A NEURAL NETWORK BASED SYSTEM IDENTIFICATION METHOD IS PROPOSED FOR ON-LINE CONTROL OF THICK FILM PRINTING PROCESS. PASSIVE COMPONENTS SUCH AS RESISTORS AND CAPACITORS ARE PRINTED AND FIRED ON CERAMIC SUBSTRATES OF THICKNESS RANGING FROM 10 TO 20 MICRONS FOR NETWORK APPLICATIONS. PARAMETER IDENTIFICATION AND MODELING OF THICK FILM PRINTING PROCESS USING TIME SERIES DATA AND NEURAL NETWORKS ARE DISCUSSED.
