PID Controller

The PID controller is the most widely implemented controller in industrial automation, combining proportional, integral, and derivative actions to achieve fast response, zero steady-state error, and good disturbance rejection simultaneously. Understanding PID design and tuning is one of the most important and frequently tested topics in control systems courses and GATE.

Core Concept Explanation

The PID controller extends the PI controller by adding derivative action. The proportional term provides immediate response proportional to current error. The integral term accumulates past error to eliminate steady-state offset. The derivative term reacts to the rate of change of error to anticipate future behavior and suppress overshoot. Each term addresses a different aspect of the dynamic response, which is why PID can achieve a quality of performance that neither P, PI, nor PD can match individually.

In the Laplace domain, the PID controller transfer function is Gc(s) = Kp + Ki/s + Kd*s = (Kd*s^2 + Kp*s + Ki) / s. This is a second-order numerator divided by a first-order denominator, giving two zeros and one pole at the origin. The two zeros provide two degrees of freedom in shaping the frequency response, allowing the designer to independently target the gain crossover frequency and phase margin with more flexibility than PI or PD alone.

The interaction among the three gains makes PID tuning the central challenge. Increasing Kp speeds response but may cause oscillation. Increasing Ki eliminates offset but reduces phase margin. Increasing Kd adds damping but amplifies noise. The tuning methods for PID are classified into model-based methods (where the plant model is known) and empirical methods (where experiments are done on the actual plant).

Mathematical Expression

The standard parallel form of PID is Gc(s) = Kp(1 + 1/(Ti*s) + Td*s) where Ti = Kp/Ki is the integral time and Td = Kd/Kp is the derivative time. An alternate form called the ideal PID expresses the same thing as Gc(s) = Kp + Ki/s + Kd*s. The closed-loop characteristic equation includes the numerator polynomial of Gc(s), effectively placing two zeros that the designer controls to shape transient behavior.

Tuning Methods

The Ziegler-Nichols step response method uses the open-loop step response of the plant to extract two parameters: the delay L (in seconds from the step to the inflection point projected to the time axis) and the time constant T (from that intersection to where the response reaches 63.2% of final value). Then Kp = 1.2*T/L, Ti = 2*L, and Td = 0.5*L are the Ziegler-Nichols settings.

The Ziegler-Nichols frequency response method (also called the ultimate gain method) increases the proportional gain only until the closed-loop system produces sustained oscillations. The gain at that point is called the ultimate gain Ku, and the oscillation period is Tu. Then Kp = 0.6*Ku, Ti = 0.5*Tu, and Td = 0.125*Tu. Both methods give starting points for tuning, not final optimized values.

Practical Understanding

PID controllers are found in virtually every industrial process: temperature control in furnaces, speed control in motors, pressure control in pipelines, and position control in servo drives. Modern PID implementations include features like derivative filtering (to limit noise), anti-windup (to prevent integrator saturation), setpoint weighting (to reduce overshoot on large setpoint changes), and bumpless transfer (to avoid sudden output jumps when switching modes).

Given:
Plant step response: L = 0.5 s, T = 2 s (S-shape curve)
Use Ziegler-Nichols step response method

Why this formula applies:
Plant has S-shaped open-loop step response indicating
first-order with delay behavior, suitable for Z-N method.

Formula:
Kp = 1.2 * T / L
Ti = 2 * L
Td = 0.5 * L

Substitution:
Kp = 1.2 * 2 / 0.5 = 4.8
Ti = 2 * 0.5 = 1.0 s
Td = 0.5 * 0.5 = 0.25 s

Calculation:
Ki = Kp / Ti = 4.8 / 1.0 = 4.8
Kd = Kp * Td = 4.8 * 0.25 = 1.2

Final Answer:
Kp = 4.8, Ki = 4.8, Kd = 1.2
Gc(s) = 4.8 + 4.8/s + 1.2*s

Exam Tip: In Ziegler-Nichols step response tuning, memorize Kp = 1.2T/L, Ti = 2L, Td = 0.5L. For ultimate gain method, Kp = 0.6Ku, Ti = 0.5Tu, Td = 0.125Tu. GATE often directly gives L, T or Ku, Tu and asks to find PID parameters.

Quick Revision

• PID transfer function: Gc(s) = Kp + Ki/s + Kd*s = (Kd*s^2 + Kp*s + Ki)/s. Two zeros, one pole at origin.
• P action: responds to current error. I action: eliminates steady-state offset. D action: improves damping.
• Z-N step response: Kp = 1.2T/L, Ti = 2L, Td = 0.5L where L = delay, T = time constant.
• Z-N ultimate gain: Kp = 0.6Ku, Ti = 0.5Tu, Td = 0.125Tu where Ku = ultimate gain, Tu = oscillation period.
• Increasing Ki alone reduces phase margin; integral windup is a practical concern in saturating actuators.
• Derivative term requires filtering in practice to avoid amplifying measurement noise.
• Exam trap: ideal PID has no finite pole other than at origin. Real PID adds a filter pole for the derivative term.