Developmental Research of Learning and Adaptive Control System for Artificial Hearts

本研究的目的是开发一种以大脑心血管中心为参考模型的人工心脏学习自适应控制系统，使该系统能够对结构和参数未知、时变的心血管系统进行人工循环控制。首先，提出了一种全人工心脏控制算法（Peripheral Vascular Resistance-Based Adaptive control， PVRBAC）。该算法可以使控制器通过外周血管阻力（PVR）被自主神经系统反映。这可能是因为解耦控制用于消除PVR对主动脉压的影响。这可能使心血管中心通过操纵PVR来控制心输出量。其次，在个人计算机系统中实现了基于上述算法的动物实验控制器。实验采用有效全人工心脏系统来确定PVRBAC的充分性。从这个系统可以看出，右心房压对心率的贡献与主动脉压一样好。这意味着PVRBAC是足够的，因为它使用了由右心房压和主动脉压组成的PVR。第三，PVRBAC不仅适用于脉动式人工心脏，也适用于连续流人工心脏。这表明PVRBAC是多功能的，不依赖于血泵的种类或驱动类型。

The purpose of this research was to develop a learning and adaptive control system of artificial hearts whose reference model is the cardiovascular center in the brain and to enable the system to carry out an artificial circulatory control, coping with the cardiovascular system whose structure and parameters are unknown and time varying.First, a control algorithm of total artificial heart (Peripheral Vascular Resistance-Based Adaptive Control ; PVRBAC) was developed. This algorithm could make the controller be reflected by the autonomic nervous system via the peripheral vascular resistance (PVR). This may be because the decoupling control is used to remove the effect of the PVR on aortic pressure. This may enable the cardiovascular center to control cardiac output by manipulating the PVR.Second, a controller for animal experiments based on the above algorithm was implemented in a personal computer system. An experimental system named effectively total artificial heart system was used to ascertain the adequacy of the PVRBAC.From this system, it could be shown that the contribution of right atrial pressure to heart rate was as good as that of aortic pressure. This implies that the PVRBAC is adequate because it uses the PVR consisting of both right atrial pressure and aortic pressure.Third, it was shown that the PVRBAC is feasible for continuous-flow artificial heart as well as pulsatile-flow one. This indicates that the PVRBAC is versatile and does not depend on the kind of blood pump or drive type.