Estimating an optimal setpoint to lessen errors in filling weighing system based on Kalman filtering

A weighing system in which a sensor is not mounted to a discharger especially in vertical filling gives rise to an excess of weight added to the given target of weight. In addition, the excess is not constant on account of some factors, such as vibration of the machine, flow of the substance, and cycle time of the system. These factors cause the surplus to oscillate. To overcome this problem, Kalman filtering is performed to predict the optimal setpoint to meet the defined target. To illustrate the performance of the proposed technique, the resulting outcome is compared with that of using the conventional statistical method. The results have shown that the proposed approach has significantly increased the speed and lowered the error. It is pointed out that the proposed algorithm may be preferable to the traditional statistical technique due to its effectiveness and its simple implementation.