Advanced Materials Research Vols. 971-973

Abstract: In this paper, a small multi-agent system (MAS) is proposed based on behavioral approach for the complex grinding processes. Causal association agents were established according to the material balance of grinding processes, and prediction agents and stability control agents were built by adding prediction and control algorithms. The simulation results prove that the system has good stability and anti-interference performance. Multi-agent behavioral control method can be considered as a potential solution to stabilize grinding processes.

Abstract: Anti-backlash gear can improve the static transmission precision of system. Besides, the dynamic characteristics of anti-backlash gear system have a significant effect on the performance of overall mechanism, and the torsion spring preload of anti-backlash gear is an important factor to affect the dynamic characteristic. In order to study dynamic characteristics of the anti-backlash gear, a rigid-flexible coupling model of single-stage anti-backlash system was established based on ADAMS / Flex, and the simulation accuracy was compared with the pure rigid model. The effect of the torsion spring preload on frequency response of the anti-backlash system was studied by virtual sweep experiments.

Abstract: A new delay compensation control algorithm is applied to the networked control system (NCS), in which the characters of random time-delay and packet loss involved. Firstly, the ideal model of networked control system is considered as a reference model, and then the Lyapunov stability theory is used to design an adaptive control algorithm which generates a bounded control input, it is guaranteed that generalized output error is closing to zero. Narandra stability adaptive controller is used in the networked control system, which has many advantages including simple algorithm, less time computing, fast response. The simulation results are shown that the networked control system is well controlled by this method.

Abstract: The inverted pendulum system is characterized as a typical nonlinear, fast multi-variable, essentially unstable system. It is difficult to control because of its instability .In order to improve balance control, the mathematical model of the single inverted pendulum is established, a LQR controller is designed which is based on improved artificial bee colony. Experiments show that the improved algorithm has better performance than standard artificial bee colony algorithm on convergence and rate balance control to meet the requirements of the single inverted pendulum.

Abstract: Neutral Beam Injection (NBI) is an effective and powerful method of plasma heating and current driving on magnetic confinement fusion device. Control system of NBI (NBICS) on Experimental Advanced Superconducting Tokamak (EAST) is a distributed control system. NBICS based on network technologies, classified according to control level, consists of remote monitoring layer, server control layer and field control layer. The 3-layer control architecture is easy for extending system functions and upgrading devices. Timing system provides reference clock in synchronization and interlock protection in the NBI system. Operators in the control room have full authorities to operate the NBI machine.

Abstract: A Planar Parallel Manipulator (PPM) with non-planar links overcomes most of the shortcomings in conventional manipulators. Due to cantilever action, the links of a PPM develop bending stress. When the links are non-planar (having inclination with the horizontal plane), this cantilever action reduces and the same manipulator can have higher payload capacity. In this project, a planar 3-RRR manipulator with non-planar links is investigated for rectifying singularity at home configuration, and a novel method of position feedback of the motor is designed, fabricated and tested with rest of the motion control components.

Abstract: A new method for distribution network planning based on group search optimizer (GSO) is proposed. In the population the individual with minimal individual fitness value is chosen as the producer, and the residuals are respectively chosen as scroungers and rangers. During the searching, the local physical search of producer is performed by quick branch-exchange method, and the scrounger are close to the producer step by step to carry out following search in the solution space. The proposed method realizes satisfied coordination of global search with local search, thus the searching efficiency is improved and a better global convergence is implemented. Results of a 10-node 10 kV system verify the effectiveness of the proposed method.

Abstract: The process of grinding chatter is divided into three states: stable grinding state, chatter gestation state and chatter state. The vibration signals of grinding process contain chatter features can correspond well to the changes of grinding process. By analyzing the natural frequency band energy ratio of grinding process, a method of grinding chatter prediction is proposed. Experimental results show that the natural frequency band energy ratio is beyond a certain threshold, chatter occurred, otherwise no chatter happened. The method of grinding prediction can provide reference for vibration monitoring in practice.

Abstract: Aim at the problem of wind turbine gearbox fault diagnosis , a method of Hilbert space feature entropy is proposed. Using the information entropy theory, the faults information obtained in the HHT spectrum quantify to the value of Hilbert space feature entropy, thus achieving quantitative analysis of the failure of the wind turbine gearbox. The method was applied to fault diagnosis of the wind turbine gearboxes, which proved that the method is efficacious.

Abstract: In order to diagnosis the fault in a a comprehensive, real-time and simple way, a multi-input predictive fault diagnosis system was promoted based on acceleration mainly include the sensor, Signal processing,display and stepper motors.Sensor was used to acquisition the inputs such as grain loss, clogging and engine vibration.Then,the inputs was processed by the fault diagnosis algorithm promoted in this thesis to obtain diagnostic results and display the rusults in button display module. When a fault occured, stepper motor would start to work controled by a control signal to minimized the failure harm.Furtherly,the effectiveness was improved by an example. The experimental results show that the prediction method can achieve the predictive fault diagnosis and effectively simplify the computational complexity with good practicality and reality.