Repository navigation
Expand file tree
/
Copy pathabs.html
More file actions
11 lines (11 loc) · 8.45 KB
/
Copy pathabs.html
File metadata and controls
11 lines (11 loc) · 8.45 KB
1
2
3
4
5
6
7
8
9
10
11
<!DOCTYPE html>
<html lang="en"><head><meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1">
<title>ABS Simulation | Stephen Jones</title><meta name="description" content="Stephen Jones: robotics and mechatronics engineer. Georgia Tech BSME, OMSCS (Computational Perception and Robotics).">
<meta property="og:type" content="website"><meta property="og:title" content="Stephen Jones | Robotics Engineer"><meta name="theme-color" content="#1b86d1"><link rel="icon" href="assets/opt/headshot.jpeg"><link rel="stylesheet" href="style.css"></head><body data-bg="car"><main><section class="panel" data-bg="car"><h1><svg class="ico" aria-hidden="true"><use href="#i-gauge"/></svg>ABS braking simulation</h1><div class="strip"><img src="assets/gifs/redcar_braking.gif" alt="Animated red car braking" width="500" height="45"></div><p class="lead">A block-diagram model of the anti-lock braking system in a car, built in MATLAB with two teammates (Josh Chung and KC Hinz) for ME 3017 in Spring 2025. It compares a full nonlinear model against a linearized one, and tests three kinds of brake pedal input as the road turns from pavement to ice.</p><div class="chips"><span class="chip"><b>1,300 kg</b> car</span><span class="chip"><b>3</b> pedal inputs</span><span class="chip">Road to ice at <b>30 s</b></span><span class="chip">Nonlinear <b>vs.</b> linearized</span></div><p><a class="gel blue" href="assets/ABS%20Braking%20Sim/Final_Code_1.m">View the MATLAB script</a></p></section><section class="panel" data-bg="car"><h2><svg class="ico" aria-hidden="true"><use href="#i-cpu"/></svg>How the model works</h2><ol class="steps"><li><b>Driver input.</b> The brake pedal angle goes up to 15 degrees. Braking starts at 10 s and uses a ramp, a step, or a cosine wave.</li><li><b>Pedal assembly.</b> The pedal turns into brake fluid pressure. The nonlinear model uses sin θ, and the linearized model uses the small-angle form θ.</li><li><b>Tire block.</b> The friction the road can supply is F<sub>f</sub> = m g μ, where μ drops from pavement to ice halfway through.</li><li><b>ABS module.</b> It compares brake force to available friction. If the brakes are weaker than friction it adds 10% pressure, and if they are stronger it removes 10%. This is a simple bang-bang rule.</li><li><b>Brake assembly and vehicle.</b> Pressure becomes clamping force through the disc brake geometry, four wheels apply it, and deceleration is a = F<sub>b</sub> / m.</li></ol></section><section class="panel" data-bg="car"><h2><svg class="ico" aria-hidden="true"><use href="#i-waves"/></svg>Equations and parameters</h2><div class="eq">p<sub>c</sub> = K · l<sub>1</sub> / l<sub>p</sub> (K = 1×10<sup>6</sup> Pa)<br>p<sub>b</sub> = p<sub>c</sub> sin θ (nonlinear) | p<sub>b</sub> ≈ p<sub>c</sub> θ (linearized)<br>F<sub>f</sub> = m g μ(t) F<sub>b</sub> = 4 · p<sub>b</sub> · F<sub>b,c</sub><br>F<sub>b,c</sub> = [ μ<sub>pr</sub> π d<sub>s</sub><sup>2</sup> / (3 r<sub>t</sub>) ] · (r<sub>ro</sub><sup>3</sup> − r<sub>ri</sub><sup>3</sup>) / (r<sub>ro</sub><sup>2</sup> − r<sub>ri</sub><sup>2</sup>)<br>F<sub>fb</sub> = F<sub>f</sub> − F<sub>b</sub> a = F<sub>b</sub> / m</div><table><thead><tr><th>Parameter</th><th>Value</th></tr></thead><tbody><tr><th>Vehicle mass</th><td>1,300 kg</td></tr><tr><th>Pad and rotor friction</th><td>0.4</td></tr><tr><th>Master / caliper piston diameter</th><td>0.05 m / 0.10 m</td></tr><tr><th>Brake disc radii</th><td>0.115 m to 0.165 m</td></tr><tr><th>Tire radius</th><td>0.65 m</td></tr><tr><th>Road friction, then ice</th><td>0.8 to 0.9, then 0.2 to 0.4 after 30 s</td></tr><tr><th>Time</th><td>0 to 60 s in 0.1 s steps</td></tr></tbody></table><pre><i>% ABS module: keep brake force near the traction limit</i>
if F_fb > 0.0001
p_b = 1.10 .* p_b; <i>% brakes weaker than friction: add pressure</i>
elseif F_fb < -0.0001
p_b = 0.90 .* p_b; <i>% brakes stronger than friction: release</i>
end
F_b = 4 .* (p_b .* F_b_c); <i>% four wheels</i>
F_fb = F_f - F_b; <i>% feedback error</i></pre></section><section class="panel" data-bg="car"><h2>Inputs</h2><div class="figs wide"><figure><div class="fi"><img src="assets/ABS%20Braking%20Sim/abs-inputs.png" alt="Road friction drops to ice at 30 s, and the three brake pedal inputs." loading="lazy" decoding="async"></div><figcaption>Road friction drops to ice at 30 s, and the three brake pedal inputs.</figcaption></figure></div><p>Friction is drawn at random every time the script runs (0.8 to 0.9 on pavement, 0.2 to 0.4 on ice), so each run is a slightly different road.</p></section><section class="panel" data-bg="sky"><h2><svg class="ico" aria-hidden="true"><use href="#i-scan-search"/></svg>Results: nonlinear vs. linearized</h2><div class="figs wide"><figure><div class="fi"><img src="assets/ABS%20Braking%20Sim/abs-compare-step.png" alt="Step input: nonlinear vs. linearized" loading="lazy" decoding="async"></div><figcaption>Step input: the nonlinear and linearized models overlap almost exactly (ramp and cosine behave the same way).</figcaption></figure></div><table><thead><tr><th>Pedal input</th><th>Avg. deceleration on road (m/s²)</th><th>On ice (m/s²)</th><th>Nonlinear vs. linearized</th></tr></thead><tbody><tr><th>Ramp</th><td>7.9</td><td>2.9</td><td>about 1%</td></tr><tr><th>Step</th><td>8.3</td><td>2.9</td><td>about 1%</td></tr><tr><th>Cosine</th><td>3.4</td><td>1.4</td><td>under 0.1%</td></tr></tbody></table><ul><li>With the ABS rule, deceleration follows the road: about 8 m/s² on pavement (friction near 0.85) and about 2.9 m/s² on ice (friction near 0.3), which is what friction times gravity predicts.</li><li>The linearized model is within about 1% because sin θ is almost equal to θ at 15 degrees.</li><li>Peaks near 9.6 m/s² appear when the 10% pressure steps overshoot the traction limit.</li></ul></section><section class="panel" data-bg="sky"><h2><svg class="ico" aria-hidden="true"><use href="#i-images"/></svg>Original MATLAB figures</h2><div class="figs" data-car data-label="Original MATLAB figures"><figure><div class="fi"><img src="assets/ABS%20Braking%20Sim/fig1.png" alt="Inputs: road friction drops to ice, plus the three pedal inputs" loading="lazy" decoding="async"></div><figcaption>Inputs: road friction drops to ice, plus the three pedal inputs</figcaption></figure><figure><div class="fi"><img src="assets/ABS%20Braking%20Sim/fig2.png" alt="Nonlinear model, ramp input" loading="lazy" decoding="async"></div><figcaption>Nonlinear model, ramp input</figcaption></figure><figure><div class="fi"><img src="assets/ABS%20Braking%20Sim/fig3.png" alt="Nonlinear model, step input" loading="lazy" decoding="async"></div><figcaption>Nonlinear model, step input</figcaption></figure><figure><div class="fi"><img src="assets/ABS%20Braking%20Sim/fig4.png" alt="Nonlinear model, cosine input" loading="lazy" decoding="async"></div><figcaption>Nonlinear model, cosine input</figcaption></figure><figure><div class="fi"><img src="assets/ABS%20Braking%20Sim/fig5.png" alt="Linearized model, ramp input" loading="lazy" decoding="async"></div><figcaption>Linearized model, ramp input</figcaption></figure><figure><div class="fi"><img src="assets/ABS%20Braking%20Sim/fig6.png" alt="Linearized model, step input" loading="lazy" decoding="async"></div><figcaption>Linearized model, step input</figcaption></figure><figure><div class="fi"><img src="assets/ABS%20Braking%20Sim/fig7.png" alt="Linearized model, cosine input" loading="lazy" decoding="async"></div><figcaption>Linearized model, cosine input</figcaption></figure></div></section><section class="panel" data-bg="sky"><h2>Limits and what I would add next</h2><div class="callout warn"><div><b>Honest limits:</b> this is a force-balance model, not a full wheel-dynamics model. It does not compute wheel speed or slip ratio, the cosine case multiplies its output by a gain of 20 to scale it, and the random friction changes every run.</div></div><ul><li>Add wheel rotation and slip ratio so the ABS can act on slip directly.</li><li>Move the model to Simulink and sweep the 10% pressure step size.</li><li>Fix the random friction seed so runs can be compared exactly.</li></ul></section><section class="panel pager"><a class="gel blue" href="vision.html">Previous</a><a class="gel blue" href="index.html#projects">All projects</a><a class="gel blue" href="leadership.html">Next</a></section></main><script src="main.js"></script></body></html>