Speaker
Description
Particle accelerators, such as cyclotrons, are intricate and extensive installations that demand meticulous maintenance and precise control over their operational conditions. It is inevitable that some minor malfunctions will occur. These malfunctions can lead to an increase in pressure within the vacuum vessel of the facility. Nevertheless, there are situations during the operation of particle accelerators where a gauge malfunctions while the pressure in the vacuum pipeline remains below the pre-estimated critical value. In such instances, it is not necessary to halt the operation, and the faulty gauge can be conveniently replaced during the next scheduled maintenance.
In this Oral Presentation, a novel approach that integrates Monte Carlo pressure simulation with a linear regression model is introduced. This integrated method is designed to forecast the pressure distribution along the axis of the pipeline in particle accelerators similar to cyclotrons. Through this approach, a trained AI model, which is founded on a linear regression algorithm, can accurately predict the maximum pressure within the vacuum pipeline. Moreover, it can determine the pressure value that a malfunctioning gauge would show by referring to the measurements from other gauges. The outcomes of this research offer significant insights that can serve as a valuable reference for the efficient operation of vacuum systems in particle accelerators.
