Abstract:
This thesis focuses on the development of a mathematical model for a continuously moving weight conveyor to provide a more accurate weight measurement than current conveyors. While material is being transported by the conveyor system, the output weight measurement signals from the conveyor scales are always contaminated with noises. Kalman filter, which is best fit for stochastic estimation from noisy sensor measurements, is used to develop a mathematical model. In the model, the weighing operation is started after the object has come to be entirely on the weighing conveyor. A weight sensing device such as scale senses the weight of the object being transferred and produces a signal indicative of the sensed weight state. If a weight feeding machine is first started or a change in operating conditions is presented, the machine is set to identification mode. In the identification mode, control and noise parameters are identified to allow Kalman Filter to operate optimally. If the system is not started or re-identification is not needed, the weight feeding machine is set to operational mode. In the operational mode, the signal produced by the scale is applied to the control unit which produces a signal to control the speed of the DC motor. This signal is used to increase or decrease of the motor speed to produce desired mass flow. Consequently, the mass flow of the material is driven to the desired level in the presence of the unavoidable noises. Mathematical model is simulated using Matlab to obtain the results. It is observed that the developed model provides satisfactory results in providing accurate weight measurement.
