pdf, 4.

48 mb
pid control : new identification and design methods

1 pid control technology

learning objectives
1.1 basic industrial control
1.1.1 process loop issues � a summary checklist
1.2 three-term control
1.2.1 parallel pid controllers
1.2.2 conversion totime constant pid forms
1.2.3 series pid controllers
1.2.4 simple pid tuning
1.3 pid controller implementation issues
1.3.1 bandwidth-limited derivative control
1.3.2 proportional kick
1.3.3 derivative kick
1.3.4 integral anti-windup circuits
1.3.5 reverse-acting controllers
1.4 industrial pid control
1.4.1 traditional industrial pid terms
1.4.2 industrial pid structures and nomenclature
1.4.3 the process controller unit
1.4.4 supervisory control and the scada pid controller
acknowledgements
references
2 some pid control fundamentals
learning objectives
2.1 process system models
2.1.1 state space models
2.1.2 convolution integral process models
2.1.3 laplacetransfer function models
2.1.4 common laplace transform process models
2.2 controller degrees of freedom structure
2.2.1 one degreeof freedom control
2.2.2 two degree of freedom control
2.2.3 three degree of freedom structures
2.3 pid control performance
2.3.1 controller performance assessment �general considerations
2.3.2 controller assessment � the effectiveness of pid control
2.3.3 classical stability robustness measures
2.3.4 parametric stability margins for simple processes
2.4 state space systems and pid control
2.4.1 linear reference error feedback control
2.4.2 two degreeof freedom feedback control system
2.4.3 state feedback with integral error feedback action
2.4.4 state space analysis for classical pi control structure
2.5 multivariable pid control systems
2.5.1 multivariable control
2.5.2 cascade control systems
acknowledgements
references
3 on-line model-free methods
learning objectives
3.1 introduction
3.1.1 a model-free control design paradigm
3.2 iterative feedback tuning
3.2.1 generating the cost function gradient
3.2.2 case study � a wastewater process example
3.2.3 some remarks on iterative feedback tuning
3.3 the controller parameter cycling tuning method
3.3.1 generating the gradient and hessian� some theory
3.3.2 issues for a controller parameter cycling algorithm
3.3.3 the controller parameter cycling algorithm
3.3.4 case study �multivariable decentralised control
3.4 summary and future directions
acknowledgements
appendix 3.a
references
4 automatic pid controller tuning � the nonparametric approach
learningobjectives
4.1 introduction
4.2 overview of nonparametric identification methods
4.2.1 transient response methods
4.2.2 relay feedback methods
4.2.3 fourier methods
4.2.4 phase-locked loop methods
4.3 frequency response identification with relay feedback
4.3.1 basic idea
4.3.2 improved estimation accuracy
4.3.3 estimationof a general point
4.3.4 estimation of multiple points
4.3.5 on-line relay tuning
4.4 sensitivity assessment using relay feedback
4.4.1 control robustness
4.4.2 maximum sensitivity
4.4.3 construction of thel f - chart
4.4.4 stability margins assessment
4.5 conversion to parametric models
4.5.1 single and multiple lag processes
4.5.2 second-order modelling
4.6 case studies
case study 4.1: improved estimation accuracy for the relay experiment
case study 4.2:estimationof a general point
case study 4.3:estimation of multiple points
case study 4.4:on-linerelaytuning
case study 4.5:sensitivity assessment
references
5 relay experiments for multivariable systems
learningobjectives
5.1 introduction
5.2 critical points of a system
5.2.1 critical points fortwo-input,two-output systems
5.2.2 critical points for mimo systems
5.3 decentralised relay experiments for multivariable systems
5.3.1 finding system gains at particular frequencies
5.3.2 decentralised relay control systems � sometheory
5.3.3 a decentralised two-input, two-output pid control system relay-based
procedure
5.4 a decentralised multi-input,multi-output pid control system relay-based
procedure
5.5 pid control design at bandwidth frequency
5.6 case studies
5.6.1 case study 1:the wood and berry process system model
5.6.2 case study 2:a three-input,three-output process system
5.7 summary
references
6 phase-locked loop methods
learningobjectives
6.1 introduction
6.1.1 the relay experiment
6.1.2 implementation issues for the relay experiment
6.1.3 summary conclusions on the relay experiment
6.2 some constructive numerical solution methods
6.2.1 bisection method
6.2.2 prediction method
6.2.3 bisection and prediction method� a comparison and assessment
6.3 phase-locked loop identifier module � basic theory
6.3.1 the digital identifier structure
6.3.2 noise management techniques
6.3.3 disturbance management techniques
6.4 summary and discussion .
references
7 phase-locked loop methods and pid control
learningobjectives
7.1 introduction� flexibility and applications
7.2 estimation of the phase margin
7.3 estimationof the parameters of asecond-order underdamped system
7.4 identification of systems in closed loop
7.4.1 identification of an unknown system in closed loop with an unknown
controller
7.4.2 identification of an unknown system in closed loop with a known controller
7.5 automated pi control design
7.5.1 identification aspects for automated pid control design
7.5.2 pi control withautomated gain and phase margin design
7.5.3 pi control with automated maximum sensitivity and phase margin design
7.6 conclusions
references
8 process reaction curve and relaymethods identification and pid tuning
learningobjectives
8.1 introduction
8.2 developing simple models from the process reaction curve
8.2.1 identification algorithm for oscillatory step responses
8.2.2 identification algorithm for non-oscillatory responses without overshoot
8.3 developing simple models from a relay feedback experiment
8.3.1 on-line identification of fopdt models
8.3.2 on-line identification of sopdt models
8.3.3 examples for theon-line relay feedback procedure
8.3.4 off-line identification
8.4 aninverse process model-based design procedure for pid control
8.4.1 inverse process model-based controller principles
8.4.2 pi/pid controller synthesis
8.4.3 autotuning of pid controllers
8.5 assessment of pi/pidcontrol performance
8.5.1 achievable minimal iae cost and risetime
8.5.2 assessmentof pi/pidcontrollers
references
9 fuzzy logic and genetic algorithm methods in pid tuning
learning objectives
9.1 introduction
9.2 fuzzy pid controller design
9.2.1 fuzzy pi controller design
9.2.2 fuzzy d controller design
9.2.3 fuzzy pid controller design
9.2.4 fuzzification
9.2.5 fuzzy control rules
9.2.6 defuzzification
9.2.7 a control example
9.3 multi-objective optimised genetic algorithm fuzzy pid control
9.3.1 genetic algorithm methods explained
9.3.2 case study a: multi-objective genetic algorithm fuzzy pid control of a
nonlinear plant
9.3.3 case studyb:control of solar plant
9.4 applications of fuzzy pid controllers to robotics
9.5 conclusions and discussion
acknowledgments
references
10 tuning pid controllers using subspace identification methods
learningobjectives
10.1 introduction
10.2 a subspace identification framework for process models
10.2.1 the subspace identification framework
10.2.2 incremental subspace representations
10.3 restricted structure single-input, single-output controllers
10.3.1 controller parameterisation
10.3.2 controller structure and computations
10.4 restricted-structure multivariable controller characterisation
10.4.1 controller parameterisation
10.4.2 multivariable controller structure
10.5 restricted-structure controller parametercomputation
10.5.1 cost index
10.5.2 formulationas aleast-squares problem
10.5.3 computing the closed-loopsystem condition
10.5.4 closed-loopstability conditions
10.5.5 the controllertuning algorithm
10.6 simulation case studies
10.6.1 activated sludge wastewater treatment plant layout
10.6.2 case study 1:single-input, single-outputcontrol structure
10.6.3 case study 2: control of two reactors with a lower triangular controller
structure
10.6.4 case study 3:control of three reactorswitha diagonalcontroller structure
10.6.5 case study 4: control of three reactors with a lower triangular controller
structure
references
11 design of multi-loop and multivariable pid controllers
learning objectives
11.1 introduction
11.1.1 multivariable systems
11.1.2 multivariable control
11.1.3 scopeof the chapter and some preliminary concepts
11.2 multi-loop pid control
11.2.1 biggest log-modulus tuning method
11.2.2 dominant pole placement tuning method
11.2.3 examples
11.3 multivariable pid control
11.3.1 decoupling control and design overview
11.3.2 determination of the objective loop performance
11.3.3 computation of pid controller
11.3.4 examples
11.4 conclusions
references
12 restricted structure optimal control
learning objectives
12.1 introduction to optimal lqg control for scalar systems
12.1.1 system description
12.1.2 cost functionand optimisation problem .
12.2 numerical algorithms for siso system restricted structure control
12.2.1 formulating a restricted structure numerical algorithm
12.2.2 iterative solution for the siso restricted structure lqg controller
12.2.3 properties of the restricted structure lqg controller
12.3 designof pid controllers using therestricted structure method
12.3.1 general principles for optimal restricted controller design
12.3.2 example of pid control design
12.4 multivariable optimal lqg control:an introduction
12.4.1 multivariable optimallqg control and cost function values
12.4.2 design procedures for an optimal lqgc ontroller
12.5 multivariable restricted structure controller procedure
12.5.1 analysis for a multivariable restricted structures algorithm
12.5.2 multivariable restricted structure algorithm and nested restricted
structure controllers
12.6 an application of multivariable restricted structure assessment � control of
the hotstrip finishing mill looper system
12.6.1 the hotstrip finishing mill looper system
12.6.2 an optimal multivariable lqg controller for the looper system
12.6.3 a controller assessment exercise for the hotstrip looper system
12.7 conclusions
acknowledgements
references
13 predictive pid control
learningobjectives
13.1 introduction
13.2 classical process control model methods
13.2.1 smith predictor principle
13.2.2 predictive pi with a simple model
13.2.3 method application and an example
13.3 simple process models and gpc-based methods
13.3.1 motivation for the process model restriction
13.3.2 analysis for a gpc pid controller
13.3.3 predictive pid control: delay-free system h=0
13.3.4 predictive pid control: systems with delay h > 0
13.3.5 predictive pid control: anillustrative example
13.4 control signal matching and gpc methods
13.4.1 design of siso predictive pid controllers
13.4.2 optimal values of predictive pid controller gains
13.4.3 design of mimo predictive pid controllers
acknowledgements
appendix 13.a
13.a.1 proof of lemma 13.1
13.a.2 proof of theorem13.1
13.a.3 proof of lemma 13.2
references
about the contributors
index

download : rapidshare.com/files/438534/pid_control.rar