A new predictive control of water conductivity using a microcontroller applied to off-set printing

Water conductivity is a decisive factor in off-set printing nowadays, especially in daily newspapers. Keeping it stable affects printing brightness directly, what is relevant in the calculation of advertisement costs. Besides, print jobs are interrupted because the printer has to be cleaned after some time, delaying the process. Furthermore, if conductivity is not controlled, the ink is not adequately fixed to the printing surface. Within this context, this work proposes a generalized predictive control (GPC), with three steps ahead for plant controller, using the extended least-squares algorithm in the plant identification. A low cost solution with easy implementation results then, which can be adapted to small or big printing companies, minimizing delays. A variation equal to about 1.5% around the setpoint in steady-state is verified, but it is acceptable since the chemical process is very noisy. In addition to this, such error is much smaller than that obtained with manual process, which is about 20% to 30%.