (MRR) are selected as the quality characteristics in this paper. The fitting response surface models (RSM) of KW and MRR are considered as the optimization objective function in pulsed Nd: YAG laser cutting of alloy steel for multi-objective optimization. An improved Pareto genetic algorithm is used in the optimization, and the significant factors have been found. The predicted results are basically consistent with the experimental. Therefore, the method used in this paper can be used for optimization of KW and MRR in pulse Nd: YAG laser cutting. The study can provide theoretical basis for the prediction and optimization of quality in laser cutting.

Abstract: The traditional algorithm of mining association rules, or slowly produces association rules, or produces too many redundant rules, or it is probable to find an association rule, which posses high support and confidence, but is uninteresting, and even is false. Furthermore, a rule with negative-item cant be produced. This paper puts forward a new algorithm MVNR(Mining Valid and non-Redundant Association Rules Algorithm),which primely solves above problems by using the minimal subset of frequent itemset.