Closed-loop instrumental variable identification of propofol anesthesia

被引:0
作者
van Heusden, Klaske [1 ]
Ansermino, J. Mark [2 ]
Dumont, Guy A. [1 ]
机构
[1] Univ British Columbia, Dept Elect & Comp Engn, Vancouver, BC, Canada
[2] Univ British Columbia, Dept Anesthesiol Pharmacol & Therapeut, Vancouver, BC, Canada
来源
2017 IEEE CONFERENCE ON CONTROL TECHNOLOGY AND APPLICATIONS (CCTA 2017) | 2017年
关键词
MULTIPLE-INPUT SYSTEMS; CHILDREN; MODEL;
D O I
暂无
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
One of the challenges in the development of high-performance closed-loop anesthetic drug delivery systems is the lack of accurate models. Physiological models have limited accuracy and drug effect varies largely between patients, while data-driven modeling of individual responses is challenging due to limited excitation and disturbances. This paper proposes a multi-input single-output (MISO) approach to deal with the effect of disturbances by identifying a plant model as well as a disturbance model. Furthermore, a MISO extension to closed-loop instrumental variable (IV) identification is proposed. Closed-loop IV methods are consistent without the need for identification of intermediate variables or noise-model parameters. Identification of fewer parameters is expected to be advantageous in this application where excitation is limited. The proposed approach is compared to closed-loop prediction-error methods. IV estimation achieved similar performance to a tailor-made parametrization. Bias in direct output-error (OE) estimates due to noise is limited. Closed-loop methods that require a controller description or that introduce additional computational complexity do not significantly improve model accuracy compared to direct OE estimation in this application.
引用
收藏
页码:1165 / 1170
页数:6
相关论文
共 23 条
  • [1] 'Paedfusor' pharmacokinetic data set
    Absalom, A
    Kenny, G
    [J]. BRITISH JOURNAL OF ANAESTHESIA, 2005, 95 (01) : 110 - 110
  • [2] [Anonymous], 1998, ANNU REV CONTROL, DOI DOI 10.1016/S1367-5788(98)00016-9
  • [3] Bibian S, 2006, IFAC S MOD CONTR BIO
  • [4] Brodie S. M., 2015, ANN M INT AN RES SOC
  • [5] Brogi E., 2016, ANESTHESIA ANALGESIA
  • [6] da Silva M. M., 2010, AM CONTR C
  • [7] Closed-Loop Control of Anesthesia: A Primer for Anesthesiologists
    Dumont, Guy A.
    Ansermino, J. Mark
    [J]. ANESTHESIA AND ANALGESIA, 2013, 117 (05) : 1130 - 1138
  • [8] Closed-loop identification revisited
    Forssell, U
    Ljung, L
    [J]. AUTOMATICA, 1999, 35 (07) : 1215 - 1241
  • [9] An optimal IV technique for identifying continuous-time transfer function model of multiple input systems
    Garnier, H.
    Gilson, M.
    Young, P. C.
    Huselstein, E.
    [J]. CONTROL ENGINEERING PRACTICE, 2007, 15 (04) : 471 - 486
  • [10] Gevers M., 1999, Proceedings of the 1999 American Control Conference (Cat. No. 99CH36251), P326, DOI 10.1109/ACC.1999.782794