Robust decentralised control of a hydrodynamic human circulatory system simulator. (July 2015)

• Record Type: Journal Article
• Authors: Misgeld, Berno J.E.; Rüschen, Daniel; Schwandtner, Sebastian; Heinke, Stefanie; Walter, Marian; Leonhardt, Steffen
• Is Part Of: Biomedical signal processing and control. Volume 20(2015)
• Page Start: 35
• Page End: 44
• Publication Date: 2015-07
• Subjects: Mock circulatory loop -- Hardware-in-the-loop -- Ventricular assist device -- Robust control -- Multivariable control; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2015.04.004
• Languages: English
• ISSNs: 1746-8094

Abstract

Abstract : Highlights: We develop a novel decentralised feedback controller for a hydrodynamic human circulatory system simulator. We present a detailed model of the human circulatory system simulator. Nonlinear simulation and experimental results underline the performance of the proposed controller. In-vitro experiments with a blood-pump show sufficiently fast reference tracking to realise aortic and ventricle pressures. Abstract: A novel feedback controlled hydrodynamic human circulatory system simulator, well-suited for in-vitro validation of cardiac assist devices, is presented in this paper. The cardiovascular system simulator consists of high-bandwidth actuators allowing a high precision hardware-in-the-loop hydrodynamic interface in connection with physiological circulatory models calculated in real-time. The hydrodynamically coupled process dynamics consist of several actuator loops and demand a multivariable control design approach in the face of system nonlinearities and uncertainties. Based on a detailed model employing the Lagrange formalism, a robust decentralised controller is designed. Fixed structural constraints and the minimisation of the H ∞ -norm necessitate the application of nonsmooth optimisation techniques. The robust decentralised norm-optimal controller is tested in extensive in-vitro experiments and shows good performance with regard to reference tracking and system coupling. In-vitro experiments include multivariable reference step tests and frequency analysis tests of the vascular impedance transfer function.