No abstract

Upgrading an analog DC drive to digital control does not always require a complete drive replacement. By retrofitting the existing power components with digital controls, all of the functionality of a modem digital DC drive can be achieved. Retrofitting can offer a sigruficant cost savings over a complete drive replacement, since the existing power components, motors, power cables and cable support system may be reused. The benefits of an upgrade to digital control include improved drive response, simplified troubleshooting, elimination of tuning parameter "drift", greater flexibility m operating with m e r e n t paper grades and orders, and an improved ability to control roll hardness. These benefits translate into reduced downtime, increased production efficiency and improved roll quality. The retrofit described in this case study has achieved the desired results in a cost-effective manner. A. Machine DowntimeThe existing analog master section was wired using pressure-crimping methods. Drive trips were being caused by insulation breakdown in this wiring, resulting in random grounding of circuits. Once a fault occurred, it was an extremely difficult ancl time-consuming task to locate the faulty wiring. The direct result was excessive machine downtime and lost production. The hardwired relay logic used for winder sequencing was another potential source of machine downtime due to troubleshooting d~fficulty.

An inverter designed for protective relay testing (PRT) purpose is modeled and analyzed in frequency domain, which is named as PRT inverter. By introducing the concepts of "biased on-time" and "time-voltage transfer coefficient", a modified model is proposed, which realizes the linear transition from duty cycle to output voltage and breaks the frequency limitation of the linearized models. Quasi-PID controller that is able to eliminate the inherent tracking error of the PI controller was assigned to control the modified model. To deal with the impacts of parameters drifting on the controller, a single neuron adaptive quasi-PID controller based on perceptron supervised Hebb learning rule (PSHLR) is put forward, which ameliorates the control performance excellently and makes the strategy fulfill the accuracy requirements of PRT, especially electrical fault waveform regeneration (EFWR).I.