Graduate School of Engineering Science, Osaka University, Toyonaka, Osaka 5608531, Japan.
J Theor Biol. 2012 Oct 7;310:55-79. doi: 10.1016/j.jtbi.2012.06.019. Epub 2012 Jun 23.
Human upright posture, as a mechanical system, is characterized by an instability of saddle type, involving both stable and unstable dynamic modes. The brain stabilizes such system by generating active joint torques, according to a time-delayed neural feedback control. What is still unsolved is a clear understanding of the control strategies and the control mechanisms that are used by the central nervous system in order to stabilize the unstable posture in a robust way while maintaining flexibility. Most studies in this direction have been limited to the single inverted pendulum model, which is useful for formalizing fundamental mechanical aspects but insufficient for addressing more general issues concerning neural control strategies. Here we consider a double inverted pendulum model in the sagittal plane with small passive viscoelasticity at the ankle and hip joints. Despite difficulties in stabilizing the double pendulum model in the presence of the large feedback delay, we show that robust and flexible stabilization of the upright posture can be established by an intermittent control mechanism that achieves the goal of stabilizing the body posture according to a "divide and conquer strategy", which switches among different controllers in different parts of the state space of the double inverted pendulum. Remarkably, it is shown that a global, robust stability is achieved even if the individual controllers are unstable and the information exploited for switching from one controller to another is severely delayed, as it happens in biological reality. Moreover, the intermittent controller can automatically resolve coordination among multiple active torques associated with the muscle synergy, leading to the emergence of distinct temporally coordinated active torque patterns, referred to as the intermittent ankle, hip, and mixed strategies during quiet standing, depending on the passive elasticity at the hip joint.
人类直立姿势作为一个机械系统，其特点是马鞍型不稳定，涉及稳定和不稳定的动态模式。大脑通过产生主动关节扭矩来稳定这种系统，根据时滞神经反馈控制。目前仍未解决的问题是，中枢神经系统用于以稳健的方式稳定不稳定姿势的控制策略和控制机制的清晰理解，同时保持灵活性。这方面的大多数研究都局限于单倒立摆模型，该模型对于形式化基本机械方面很有用，但不足以解决与神经控制策略有关的更普遍的问题。在这里，我们考虑在矢状面的双倒立摆模型，踝关节和髋关节具有小的被动粘弹性。尽管在存在大反馈延迟的情况下稳定双摆模型存在困难，但我们表明，通过间歇控制机制可以实现直立姿势的稳健和灵活稳定，该机制根据“分而治之策略”实现稳定身体姿势的目标，根据双倒立摆的状态空间的不同部分切换不同的控制器。值得注意的是，即使个别控制器不稳定且用于从一个控制器切换到另一个控制器的信息严重延迟（如在生物现实中发生的情况），也可以实现全局稳健稳定性。此外，间歇控制器可以自动解决与肌肉协同作用相关的多个主动扭矩之间的协调问题，导致在安静站立期间出现明显的时间协调的主动扭矩模式，称为间歇踝关节、髋关节和混合策略，具体取决于髋关节的被动弹性。
