O'Connor Shawn M, Wong Jeremy D, Donelan J Maxwell
School of Exercise and Nutritional Sciences, San Diego State University, San Diego, California; and
Department of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, British Columbia, Canada.
J Appl Physiol (1985). 2016 Dec 1;121(6):1363-1378. doi: 10.1152/japplphysiol.00274.2016. Epub 2016 Sep 15.
While forcing of end-tidal gases by regulating inspired gas concentrations is a common technique for studying cardiorespiratory physiology, independently controlling end-tidal gases is technically challenging. Feedforward control methods are challenging because end-tidal values vary as a dynamic function of both inspired gases and other nonregulated physiological parameters. Conventional feedback control is limited by delays within the lungs and body tissues and within the end-tidal forcing system itself. Consequently, modern end-tidal forcing studies have generally restricted their analysis to simple time courses of end-tidal gases and to resting steady-state conditions. To overcome these limitations, we have designed and validated a more generalized end-tidal forcing system that removes the need for manual tuning and rule-of-thumb based control heuristics, while allowing for accurate control of gases along spontaneous and complicated time courses and under nonsteady physiological conditions. On average during resting, steady walking, and walking with time varying speed, our system achieved step changes in Pet within 3.0 ± 0.9 (mean ± SD) breaths and Pet within 4.4 ± 0.9 breaths, while also maintaining small steady-state errors of 0.1 ± 0.2 mmHg for Pet and 0.3 ± 0.8 mmHg for Pet The system also accurately tracked more complicated changes in end-tidal values through a bandwidth of 1/10 the respiratory (sampling) frequency, a practical limit of feedback control systems. The primary mechanism enabling this controller performance is a generic mathematical model of the cardiopulmonary system that captures the breath-by-breath relationship between inspired and end-tidal gas concentrations, with secondary contributions from reduced delays in controlled air delivery.
虽然通过调节吸入气体浓度来强制呼出末气体是研究心肺生理学的常用技术，但独立控制呼出末气体在技术上具有挑战性。前馈控制方法具有挑战性，因为呼出末值会随着吸入气体和其他未调节生理参数的动态函数而变化。传统的反馈控制受到肺部和身体组织以及呼出末强制系统本身延迟的限制。因此，现代呼出末强制研究通常将其分析限制在呼出末气体的简单时间过程和静息稳态条件下。为了克服这些限制，我们设计并验证了一种更通用的呼出末强制系统，该系统无需手动调整和基于经验法则的控制启发式方法，同时允许在自发和复杂的时间过程以及非稳态生理条件下精确控制气体。在静息、稳定行走和随时间变化速度行走期间，我们的系统平均在3.0±0.9（平均值±标准差）次呼吸内实现了呼气末二氧化碳分压（Pet）的阶跃变化，在4.4±0.9次呼吸内实现了呼气末氧气分压（Pet）的阶跃变化，同时还保持了Pet的小稳态误差为0.1±0.2 mmHg，Pet的小稳态误差为0.3±0.8 mmHg。该系统还通过呼吸（采样）频率的1/10带宽精确跟踪呼出末值的更复杂变化，这是反馈控制系统的实际限制。实现这种控制器性能的主要机制是心肺系统的通用数学模型，该模型捕捉了吸入和呼出末气体浓度之间的逐次呼吸关系，次要贡献来自受控空气输送中减少的延迟。
