[en] Pyrofusion neutron sources are compact, low power systems which may be capable of producing short intense pulses of neutrons from D-D fusion reactions. The neutron pulse duration has been too short for use as a continuous wave source, and too long for application in interrogation techniques. However, initial analysis of Pyrofusion dynamics has indicated the potential to manipulate the pulse characteristics through system control. This paper introduces the development of state-space multimode system models in MATLAB/Simulink that may be used to support the predictability and control of the neutron pulse. Preliminary examples of frequency-domain studies, stability analysis and time-domain simulations are reported. (authors)

[en] Pyrofusion sources are compact, low power systems which are capable of producing short intense pulses of neutrons from D-D fusion reactions. A model structure has been derived from the underlying physics principles and it is shown that the dominant non-linear dynamics pose significant challenges to the system control. This paper presents the methods under development to construct, estimate and analyse nonlinear grey-box models of Pyrofusion systems from measured data. The methods have been trialled on data sets obtained from the publications of previous Pyrofusion research groups. The results indicate the promising potential application of these methods to the system identification of profusion device characteristics to support model verification, pulse characterisation and prediction of operational capabilities. The identified models are shown to exhibit satisfactory analogous cause-and-effect behaviour. (authors)

[en] The description of human operator dynamic characteristics in mathematical terms compatible with control engineering practice is an essential prerequisite to the analytical treatment of manual reactor control systems. Safe reactor operation requires effective operator control through interaction with plant dynamics, manipulators and displays. Traditional static analysis methods consider only specific situations; they fail to adequately explain the mutual interactions between the operator and the reactor plant characteristics. In this paper we investigate the theory for describing operator-reactor characteristics based on the methods of conventional control engineering techniques. The primary purpose of the experiments reported is the validation of the quasi-linear operator model. (authors)

[en] Highlights: • A non-linear model structure is fit to manual-control of the VR-1 reactor. • The estimated model is a good fit to experimental data. • The frequency response indicates a good closed-loop control system performance. • The output non-linearity shows potential evidence of operator performance. - Abstract: The description of human operator dynamic characteristics in mathematical terms compatible with control engineering practice is an essential prerequisite to the analytical treatment of manual reactor control systems. Safe reactor operation requires effective operator control through interaction with plant dynamics, manipulators and displays. Traditional static analysis methods consider only specific situations; they fail to adequately explain the mutual interactions between the operator and the reactor plant characteristics. In this paper we investigate the cause-and-effect behaviours of three VR-1 research reactor operators during reactivity disturbance experiments, based on the methods of conventional control engineering techniques. The primary purpose of experiments carried out at the VR-1 training reactor and reported in this paper is the validation of a quasi-linear cause-and-effect operator model.