Applied Mechanics and Materials Vols. 347-350

Abstract: Attribute reduction is a key problem for rough set theory. While computing reduction according to the definitions is a typical NP problem. In this paper, basic concept of rough set theory is presented, one heuristic algorithm for attribution reduction based on conditional entropy is proposed. The actual application shows that the method is feasible and effective

Abstract: The Internet of Things' is far more complex than the Internet, aiming at the heterogeneous and real-time characteristics of its massive data, combined with the advantage of distributed cloud platform. In this paper,the data reduction tool reusable on an extended cloud platform is designed which solved the problem of heterogeneous data access and new type data; the characteristics of resources and information for the platform sharing requirements, take three levels of abstract data modeling of the data resource, the model can comprehensively, systematically describes a IOT data resource, and has high efficiency in storage and query terms. Finally, a new data-mining system built on top of cloud computing platform is implemented to verify the validity of data reduction module and finally through analysis of experimental data to verify the validity of the target.

Abstract: By comparing several typical multi-objective decision-making methods, an ideal objective method is proposed, which combines Similarity to Ideal Solution Method and Analytic Hierarchy Process (AHP). After validation, this ideal objective method is both scientific and accurate in the process of target key decision.

Abstract: Detect the intersect lines of multiple planes is one way to recognize and reconstruct the 3D space. First, analyzed the depth data grabbed by SR3000, sampled the depth data into gray bitmap and detect the depth gradient edge. By using the improved Hough Transformation method, get the parameters of spatial intersection lines.

Abstract: In daily life, the nutrients taken from the food of the human body are often insufficient or surplus, which has a negative impact on human health and waste of agricultural nutrients as well. There is some catering software that can guide people about the intake of the nutrients at home and abroad, but they can only realize the analysis function, that is, tell the user that the nutrients intake is inadequate or exceed the standard. This paper constructs an optimal allocation system of the agricultural products components of the nutrition and health care based on internet of things, allowing people to use it at home or in the hotel. Based on the users free ordering meal, analyze and optimize the users meal according to different types of users with different matching allocation model, guide the user to adjust the meal to meet their nutrient needs, finally reach the purpose of users adjustment of the dietary structure.

Abstract: Oilfield construction simulation training can provide the trainees a interactive training which is realistic, safe and standardized, and it is an important aspect to build the digital oilfield. It is still an essential hotspot to reduce development costs and improve development efficiency today. This paper summarizes the development method to reuse the oilfield simulation basic module library based on component reuse ideas, through the development of a number of oilfield construction simulation training system. Use the Layered architecture and the atomic model principle to design and realize the basic model library which is for oilfields construction simulation training system, three-dimensional figure the basis of action library, basic script library, basic plug-in library, the underlying database, through on-demand dynamic call basis library file portfolio development system, thereby enhancing the efficiency of the system development.

Abstract: It is based on DEM for battlefield environment viewport realization, the article mainly explains how the system was realized based on quadtree algorithm, and system framework was constructed under windows manipulation system and C++ setting, the result indicates the experiment is feasible.

Abstract: This paper presented a novel approach to solving the problem of robot avoidance collision planning. A Learning Classifier System is a accuracy-based machine learning system using gradient descent that combines covering operator and genetic algorithm. The covering operator is responsible for adjusting precision and large search space according to some reward obtained from the environment. The genetic algorithm acts as an innovation discovery component which is responsible for discovering new better path planning rules. The advantages of this approach are its accuracy-based representation, that can be easily reduce learning space,online learning ability,robustness due to the use of genetic algorithm.

Abstract: Against the defects that the favoritism method and average method in the multi-sensor image fusion are apt to impair the image contrast, an image fusion algorithm based on NSCT is proposed. Firstly, this algorithm applied NSCT to the rectified multi-sensor images from the same scene, then different fusion strategies were adopted to fuse the low-frequency and high-frequency directional sub-band coefficients respectively: regional energy adaptive weighted method was used for low-frequency sub-band coefficient; the directional sub-band coefficient adopted a regional-energy-matching program that combined weighted average method and selection method. Finally, the fusion image was obtained by NSCT inverse transformation. Experiments were conducted to IR and visible light image and multi-focus image respectively. And the fusion image was evaluated objectively. The experimental results show that the fusion image obtained through this algorithm has better subjective visual effects and objective quantitative indicators. It is also superior to the traditional fusion method.

Abstract: The computation efficiency of traditional algorithm is not high, and there is more time consuming. This paper presents an effective method for improved hausdorff distance, depth correction of logging curves is based on improved Hausdorff distance. In this method. On the basis of existing LTS hausdorff distance, the contrast curve segment is divided into neighborhood in an area, the LTS hausdorff distance is calculated by using engineering approximate, and the improving methods of search path is put forward, which ensures that the improved algorithm is better than the original algorithm has high computing efficiency and accuracy in theory.