The document discusses PID controllers as a common feedback control system. It describes the basic structure of a PID controller using proportional, integral and derivative terms to minimize error based on the difference between a reference input and system output. Tuning the gains for each term is important for system performance and stability. While PID controllers provide good, low-cost control, limitations exist which have led to practical implementations including additional operating strategies and gain scheduling over different conditions.
The document discusses PID controllers as a common feedback control system. It describes the basic structure of a PID controller using proportional, integral and derivative terms to minimize error based on the difference between a reference input and system output. Tuning the gains for each term is important for system performance and stability. While PID controllers provide good, low-cost control, limitations exist which have led to practical implementations including additional operating strategies and gain scheduling over different conditions.
The document discusses PID controllers as a common feedback control system. It describes the basic structure of a PID controller using proportional, integral and derivative terms to minimize error based on the difference between a reference input and system output. Tuning the gains for each term is important for system performance and stability. While PID controllers provide good, low-cost control, limitations exist which have led to practical implementations including additional operating strategies and gain scheduling over different conditions.
The document discusses PID controllers as a common feedback control system. It describes the basic structure of a PID controller using proportional, integral and derivative terms to minimize error based on the difference between a reference input and system output. Tuning the gains for each term is important for system performance and stability. While PID controllers provide good, low-cost control, limitations exist which have led to practical implementations including additional operating strategies and gain scheduling over different conditions.
Slide 2 of 13 SEDEMAC Intro: Feedback Control Systems - I ● PID Controller is a Feedback Control System ● System – Relationship between inputs and outputs – Inputs/Outputs: Physical quantities (speed, voltage, air-to-fuel ratio, fuel, position…)
Inputs Outputs System
● Control system: a system that “controls” another system
● Feedback control system: uses output of controlled system Slide 3 of 13 SEDEMAC Intro: Feedback Control System - II ● Structure of a feedback control system:
Control Reference “r” Input “u” Output “y” Controller Plant
● Identify Controller, Plant, u and y for:
– Electronic governor – Digital AVR – FI ECU: Closed loop fuel control – EGR ECU
Slide 4 of 13 SEDEMAC Feedback Control Problem Formulation ● Structure of a feedback control system:
Control Reference “r” Input “u” Output “y” Controller Plant
● Feedback control problem
– Choose u(t) = f(r(t), y(t)) such that y(t) = r(t) – Time varying control input: real-time control ● Several possible solutions – Studied formally under control systems theory: researched for >60 years ● Over time, A specific form of f(.) has become popular: the PID controller
Slide 5 of 13 SEDEMAC PID Controller Structure - I ● Error based controller – Generic control structure: u(t) = f(r(t), y(t)) – Define e(t) = r(t) – y(t): the “error signal” – Control input u(t) expressed as a function of e(t)
Slide 6 of 13 SEDEMAC PID Controller Structure - II ● Block Diagram PID Controller ∫ KI
+ r e u y KP Plant + + + -
d KD dt
● PID terms – P: “present”, based on instantaneous error, fast response – I: “past”, based on history, removes steady-state error – D: “future”, based on estimated future, opposes future changes Slide 7 of 13 SEDEMAC PID Controller: Virtues ● Criteria for a good control system – Good closed-loop performance – Simple to realize ● System cost should be low (hardware, software) – Simple to tune ● Tuning is critical in a closed loop control ● Closing the loop can de-stabilize an open-loop-stable system ● Number of tuning parameters, ease of tuning important ● How PID controller fairs – Good performance for most practical systems – Simple equations, implementable on low-cost MCUs/ hardware – Relatively simple to tune, plenty of experience built Slide 8 of 13 SEDEMAC PID Controller Tuning Methods - I ● Tuning approaches – Model based (automatic/semi-automatic) – Model-free
● Model based approaches
– Mathematics based method of choosing controller gains to meet certain performance criteria – Methodology ● Express Plant behavior as a mathematical model P(θ); θ : parameters ● Express controller as a function of plant model parameters: C(θ) ● Identify θ from experiments or from knowledge – Several methods used in practice ● Ziegler Nicholas ● Relay feedback Slide 9 of 13 SEDEMAC PID Controller Tuning Methods - II ● Model free approach – Methodology ● Define performance criteria – Steady-state error, response time, error during step disturbance.. ● Based on generic understanding of effect of tuning parameters on performance criteria, tune manually – Requires understanding on how tuning parameters effect performance Steady-state Transient error Transient Stability error recovery time Kp Strong decrease Strong decrease Weak decrease Weak decrease Ki No effect No effect Strong decrease Weak decrease Kd Weak decrease Strong decrease Weak increase Strong increase
– Used currently in our governing, AVR and FI ECU products
Slide 10 of 13 SEDEMAC PID Controller: Limitations/ Improvements ● “Classical PID”: A single linear equation relates u(t) to e(t) for all operating conditions ● Limitations – Plant behavior changes (over time, over operating conditions) – Sometimes closed-loop control is not desired ● Start event, stop event, limiting control input, idling – Better performance possible with ● Changes in control structure with operating conditions ● Incorporating system knowledge (calibration) ● Practical control system – OL Strategies for start/stop + PID controller for regulation + Gain scheduling/ OL boost modes
Slide 11 of 13 SEDEMAC Summary ● PID controller is a particular implementation of feedback control system
● Simple, cost-effective implementation, good performance has lead to
widespread use
● Tuning is critical in practice, model based/ model free methods used
