Real-time data-driven predictive feedback control of process-state dynamics for pulsed laser deposition

Pulsed laser deposition (PLD) is widely used for fabricating kilometer-length coated conductors of superconducting YBa₂Cu₃O₇. Stable deposition and consequently stable superconducting properties are required even under unexpected drifts in process conditions. We developed a process stabilization system consisting of real-time ablation plume monitoring and adjustment of input parameters. Proportional–integral–derivative (PID) control and data-driven predictive feedback control (DPFC) were investigated for suppressing plume height drift caused by lens position and oxygen gas flow variations. DPFC enabled flexible control of plume height under both drift conditions, although PID can exhibit better control performance depending on the drift source and controller tuning. The combination of real-time monitoring, data-driven modeling, and control based on future state predictions enhances deposition stability in PLD. This framework can be extended to a broad class of nonequilibrium manufacturing processes and could contribute to the development of autonomous manufacturing systems capable of self-stabilization under time-varying process conditions.