A novel design of a multi-functional controller for an electronic belt scale

This paper is to present a novel design of a multi-functional controller for the electronic belt scale. The data acquisition unit in this controller is separated from the main control unit. As the control unit is located at a long distant from the electronic belt scale, data communication using RS-485 avoid error caused by EMI. The data acquisition unit and control unit are portable and are easy to be upgraded. Since this controller applied a new fuzzy control arithmetic, the simulation result shows that it has reduced 0.5 second of calculation time without overshooting in comparison of a composite PID control algorithm based on the BP neural network and Smith predictor. Since the proposed design employs an off-line look-up table, and only needs to calculate on-line for the fuzzy function, self-tuning, it reduces computational complexity and only has 20% computational complexity of the model based on BP neural network and Smith predictor.