Artificial intelligence-assisted printed electronics self-guided optimization method

An artificial intelligence-assisted printed electronics self-guided optimization method, comprising:

• Step 1, setting up factors that impact printing quality and experimental groups:
• determining a printer, a printing ink and a printing substrate are suitable;
• dividing six variables, the number of jetting holes of the printer, the number of times of printing, a printing speed, a temperature of the printing substrate, the distance between a nozzle and the substrate, and an inkjet intensity of the nozzle, into six groups, where each group consists of four uniformly varying parameters, totaling 24 printing parameter combinations, the six variables being factors that impact printing quality;
• when printing with any one of the parameter condition groups, setting the rest five parameter condition groups to have fixed printing parameters;
• Step 2, designing a printing pattern:
• after the printing parameters are determined, designing a printing pattern, where straight lines of the printing pattern have a line width of 10 μm and a line distance of 110 μm, curves of the printing pattern have a line width of 80 μm and a line distance of 160 μm;
• Step 3, printing sample patterns according to the 24 printing parameter combinations:
• setting printing parameters according to the 24 printing parameter combinations;
• printing out actual patterns according to the designed printing pattern, the actual patterns being sample patterns;
• Step 4, characterizing printing effects:
• characterizing straight line positions and curve positions of the sample patterns by an optical microscope, where as the number of jetting holes increases, details of the printing pattern deteriorate and lines connect with one another;
• as the distance between the nozzle and the substrate increases, a curvature of the lines increases;
• the smaller an average of line widths is, the closer the average of line widths is to a designed value and the smaller a standard deviation of the line widths is, the better the printing effect is;
• and data characterizing printing effects includes an average of sample pattern line widths and a standard deviation of sample pattern line widths for each combination of the 24 printing parameter combinations;
• Step 5, analyzing the data by machine learning:
• uploading the data characterizing printing effects to a computer;
• analyzing the data by machine learning to obtain printing parameters corresponding to a best printing effect;
• Step 6, returning the parameters to a user computer, and guiding the user in improving printing quality:
• transmitting by the computer the printing parameters obtained in step 5 back to a printer control program;
• modifying automatically by the control program printing parameters of the microelectronic printer and printing, to obtain an improved printing pattern;
• characterizing the improved printing pattern under an optical microscope, uploading line width averages and line width standard deviations to a computer, performing machine learning and improving printing effect.