ACE Matlab Toolbox Tutorial: A Comprehensive Guide to Adaptive Optics

HIL systems require separation between hardware simulation and control software. The computation block in MATLAB/Simulink contains all hardware models (power sources, filters, loads) and executes on the HIL system's FPGA. The control block contains the user interface for parameter adjustment and displays results on the Host PC. The Host PC runs MATLAB with RT-Lab software for simulation management, while the Target CPU inside the HIL system runs on Linux. The FPGA executes hardware models at approximately 200 nanoseconds per step, enabling realistic real-time simulation. Signal flow includes voltage source converter (3-phase) with LCL filter and grid connection. Signal scaling is critical: for a 5 kW system at 380V with 10.7A peak current, signals must be scaled from 3.3V DSP output to 5V HIL input, then to appropriate ranges for the DSP's ADC (16-bit, 4095 counts) to ensure maximum resolution.

Hardware-in-the-loop (HIL) simulation integrates real-time simulation with physical hardware to verify control systems before deployment. This webinar demonstrates HIL using Simulink Real-Time and Speedgoat hardware with an oil rig draw works example where a PLC controls an induction motor and hydraulic brake. A typical HIL test bench comprises three components: the system under test (embedded hardware like PLCs), the HIL simulator hardware running Simulink Real-Time, and the test monitor. The workflow involves deploying controller models to embedded hardware, replacing them with I/O drivers for communication, and deploying the test harness. Key benefits include safe lab testing with 100% repetition accuracy, automated testing with reusable test vectors, proof of functional equivalence between simulation and deployed hardware, elimination of complex unit integration costs, ability to test dangerous scenarios without equipment damage, and easy fault injection for comprehensive system calibration.

This section covers the complete setup process for connecting MATLAB with physical hardware. MATLAB offers a 30-day free trial accessible via public email, providing access to MATLAB, Simulink, and 80+ products. Two versions exist: MATLAB Online for virtual simulation and Desktop Software for local hardware connection. The RTL SDR dongle, a software-defined radio using R860 tuner and RTL 2832 chip, requires a monopole antenna for sub-GHz frequency reception. MATLAB add-ons through the Add-ons tab provide hardware support packages. After connecting hardware via USB with PoE power, proper drivers must be installed using tools like Zadig for RTL SDR devices. This setup enables communication between MATLAB and physical devices for real-time signal processing and analysis.

Hardware-in-the-Loop (HIL) testing enables rapid prototyping of controllers and converters by deploying Simulink models onto FPGA hardware, allowing real-time testing of complex control algorithms like PID controllers without building expensive physical hardware; the setup requires compatible MATLAB and Vivado versions, proper installation of HDL Coder support packages, data type conversion for 32-bit JTAG communication, and verification that FPGA deployment produces results matching pure simulation.

Hardware-in-the-Loop (HIL) simulation involves building a real-time model using Real Time Workshop, configuring input control through constant blocks and from workspace signals, unchecking block reduction for HIL compatibility, selecting the appropriate target file for the hardware platform, and generating HIL files for hardware testing with subsequent result analysis in MATLAB.

Adaptive optics corrects atmospheric distortion by measuring wavefront aberrations in real-time and commanding deformable mirrors to compensate. The system uses wavefront sensors and controllers operating at ~600 Hz. Since natural guide stars aren't always available, laser guide stars excite sodium atoms at 90km altitude to create artificial reference points. Multi-conjugate AO uses two deformable mirrors conjugated to different altitudes (ground layer and 11km) for volumetric correction. Canada provides the multi-conjugate AO instrument with deformable mirrors having 3,100+ and 4,500+ actuators respectively. The system accepts AO-corrected light through beam transfer optics from below the primary mirror to the top end. This technology enables diffraction-limited imaging, transforming ground-based telescopes into competitive instruments for high-resolution astronomy.

Multi-Conjugate Adaptive Optics (MCAO) is an advanced technique that uses multiple deformable mirrors, multiple wavefront sensors, and multiple reference sources (both natural guide stars and laser guide stars) to provide three-dimensional correction of atmospheric turbulence across a wide field of view, enabling Extremely Large Telescopes (ELTs) with 25-40m diameters to achieve diffraction-limited imaging performance; this approach addresses the fundamental challenge that single-conjugate AO systems can only correct atmospheric turbulence effectively near the ground layer, while MCAO extends correction to higher altitude layers through tomographic reconstruction of the atmospheric phase perturbation.

A complete AO system consists of wavefront sensors, deformable mirrors, high-speed control systems, and science cameras. The Shack-Hartmann sensor uses lens arrays to sample wavefronts, creating spot patterns indicating local slopes. Deformable mirrors with piezoelectric actuators correct distortions by moving a few microns at high frequencies. Multi-object AO (MOAO) extends this to observe multiple targets simultaneously using multiple wavefront sensors and deformable mirrors. Laser tomography uses multiple guide stars to sample atmospheric turbulence at different altitudes, enabling multi-conjugate AO (MCAO) that corrects different atmospheric layers. Single-mode waveguides and polymer bundles enable efficient multi-object observations by guiding light from multiple targets to single detectors.

Multi-conjugate adaptive optics extends correction to wider fields by using multiple deformable mirrors conjugated to different atmospheric heights where distortions originate, requiring tomographic reconstruction of the 3D atmospheric structure using multiple wavefront sensors. This achieves ~60 arcsecond field correction compared to classical AO's narrower region. Modern implementations use parallel computing with multi-core CPUs rather than dedicated hardware, enabling real-time processing of thousands of cross-correlation calculations per frame. Despite complexity, complete correction remains challenging due to the 3D nature of atmospheric turbulence, with residual distortions affecting data analysis for applications requiring multiple exposures or time-series measurements.

Adaptive optics (AO) systems compensate for atmospheric distortion to achieve diffraction-limited imaging. Key components include wavefront sensors measuring instantaneous distortions, computers calculating phase errors, and deformable mirrors correcting shapes. The process involves measuring spot locations, calculating slopes between adjacent spots, and solving linear equations to reconstruct wavefronts. Atmospheric turbulence changes rapidly at millisecond timescales, requiring AO loop closure rates of 1 kHz or higher. Finding sufficiently bright natural guide stars within small angular separations from arbitrary targets is statistically unlikely. Solutions include laser guide stars tuned to sodium transitions at 90 km altitude, creating artificial stars. However, laser guide stars cannot measure overall tip-tilt motion due to atmospheric refraction. The cone effect further limits single-laser systems, requiring multi-conjugate AO with multiple lasers and wavefront sensors for large telescopes. AO performance is measured using full width at half maximum (FWHM) comparing image angular size to the diffraction limit (λ/D), and the Strehl ratio measuring peak intensity relative to theoretical maximum. Telescope sensitivity scales approximately with D^4, meaning a 2-meter telescope with AO is 16 times more sensitive than a 1-meter telescope. Extreme adaptive optics combines coronagraphs blocking starlight with powerful AO systems suppressing scattered light, enabling direct imaging of exoplanets.

Free space optics (FSO) transmits light through unguided media like air or space without fiber optics. The satcom market, a major FSO segment, grew to $23B in 2018 with 7% annual growth projected to $41B. Atmospheric turbulence degrades FSO links by distorting wavefronts, causing signal fading. Adaptive optics (AO) corrects this by using a deformable mirror controlled in closed-loop with a wavefront sensor. The system flattens distorted wavefronts, concentrating energy into smaller spots for improved communication quality. This same principle applies to both astronomical imaging and FSO applications.

Guide stars provide reference points for measuring atmospheric distortions in AO systems. Natural guide stars are bright stars near the target, while artificial guide stars are created using laser beacons that illuminate sodium atoms in the upper atmosphere, causing fluorescence. These guide stars experience the same atmospheric turbulence as the astronomical target, enabling accurate measurements. AO has expanded beyond astronomy into multiple applications: ophthalmology uses it for retinal imaging without aberration, microscopy employs it for deeper focus and aberration correction, free-space optical communications use it to correct turbulence-induced distortions in data transmission, and lasers with AO enable precise beam shaping for atomic trapping, laser cutting, quantum computing, and precision manufacturing. In astronomy, AO enables high-contrast imaging for directly detecting exoplanets and circumstellar disks, combining AO with coronagraphy and post-processing techniques to distinguish faint planets from stellar speckles.

Adaptive Optics (AO) revolutionizes imaging by correcting atmospheric distortions that blur ground-based observations. Stars twinkle due to atmospheric turbulence causing refraction and wavefront distortion. AO systems use deformable mirrors (segmented, piezoelectric, continuous, or MEMS types) to counteract these distortions in real-time. Key components include wavefront sensors (Shack-Hartmann and pyramid types) that measure aberrations and guide stars (natural or laser-created) as reference points. AO enables high contrast imaging for exoplanet detection and extends to applications in vision science, microscopy, free space optical communications, and laser technology for precision manufacturing and quantum computing.

Adaptive Optics corrects wavefront aberrations degrading biological imaging. In fluorescence microscopy, scattering, absorption, and aberrations distort images, with aberrations causing signal loss and resolution degradation with depth. In ophthalmology, the cornea and crystalline lens introduce aberrations limiting retinal imaging. Two correction approaches exist: direct wavefront sensing (Shack-Hartmann sensor + deformable mirror in closed-loop) offers high accuracy (λ/100) but requires dedicated light and complex setup; sensorless AO uses only a deformable mirror with iterative image-based optimization, simpler but slower. Closed-loop suits dynamic samples and high accuracy; sensorless suits static samples. Four phase modulation technologies serve these needs: liquid crystal spatial light modulators, deformable mirrors, transmissive adaptive lenses, and electromagnetic deformable mirrors. For life sciences, key requirements include sufficient actuators for high-order aberration correction, adequate stroke amplitude (tens of microns for ophthalmology), high linearity for rapid convergence, temporal stability over hours, compact integration, and improved price-performance.

Adaptive optics (AO) systems compensate for atmospheric turbulence that degrades astronomical images. The system uses a wavefront sensor to measure instantaneous wavefront distortions, calculates slopes at each aperture, determines phase errors, and generates signals to deform a mirror in real-time. Without AO, starlight appears as scattered blobs; with AO activated, light concentrates into a sharp Airy disk matching the telescope's diffraction limit. Atmospheric changes occur on millisecond timescales, requiring AO loops operating at kilohertz frequencies or faster for large telescopes.