Frequency Design

Hey students! 🎯 Welcome to one of the most practical and powerful topics in control engineering - frequency design! This lesson will teach you how to design controllers that work beautifully in the real world by shaping how they respond to different frequencies. You'll learn three essential techniques: loop-shaping, lead-lag compensation, and notch filters. By the end of this lesson, you'll understand how engineers design everything from airplane autopilots to smartphone camera stabilizers! ✈️📱

Understanding Frequency Design Fundamentals

Frequency design is like being a DJ for control systems! 🎧 Just as a DJ adjusts different frequency bands to create the perfect sound, control engineers adjust how their systems respond to different frequencies to achieve perfect performance.

Think about your car's suspension system. When you hit a small bump (high frequency), you want the suspension to respond quickly and dampen the vibration. But when you drive over a hill (low frequency), you want the car to follow the road smoothly without bouncing. This is exactly what frequency design does - it shapes how your control system responds to different types of disturbances and commands.

The foundation of frequency design lies in the frequency response of your system. This tells us how the system's output amplitude and phase change when we input sinusoidal signals of different frequencies. The key tool we use is the Bode plot, which shows magnitude (in decibels) and phase (in degrees) versus frequency on a logarithmic scale.

In frequency design, we focus on three critical frequency regions:

• Low frequencies: Determine steady-state accuracy and disturbance rejection
• Crossover frequency: Where the open-loop gain equals 1 (0 dB), determining bandwidth and speed of response
• High frequencies: Affect noise rejection and system robustness

The magic happens when we design controllers that shape the open-loop transfer function $L(s) = G(s)H(s)$ where $G(s)$ is your plant and $H(s)$ is your controller. Real-world systems like the Boeing 787's flight control system use these exact principles to maintain stable flight even in turbulent conditions!

Loop-Shaping: The Art of System Sculpting

Loop-shaping is like sculpting with mathematics! 🎨 You start with the raw frequency response of your plant and carefully shape it until it meets all your performance requirements. This technique is used extensively in aerospace, automotive, and robotics industries.

The process begins with analyzing your open-loop transfer function $L(jω) = G(jω)H(jω)$. The goal is to shape this function so that:

• At low frequencies, the gain is high enough for good steady-state accuracy
• At the crossover frequency, the slope is approximately -20 dB/decade for good stability margins
• At high frequencies, the gain rolls off quickly to reject noise

Let's consider a practical example: designing a controller for a robotic arm. The plant might have a transfer function like $G(s) = \frac{100}{s(s+10)}$. Without any controller, this system would be unstable and have poor performance. Through loop-shaping, we design a controller $H(s)$ that transforms the overall response.

The gain margin and phase margin are your safety nets. Gain margin tells you how much you can increase the gain before the system becomes unstable, while phase margin indicates how much additional phase lag the system can tolerate. Industry standards typically require at least 6 dB gain margin and 45° phase margin for robust performance.

Modern loop-shaping often uses sensitivity functions. The sensitivity function $S(jω) = \frac{1}{1+L(jω)}$ tells you how well your system rejects disturbances at different frequencies. The complementary sensitivity function $T(jω) = \frac{L(jω)}{1+L(jω)}$ shows how well the system follows reference commands. The fundamental constraint $S(jω) + T(jω) = 1$ means you're always making trade-offs - you can't have perfect disturbance rejection AND perfect tracking at the same frequency!

Companies like Tesla use loop-shaping in their Autopilot systems to ensure the car responds appropriately to steering commands while rejecting road disturbances and sensor noise.

Lead-Lag Compensation: The Dynamic Duo

Lead-lag compensation is like having a sports car with both turbo boost and cruise control! 🏎️ The lead part gives you quick response and good stability margins, while the lag part provides steady-state accuracy and disturbance rejection.

A lead compensator has the form:

$$H_{lead}(s) = K_c \frac{s + \frac{1}{\tau}}{s + \frac{1}{\alpha\tau}}$$

where $\alpha < 1$. This compensator adds phase lead (positive phase) around its center frequency, which improves stability margins and transient response. It's like adding anticipation to your system - the controller starts responding before the error gets too large.

The lag compensator has the form:

$$H_{lag}(s) = K_c \frac{s + \frac{1}{\tau}}{s + \frac{1}{\beta\tau}}$$

where $\beta > 1$. This adds gain at low frequencies without significantly affecting the crossover frequency, improving steady-state accuracy.

When combined, a lead-lag compensator becomes:

$$H_{lead-lag}(s) = K_c \frac{(s + \frac{1}{\tau_1})(s + \frac{1}{\tau_3})}{(s + \frac{1}{\alpha\tau_1})(s + \frac{1}{\beta\tau_3})}$$

The design process is systematic:

1. Lead design: Choose the lead frequency to add maximum phase at the desired crossover frequency
2. Lag design: Place the lag frequency about a decade below crossover to improve low-frequency gain
3. Gain adjustment: Set the overall gain to achieve the desired crossover frequency

Real-world example: The Hubble Space Telescope uses lead-lag compensation in its pointing control system. The lead compensation ensures the telescope can quickly and accurately point to new targets, while the lag compensation maintains precise pointing despite thermal disturbances and orbital mechanics effects.

NASA's Mars rovers also employ lead-lag compensation in their wheel control systems. The lead part helps the wheels respond quickly to terrain changes, while the lag part ensures steady forward motion over long distances.

Notch Filters: Precision Noise Elimination

Notch filters are the surgical tools of frequency design! 🔧 They precisely remove unwanted frequencies while leaving everything else untouched. Think of them as noise-canceling headphones for your control system.

A notch filter has the transfer function:

$$H_{notch}(s) = \frac{s^2 + 2\zeta_n\omega_n s + \omega_n^2}{s^2 + 2\zeta_d\omega_n s + \omega_n^2}$$

where $\zeta_n < \zeta_d$ and $\omega_n$ is the notch frequency. The numerator creates zeros that cancel out the unwanted frequency, while the denominator poles ensure the filter doesn't create instability.

The quality factor $Q = \frac{1}{2\zeta_n}$ determines how sharp the notch is. A high Q creates a very narrow, deep notch that precisely targets one frequency. A low Q creates a wider, shallower notch that affects a broader frequency range.

Notch filters are incredibly useful when your system has:

• Structural resonances: Like a flexible robotic arm that vibrates at specific frequencies
• Periodic disturbances: Such as 60 Hz electrical noise or engine vibrations
• Sensor noise: Removing specific frequency components that don't contain useful information

A fascinating real-world application is in hard disk drives. These drives spin at thousands of RPM, creating vibrations at specific frequencies that could cause read/write errors. Engineers use notch filters to eliminate these vibration frequencies from the head positioning control system, allowing your computer to store terabytes of data reliably.

Optical disk players (CD/DVD/Blu-ray) use multiple notch filters to eliminate resonances from the spinning disk and mechanical vibrations. This ensures your movies play smoothly without skipping, even when someone walks across the room!

In aerospace applications, notch filters remove structural vibrations from aircraft control surfaces. The Boeing 777's fly-by-wire system uses notch filters to prevent structural modes from interfering with flight control, ensuring passenger safety and comfort.

Conclusion

Frequency design gives you the power to sculpt your control system's behavior with mathematical precision! You've learned how loop-shaping lets you craft the overall system response, how lead-lag compensation provides the perfect balance of speed and accuracy, and how notch filters surgically remove unwanted frequencies. These techniques work together like instruments in an orchestra, each playing their part to create beautiful, robust control systems that make modern technology possible. From the smartphone in your pocket to the International Space Station orbiting overhead, frequency design principles are working 24/7 to keep our world running smoothly! 🚀

Study Notes

• Frequency design shapes how control systems respond to different frequencies for optimal performance

• Bode plots show magnitude (dB) and phase (degrees) vs frequency, essential for frequency domain analysis

• Loop-shaping designs the open-loop transfer function $L(s) = G(s)H(s)$ to meet performance specifications

• Gain margin ≥ 6 dB and phase margin ≥ 45° are industry standards for robust stability

• Sensitivity function: $S(jω) = \frac{1}{1+L(jω)}$ measures disturbance rejection capability

• Lead compensator: $H_{lead}(s) = K_c \frac{s + \frac{1}{\tau}}{s + \frac{1}{\alpha\tau}}$ where $α < 1$, adds phase lead for better stability

• Lag compensator: $H_{lag}(s) = K_c \frac{s + \frac{1}{\tau}}{s + \frac{1}{\beta\tau}}$ where $β > 1$, improves steady-state accuracy

• Lead-lag combination provides both good transient response and steady-state accuracy

• Notch filter: $H_{notch}(s) = \frac{s^2 + 2\zeta_n\omega_n s + \omega_n^2}{s^2 + 2\zeta_d\omega_n s + \omega_n^2}$ eliminates specific unwanted frequencies

• Quality factor $Q = \frac{1}{2\zeta_n}$ determines notch sharpness - higher Q means narrower, deeper notch

• Crossover frequency is where open-loop gain = 0 dB, determines system bandwidth and response speed