disturbance, especially when the available computational
power is scarce.
The IDC method is a high-order control law with a
low number of free design parameters. The relative
order of the controller becomes low when the number
of process inputs is less than the number of outputs.
When SISO, MISO and SIMO systems are considered,
the method has very good vibration mitigation properties,
very high stability margins and good tolerance to
output model errors and measurement noise, with good
convergence properties. However, in MIMO systems
the convergence and tolerance to input model errors
are poor. The method is the best applicable to SISO
and SIMO systems subject to any number of disturbance
tones.
The IHC method is a very low-order controller with
a very low number of design parameters that requires
no model of the process (beside the phase and gain
information at a single frequency). The method provides
excellent performance in all fields in processes
subject to a single disturbance tone. However, in the
case of multiple disturbance tones, the stability margins,
tolerance to model errors and process noise and
vibration mitigation performance are lost. This loss of
performance is a direct consequence of the fact that the
control law is not based on the process model, thereby
resulting in severe problems in the frequency ranges
where the control effort is still effective. In essence, if
there are any significant process dynamics in between
the compensated disturbance tones, the roll-off rate of
the control is inadequate to prevent the control effort
being impacted by these dynamics. This is a fundamental
problem for any control law, which is not based on a
model. For such control approaches, the closed-loop
stability cannot be inherently guaranteed; the most trivial
example of such a control law is the celebrated proportional–
integral–derivative (PID) controller, whose
stability has to be evaluated separately in any process
it is implemented in. Fortunately, the problems considered
herein represent the worst case scenarios and,
in the most of the applications, the excluded dynamics
do not pose any particular problems. The method is the
most applicable to very complex processes subject to a
single disturbance tone, especially when the available
computational power is very low.
The DOFC method has the best overall performance
in all scenarios with good stability margins and good
tolerance to the model errors and measurement noise.
The major drawback of the controller is its high order
and the requirement for a process model. The relative
order of the controller is low when the number of process
outputs is less than the number of inputs.
The number of free design parameters is low and,
after some initial parameters are set, the properties
can be adjusted by three parameters. The same
disturbance, especially when the available computationalpower is scarce.The IDC method is a high-order control law with alow number of free design parameters. The relativeorder of the controller becomes low when the numberof process inputs is less than the number of outputs.When SISO, MISO and SIMO systems are considered,the method has very good vibration mitigation properties,very high stability margins and good tolerance tooutput model errors and measurement noise, with goodconvergence properties. However, in MIMO systemsthe convergence and tolerance to input model errorsare poor. The method is the best applicable to SISOand SIMO systems subject to any number of disturbancetones.The IHC method is a very low-order controller witha very low number of design parameters that requiresno model of the process (beside the phase and gaininformation at a single frequency). The method providesexcellent performance in all fields in processessubject to a single disturbance tone. However, in thecase of multiple disturbance tones, the stability margins,tolerance to model errors and process noise andvibration mitigation performance are lost. This loss ofperformance is a direct consequence of the fact that thecontrol law is not based on the process model, therebyresulting in severe problems in the frequency rangeswhere the control effort is still effective. In essence, ifthere are any significant process dynamics in betweenthe compensated disturbance tones, the roll-off rate ofthe control is inadequate to prevent the control effortbeing impacted by these dynamics. This is a fundamentalproblem for any control law, which is not based on amodel. For such control approaches, the closed-loopstability cannot be inherently guaranteed; the most trivialexample of such a control law is the celebrated proportional–integral–derivative (PID) controller, whosestability has to be evaluated separately in any processit is implemented in. Fortunately, the problems consideredherein represent the worst case scenarios and,in the most of the applications, the excluded dynamicsdo not pose any particular problems. The method is themost applicable to very complex processes subject to asingle disturbance tone, especially when the availablecomputational power is very low.The DOFC method has the best overall performancein all scenarios with good stability margins and goodtolerance to the model errors and measurement noise.The major drawback of the controller is its high orderand the requirement for a process model. The relativeorder of the controller is low when the number of processoutputs is less than the number of inputs.The number of free design parameters is low and,after some initial parameters are set, the propertiescan be adjusted by three parameters. The same
การแปล กรุณารอสักครู่..
