In this paper, an alternative approach to achieve speed, thrust, and flux regulation is presented for the drive system of permanent magnet synchronous linear motors (PMSLMs). The conventional proportional-integral (PI) speed regulator based direct thrust control (DTC) of PMSLM has problems of degrading the system performances, such as high thrust, flux, and current ripple, high-frequency noise caused by high thrust ripples, and the difficulty to accurately adapting to varying parameters and external loads. We propose an adaptive back-stepping control based DTC for PMSLM servo system. The PI speed regulator and two hysteresis regulators in the classical DTC system are substituted by one adaptive back-stepping controller. The uncertainties in the system are estimated online using adaptive techniques. Moreover, a space vector modulation (SVM) scheme is employed to ensure that the inverter switching frequency remains fixed. The stability of the proposed method is verified by the Lyapunov stable theory, which is difficult to achieve in the classical DTC method. Simulations are conducted to verify the effectiveness of the proposed method.

No abstract

Pulse shaping techniques are discussed in this paper for the practicability of a developed high-g shock tester. The tester is based on collision principle where there is a one-level velocity amplifier. A theoretical and experimental study of pulse shaping techniques is presented. A model was built and theoretical formulae were deduced for the shock peak acceleration and its duration. Then theoretical analysis and some experiments were conducted. The test results verify the validity of theoretical model and show that the shock tester can generate the expected high-g shock pulses by integrated usage of different impact velocities and pulse shapers made from different materials. This is important in practical applications where the items under test can be shown to excite specific resonances at predetermined acceleration levels using the shock tester.

No abstract

Considering the parameters time variation, high nonlinear, disturbance and mechanical resonance of direct-drive turing table, a double neuro-sliding mode position control based on genetic algorithms is proposed. The turning table control system consists of sliding mode control and sliding swith control, the two control is used by two neural networks respectively, and use elastic BP algorithms and genetic algorithms(GA) to online optimize the weigh of the neural network. Simulation results demonstrate that the scheme can effectively achieve tracking performance and has strong robustness. Keywords-Direct-drive turningtable; neural network; sliding mode control; genetic algorithm 4912 978-1-4244-9439-2/11/$26.00 ©2011 IEEE