Advanced Materials Research Vol. 748

Abstract: The existing x-ray computed tomography algorithm simulation assume the complete measurements of the investigated objectives to be available, but this is not true in most applications. To overcome the problem, we creatively propose a method of image reconstruction based on fuzzy clustering algorithm under limited measurements. Different from the existing algorithms, we map all measurements into a set of vectors and cluster all vectors for the image reconstruction. The proposed algorithm aims to be easily realized, lower time complexity, and applicable in a real-time manner in case of limited measurements of the investigated objectives. Experiments demonstrate the effectiveness and efficiency of the proposed algorithm.

Abstract: In electrical resistance tomography (ERT), one of the main reasons for poor quality of reconstructed image is ill-posed property of the inverse problem. In order to improve image quality, one of the feasible methods is to incorporate prior information. Based on conventional Tikhonov regularization method to reconstruct ERT image, the regularization matrix, combined with prior information vector, was proposed. The prior information included the structure property of human thorax, conductivity distribution information of organs and other dynamic information during the respiration process. The research results indicated that the proposed method can enhance the spatial resolution of the reconstructed images, and reduced the trail traces of the investigated objectives. It is promising in detecting and monitoring regional lung ventilation change by means of the ERT system with a proper imaging method.

Abstract: This article according to the characteristics of the image enhancement and the histogram of the image properties, using genetic algorithms to improve the image gray-scale. Optimization process is to optimize the shape parameters of a normalized to unit interval doublet Gaussian function. The simulation results show that: the purpose of the image enhancement is pull the distance between object and image in gray-scale coordinates, proving the effectiveness of the methods. Lay the foundation for the binary image and target feature extraction.

Abstract: This paper Analyze the features of Tibetan speech intonation, design the state of rhythm and observation of Tibetan TTS system, based on HMM the intonation forecasting model is given, and finally present to explore algorithm of the best rhythm state.

Abstract: In this paper we compare the performance of metaheuristic methods, namely simulated annealing and Tabu Search, against simple hill climbing heuristic on a supply chain optimization problem. The benchmark problem we consider is the retailer replenishment optimization problem for a retailer selling multiple products. Computation and simulation results demonstrate that simulated annealing and Tabu search improve solution quality. However, the performance improvement is less in simulations with random noise. Lastly, simulated annealing appears to be more robust than Tabu search, and the results justify its extra implementation effort and computation time when compared against hill climbing.

Abstract: The paper discuss the application of CRI method in the fuzzy reasoning in the following three aspects: (1) the influence of different implication operators on the classification result; (2) the influence of different shape of triangles on the classification result when fuzzying the example as a triangle; (3) the influence of different compound algorithms on the classification result. The paper presents the maximum information entropy methods to calculate the classification result for a given rule and compares the CRI method with the similarity-based method in the fuzzy classification problems. The experiments show that choosing proper parameters can make the classification accuracy obtained by CRI method prior to the accuracy obtained by similarity-based method. Thereby proves the validity of CRI method in the classification problems.

Abstract: Actually as a question of pattern recognition, the research of cash recognition mainly focuses on three aspects: data acquisition, feature extraction and classifier design. In order to finish the real time recognition of Chinese cash, a fuzzy recognition intelligent system is presented in this paper. We use this arithmetic in the processing of information of cash and the recognition of the cash. Experiments has proved that this method can auto organize and auto study, and meet the need of complex system for the networks no linearity and high collateral.

Abstract: This article describes the research and development of automatic switch test and control system. It introduces the hardware structure of the test system, designs and builds the signal conditioning circuit. Applying of virtual instrument design software to achieve the purpose of control the system, and acquisition and processing of test data. And take some effective measures to curb the noise interference signals. The experimental results show that the system can fulfill accurately online monitoring for the parameters in the test process. The test results can be used as a standard to determine whether the switch qualified.

Abstract: World Model is the robots view of the world and contains information for the robot to make decisions. World Modeling is the process that robots map the sensor data into the World Model. The complexity of the sensors increased, so does the data process components. In this paper, an architecture for world modeling is proposed. This architecture enhances the extensibility of the data process components by decoupling them with a middleware, and it improves the efficiency by computing in parallel. The combination of this architecture and the PDDL will also be presented.

Abstract: This Paper will introduce methods used to resolve self-location problem of humanoid robots. There are two method that will be mentioned in this paper. A traditional method which uses marks and flags on the soccer field for robot location will be introduced at first. This paper will also discuss how to solve the singularity problem in traditional self-location method. The optimization self-location method which uses sidelines on the soccer field will be introduced at the same time. This paper will introduce how to mapping the field sidelines in vision to the field sidelines of soccer field. This method is implemented and tested in Robocup 3D Soccer Simulation League. The results show that the accuracy of self-location method will be increased by taking full use of soccer field sidelines in vision information.