Biological and industrial control systems

The proposed mathematical research program includes applications specific to biological and industrial control that both involve research into control over shared networks.The biological component is to understand how reflected waves may be relevant in control of cardiac output.This problem will be tackled using a mixture of mathematical modelling of wave reflections in pulsatilenetworks with approximate solutions developed by semi-analytic means and asymptotic expansions. Wavereflections occur when the pulsatile blood flow encounters arterial bifurcations and as such carry informationabout the fluid mechanical and structural status of the entire arterial tree back up to the aorta. In recent work bythe principal applicant and colleagues, the concept of `smart' baroreception was presented. There, thesensitivity of aortic stretch receptors to change in global arterial tree characteristics was measured. The nextproposed steps are to understand how global arterial blood flow and pressure may be inferred via wavereflections and then to build a closed-loop neural control model as a canonical demonstration of `smart'baroreception. The biomedical importance of `smart' baroreception is that arterial wall stiffening followed bysympathetic neural compensation may be a component in the onset of essential hypertension.The industrial component is for analysis of predictive control of fast processes over shared networks. Thisproblem will involve the programming and use of a network simulator, mathematical analysis and design ofrobust predictive control methods along with the simulation and testing of network protocols subject tononstationary communication delays. The aim is to provide remote control of manufacturing processes basedon a predictive control design that is robust with respect to communication delays coupled with a networkprotocol designed to allow a controller to adequately compensate for such delays.