Attentive ink MLP droplet detection algorithm based on the satellite droplets threshold domain

Satellite droplet are trailing droplets caused by improper voltage waveform control in the nozzles, which directly affect the quality of inkjet printing. Traditional empirical tuning methods struggle to effectively control and predict satellite droplet. In response, this paper establishes a simulation model of a ring-shaped piezoelectric ceramic device. establishes a simulation model of a ring-shaped piezoelectric ceramic nozzle through numerical analysis, collecting droplet behavior parameters for 13,650 different waveforms. It was found that the relative differential of the negative pressure peak at the nozzle is the primary cause of satellite droplet formation. Furthermore, the critical values of various behavior parameters leading to satellite droplet formation were investigated, resulting in the construction of a "Satellite droplet formation". resulting in the construction of a "Satellite droplet threshold domain." Based on this domain, an attention-based MLP algorithm, the Attentive Ink MLP Based on this domain, an attention-based MLP algorithm, the Attentive Ink MLP, was proposed to automatically predict whether a waveform will generate satellite droplet. algorithm reduces prediction time from 40 min to 5 min, with a validation accuracy of 0.988. The numerical analysis and algorithm were further validated through experiments with a droplet. The numerical analysis and algorithm were further validated through experiments with a droplet observation system.