Sampled-data control for minimising output product variance

The paper deals with the design of a digital (sampled-data) controller to minimize output product variance. The digital controller is designed on the basis of a feedback control algorithm. Automatic and statistical process control techniques are used to develop the control algorithm. Dead-time simulation of the control algorithm gives information of when to make an adjustment in the input variable and by how much the input variable should be adjusted so that the mean of the quality variable is brought closer to the controller set point. The technique of dead-time simulation can be used for discrete (sampled-data) control of processes with time delay. The minimization of the variance of the output product quality variable is possible by computing the input feedback adjustment and by compensating exactly the forecast disturbance that inflicts a dynamic process. The discrete (sampled-data) controller built on this control algorithm has potential to reduce output product variance, which can be used to control noisy, drifting processes. It is important to minimize the variance of the output in any production process and this type of control action can play a significant role in manufacturing.