Applied Mechanics and Materials Vols. 462-463

Abstract: In todays information age, how to effectively acquire knowledge is now a prerequisite for success in life. The conception and features of the knowledge service of university library are discussed firstly, and then the problems and development hindrances are also indicated on the basis of the network environment. Therefore, how to meet the requirement of the era becomes a challenge for the information service of university library. Some suggests are made to set up a scaffold in order to strengthen the information exchange and enhance the knowledge services level of university library by the application of the network communication tools.

Abstract: We propose a layered log-likelihood-ratio-based belief propagation(LLR-BP)algorithm for Low Density Parity Check (LDPC)codes. In the conventional decoding algorithm, the process of decoding would be terminated when it reaches the maximum iterative number or the near-convergence is achieved. The proposed algorithm is based on the variable node information quantification and stop updating criterion thought. By dividing the absolute value of the variable node to different layers, a part of the check nodes stop the iteration before reaching the maximum iterative number to save iterative time. From the simulation results, we know that the improved decoding algorithm successively achieves lower computation complexity than the conventional one .And the layered LLR-BP algorithm is a better scheme for LDPC codes.

Abstract: Switched Reluctance Motor has become one of hot topics of the international transmission for its simple structure,wide speed range and low cost, etc. In this thesis, based on TI's 32-bit DSP (TMS320F240) and following the principles of structure matching, high efficiency, easy to control, switched reluctance motor speed control system is designed using a minimum program of the main switching device. The parts of design include power converter, controller and software components.

Abstract: Modular reconfigurable robots (MRRS) is an approach to building robot for various complex tasks with the promise of great versatility, robustness and low cost. It can be used extensively to meet the demands of different tasks or different working environments by changing its shapes, such as from snake first to loop and next to quadruped, and so on. Therefore they can travel over or through obstacles, and go though small pipe. Even they can walk somewhat like a person on crutches. The quantity of the motion pattern will decide the adjustable ability of MRRS. So, dynamic vision system is very important for MRRS, which can improve its moving ability and intelligent degree. With the help of the dynamic vision system, the robot can accomplish different tasks in different working environments by itself, such as deciding routes and avoiding obstacles. In this paper, the dynamic vision system for MRRS will be analyzed and researched roundly, and three control approaches, such as open loop tracking, close loop servo and active vision system, are put forward.

Abstract: A novel PM arc guideway motor (PMAGM) is developed due to the drawback of the conventional motor reciprocating within certain angle. The PMAGM can directly drive load to make reciprocating motion within a limited angle with advantages of simplification, reliability, high efficiency, etc.. Aiming at the permanent magnet (PM) magnetization and solenoid coil electrifying modes, the PM deflection angle is parameterized and the influence on electromagnetic force has been analyzed between different magnetization and electrifying modes. Results show that electromagnetic force in Y direction magnetized along the external field force lines at proposal 1 is 51.23% higher than that of vertical magnetization, and the electromagnetic resultant force increases by 45.91%. As the angle of the PM deflection within 3o, the electromagnetic force in X direction points to the right, which is not good for PM slidable assembly swing. Based on comprehensive analysis of the advantages and disadvantages of several electrifying modes, the electricity combination under full swing angle is optimized.

Abstract: To integrating of combat simulation system in existence, the barrier of differences of runtime framework, connection protocols and integration framework among the simulation systems must be solved. HLA bridge is necessary when the different RTI implements, which decline the system performance. The Data Distribution Service (DDS), which is implemented with P/S mode, with the advantage of low coupled, strong scalability real-time and huge amount, seems an ideal solution to the problem. An integration framework based on DDS is approached, which inherits the real-time communication ability of DDS and the time-management and federation management ability of HLA, the prototype system proved the ability of the framework, and the semantics communication module of the system would be the next research topic.

Abstract: This paper summarizes some current typical odor source localization strategies and algorithms. And their advantages and disadvantages are pointed out. Some typical olfactory robots and achievements are listed. It is pointed out that the current questions of odor source localization robot based on bionic olfaction are how to build an accurate gas diffusion model and combining multi-information technology.

Abstract: Based on the theory of fuzzy mathematics and system reliability, one kind of methods of fuzzy evaluation of reliability analysis about robot for work of drainpipe deposit cleaning was presented in this paper. The method is to divide the robot into several subsystems, the influence factor of every subsystems reliability are the same, but each of subsystem weight subset and single factor evaluating matrix are different from each other, therefore the method considers not only different function of influence factor in evaluation model, but also different function of influence factor to different subsystems, so the method has comprehensive characteristics even more. Finally the reliability numerical analysis of the robot was given.

Abstract: In order to solve the problem of poor performance of the planar motor speed control system under PID control, the paper proposed an auto-disturbance rejection control scheme, on a sudden load or some disturbances, which solved the contradiction between "rapidity" and "overshoot" under PID control by using ADRC to estimation, compensation and control. This paper described in detail the composition of ADRC and the construction of planar motor speed controller. And on the sudden load and friction, the results of comparing the auto-disturbances rejection control and PID control show that the ADRC has better dynamic performances and robustness.

Abstract: A new control algorithm based on incorporating proportion integration differentiation (PID) neural networks into cross-coupling technology is developed for position synchronization of biaxial system. The PID neural networks controller is be used to adjust control value injecting two velocity loops of the biaxial system.The learning algorithm of PID neural networks is demonstrate in theory. It is tested by simulation that the proposed control scheme can guarantee the stability and the effectiveness of the biaxial control system.The simulation results show that synchronization error is reduced to 35um from 56um.