作者: Paul C. Badavas , Peter D. Hansen
DOI:
关键词:
摘要: A method and apparatus for a robust process control system which utilizes neural-network based multivariable inner-loop PD controller cascaded with decoupled outer-loop controllers integral action, the combination providing nonlinear PID feedforward controller. The employs quasi-Newton iterative feedback loop structure whereby manipulated variables are computed in an fashion as function of difference between inner setpoint predicted controlled variable advanced by optimum prediction time, order to incorporate downstream limiting effects on non-limited loops. compensate unmodeled changes, unmeasured disturbances, modeling errors adjusting target values.