Hydro Application of Advanced Control Technology – Model Based Control

摘要

Proportional/ Integral/Derivative (PID) control is a classic control algorithm thathas been applied for decades. PID control has practical limitation, such as,requiring a process feedback variable, which make it unsuitable for advancedcontrol applications. Another control algorithm, Model Based Control, as it isbroadly referred to, doesn’t have that limitation. Model Based Control is not anew concept; various forms of this control theory have existed since perhapssince the 40's or 50's. Today many process control vendors are implementingthis control technique to increase the robustness, decrease the response time, orreduce the cost of their controls. The essence of this advanced control algorithmis an optimized model of the process behavior. The model is built based uponknowledge of the process dynamics (torque, deadtime, time constants, etc.).This model is then used to forecast, if you will, what should be happening withinthe “controlled” process or system. The model uses direct or indirect processfeedback to optimize the accuracy of the forecasts. Model Based Control allowsthe user to control a process variable that can’t be directly measured, or toremove a feedback device from a system that was traditionally required. Theability to forecast makes it possible to control a turbine’s servo motor position, viaa two stage distributing valve, without direct feedback of the intermediate(second) stage position. This intermediate (second) stage is traditionallyproportional ’zed using a LVDT (Linear Voltage Differential Transformer)feedback device on that stage. Removing this device without the model-basedcontrol scheme would result in the intermediate stage becoming an integrator. Inthe model-based control scheme the intermediate stage is proportional ‘zedusing the forecasted behavior in lieu of an actual measurement (LVDT). This