不确定大时延下遥操作对象模型的在线修正方法

基于预测模型的遥操作是消减通信大时延的有效方法,预测精度直接影响操作品质。分析了遥操作中的不确定大时延及有限带宽对模型修正的影响,针对模型参数的不确定性,提出了一种新的模型在线修正方法。该方法采用时标标识、同态模型、数据平滑等策略,有效克服了实测信息错配、反馈修正与实时预测过程异步以及变采样步长问题,实现了模型的精确在线修正,提高了预测精度和系统的鲁棒性。通过仿真实验验证了该方法的有效性。

Online Model-Modifying Method for Teleoperation under Large and Uncertain Time Delay

The teleoperation based on model is an effective method to eliminate the influence of communication time delay with uncertainty,and predictive precision is directly concerned to qualities of teleoperation. The effect on model-modifying was analyzed. A new model-modifying method was proposed to resolve model parameter mismatch. Strategies of time point identitying,homonymic model and response smoothing were sufficiently combined so that not only the effects of uncertain time delay and limited bandwidth could be weeded out,but also the plant model mismatch could be on-line modified effectively. A simulation example was computed to show the validity of the mentioned method.