Multi-rate simulation accuracy improvement for linear systems using hold networks
暂无分享,去创建一个
For real-time simulation of a system consisting of subsystems with widely ranging dynamics, multi-rate is a way of maintaining overall accuracy while at the same time retaining the real-time feature. This paper provides a technique, applicable to tri-rate simulations, for identifying which combinations of subsystem integration step sizes will result in an acceptable level of accuracy. This technique is then applied to studying the effects of using different types of hold networks for information transfer between simulation subsystems on simulation accuracy. Simulation results using a tri-rate sixth order linear circuit are used to illustrate the effect on accuracy using a zero, first, or second order hold. A frequency domain interpretation of the hold transfer functions provides an explanation for the improvement in accuracy when higher order holds are used.
[1] Roy E. Crosbie,et al. Stability of multi-rate simulation algorithms , 2007, SCSC.
[2] Robert M. Howe. Accuracy and stability tradeoffs in multirate simulation , 2001, SPIE Defense + Commercial Sensing.
[3] J. G. Pearce,et al. Testing of multi-rate simulations using the ESL simulation language , 2007, SCSC.