The prestigious IEEE Control Systems Magazine, an international reference in the field of Process Control Theory and Practice, publishes as leading article in the April 2018 issue, a paper entitled Optimized Adaptive Control for the MYRRHA Linear Accelerator, and grants the cover honors to ADEX Optimized Adaptive Control Technology, whose application to the proton accelerator of the MYRRHA transmutation reactor is described in said article.
The multipurpose hybrid research reactor for high-tech applications (MYRR HA) is envisaged to be the first midsize, industrial-scale demonstrator of a transmutation reactor for the incineration of nuclear waste. It will be operative in Mol, Belgium, by 2030 and rely on a proton accelerator producing a high-energy beam that initiates and maintains transmutation within the reactor.
In the operating context of MYRRHA, the essential role of the accelerator control system can hardly be overemphasized, due to the high reliability requirements. The proton beam is accelerated in a set of superconducting (SC) cavities, in which the precise control of the field amplitude and phase is of paramount importance for obtaining required stability levels in terms of beam power and position on target.
SC cavities are nonlinear and time-varying processes, characterized by their optimal operating conditions being at a certain electromagnetic resonance frequency and the necessity of avoiding exciting mechanical resonance modes. However, the resonance frequency is a very sensitive feature that depends on the shape of the SC cavity, which in turn may be modified by the operating conditions and different kinds of stochastic and unmeasured perturbations. Maintaining this resonance frequency at its optimal operating value is critical, but its control presents multiple difficulties that have challenged the research efforts of numerous scientist in the accelerator community.
This article describes the design of an optimized adaptive control system (OACS) based on the adaptive predictive expert (ADEX) control methodology and illustrates its application to the resonance frequency of a high-fidelity SC cavity simulator of the MYRRHA proton linear accelerator developed by the Centre National de la Recherche Scientifique (CNRS), France. The OACS approach enhanced the stability and control precision of the resonance frequency and proved to be reliable under unforeseen perturbations that had previously caused cavity failure under the operation of one of the most advanced control systems reported in control literature.