脉宽调速系统模型辨识与不确定性

Model Inentification of PWM Systems and Uncertainty

应用系统辨识对现场直流脉宽调速系统（ＰＷＭ）进行建模后，讨论模型的验证和参数估计值的不确定性问题。提出采用参数估值的标准差，以９５％置信概率的置信区间来界定参数估值的可能变化范围，并考虑用各参数变化的随机组合来表明由于模型参数不确定性对模型输出响应的影响。合理地估计建模中的不确定性，可望对ＰＷＭ伺服系统实现智能控制提供更切合实际的知识库和推理规则。

Discusses the model validation and uncertainty of estimate parameters based on the modeling of a real D. C. pulse width modulation (PWM) velocity governing system and proposes a method for limiting the probable variety value of the system parameter with a 95 % accaracy rate using the standard deviation of the estimated parameter. Using the same model, the influences caused by parameters uncertainty in response behavior are also simulated by means of random conbination of variable parameter values one another reasonable estimation of the uncertainty of modeling presents a more rational understanding of resources and inference rules for the intelligent contol of PWM servo systems.