Noise tuning in loss-in-weight feeding machines

The accurate control of a class of systems called the loss-in-weight feeding machine is considered. Identification of the characteristics of the stochastic noise processes within the loss-in-weight feeding machine is discussed. The noise tuning process is concerned with the adjustment of the filtering gain to minimize the system error variances. The indirect noise tuning method estimates the system noise variances, which are then used to adjust the filtering gains by the Kalman process. The direct method of noise tuning involves adjusting the filtering gains directly to achieve an optimal filter performance, and the system noise variances are estimated as a by-product. Identification of the control compensation matrix in the stochastic environment is described. Also addressed is the use of motor position sensor data and fusing the motor rate estimate with the weight rate estimate, which results in an improved loss-in-weight feeding machine performance