terrestial locomotion requires a balance between (1) displacement, (2) robustness, (3) energy and...
TRANSCRIPT
Terrestial Locomotion
Requires a balance between
(1) displacement, (2) robustness,(3) energy and (4) stability.
All four are usually inopposition:
Improving onedegrades the others.
Q: Examples?
Alexander, R. M. (2002)Principles of Animal Locomotion.Princeton University Press.
Chapter 1: The Best Way to Travel 11.1. Fitness 11.2. Speed 21.3. Acceleration and Maneuverability 21.4. Endurance 41.5. Economy of Energy 71.6. Stability 81.7. Compromises 91.8. Constraints 91.9. Optimization Theory 101.10. Gaits 12Chapter 2: Muscle, the Motor 152.1. How Muscles Exert Force 152.2. Shortening and Lengthening Muscle 222.3. Power Output of Muscles 262.4. Pennation Patterns and Moment Arms 282.5. Power Consumption 312.6. Some Other Types of Muscle 34Chapter 3: Energy Requirements for Locomotion 383.1. Kinetic Energy 383.2. Gravitational Potential Energy 393.3. Elastic Strain Energy 403.4. Work That Does Not Increase the Body's Mechanical Energy 423.5. Work Requirements 463.6. Oscillatory Movements 48Chapter 4: Consequences of Size Differences 534.1. Geometric Similarity, Allometry, and the Pace of Life 534.2. Dynamic Similarity 584.3. Elastic Similarity and Stress Similarity 60Chapter 5: Methods for the Study of Locomotion 685.1. Cinematography and Video Recording 685.2. Stationary Locomotion 705.3. Measurement of Energy Consumption 735.4. Observing Flow 745.5. Forces and Pressures 765.6. Recording Muscle Action 80
Evolutionary Robotics
Locomotion
5.7. Recording Movement at a Distance 835.8. Properties of Materials 84Chapter 6: Alternative Techniques for Locomotion on Land 866.1. Two-Anchor Crawling 866.2. Crawling by Peristalsis 886.3. Serpentine Crawling 906.4. Froglike Hopping 916.5. An Inelastic Kangaroo 936.6. A Minimal Model of Walking 956.7. The Synthetic Wheel 976.8. Walkers with Heavy Legs 986.9. Spring-Mass Models of Running 996.10. Comparisons 100Chapter 7: Walking, Running, and Hopping 1037.1. Speed 1037.2. Gaits 1097.3. Forces and Energy 1147.4. Energy-Saving Springs 1227.5. Internal Kinetic Energy 1257.6. Metabolic Cost of Transport 1287.7. Prediction of Optimal Gaits 1337.8. Soft Ground, Hills, and Loads 1367.9. Stability 1397.10. Maneuverability 143Chapter 8: Climbing and Jumping 1468.1. Standing Jumps 1468.2. Leg Design and Jumping Technique 1508.3. Size and Jumping 1538.4. Jumping from Branches 1558.5. Climbing Vertical Surfaces and Walking on the Ceiling 159Chapter 9: Crawling and Burrowing 1669.1. Worms 1669.2. Insect Larvae 1709.3. Molluscs 1719.4. Reptiles 1769.5. Mammals 179
Evolutionary Robotics
Brachiation
Locomotion
Chapter 10: Gliding and Soaring 18110.1. Drag 18110.2. Lift 18310.3. Drag on Aerofoils 18710.4. Gliding Performance 19210.5. Stability 20010.6. Soaring 201Chapter 11: Hovering 20911.1. Airflow around Hovering Animals 20911.2. Lift Generation 21311.3. Power for Hovering 221Chapter 12: Powered Forward Flight 22412.1. Aerodynamics of Flapping Flight 22412.2. Power Requirements for Flight 22812.3. Optimization of Flight 236Chapter 13: Moving on the Surface of Water 24013.1. Fisher Spiders 24013.2. Basilisk Lizards 24413.3. Surface Swimmers 246Chapter 14: Swimming with Oars and Hydrofoils 24914.1. Froude Efficiency 24914.2. Drag-Powered Swimming 25014.3. Swimming Powered by Lift on Limbs or Paired Fins 25514.4. Swimming with Hydrofoil Tails 26114.5. Porpoising 264Chapter 15: Swimming by Undulation 26615.1. Undulating Fishes 26615.2. Muscle Activity in Undulating Fishes 27715.3. Fins, Tails, and Gaits 28215.4. Undulating Worms 284Chapter 16: Swimming by Jet Propulsion 28816.1. Efficiency of Jet Propulsion 28816.2. Elastic Mechanisms in Jet Propulsion 296
Evolutionary Robotics
Q: Why evolve awayFrom snake-like or reptilianlocomotion?
Locomotion
Chapter 17: Buoyancy 30117.1. Buoyancy Organs 30117.2. Swimming by Dense Animals 30317.3. Energetics of Buoyancy 30717.4. Buoyancy and Lifestyle 311Chapter 18: Aids to Human Locomotion 31618.1. Shoes 31618.2. Bicycles 31818.3. Scuba 32118.4. Boats 32218.5. Aircraft without Engines 324Chapter 19: Epilogue 32719.1. Metabolic Cost of Transport 32719.2. Speeds 32819.3. Gaits 33019.4. Elastic Mechanisms 33119.5. Priorities for Further Research 331REFERENCES 333INDEX 367
Evolutionary Robotics
Evolutionary RoboticsLegged Locomotion
Question 1:
For a given gait,at a given speed,how much energy is consumed?
Question 2:
For walking,How does stride lengthchange the metabolic cost?
Evolutionary RoboticsLegged LocomotionWalking
Trotting
Galloping
Stance phase(at least one footon the ground)Flight phase(no feet on theground)
Evolutionary Robotics
Walking
Trotting
Galloping
Static stability(if movement stops,animal will not fall over)
Examples?
Dynamic stability(an observed pattern overtime continues, despiteexternal perturbations)
Examples?
Polygon of support
L = leftR = rightF = frontB = backCOM = centerof mass
Evolutionary Robotics
Walking
Trotting
Galloping
Static stability(if movement stops,animal will not fall over)
Examples?
Dynamic stability(an observed pattern overtime continues, despiteexternal perturbations)
Examples?
Polygon of support
L = leftR = rightF = frontB = backCOM = centerof mass
Evolutionary Robotics
Dynamic stability(an observed pattern overtime continues, despiteexternal perturbations)
Examples?
BigDog, Boston Dynamics
“BigDog”, ?
Evolutionary RoboticsLegged Locomotion
*Ponies trained to exhibit a specificgait at different speeds.
Bongard, J. C. and R. Pfeifer (2002) A Method for Isolating Morphological Effects on Evolved Behaviour.
Q: Given 10 robots with different bodies but the same…
number of sensors and motors,neural network,fitness function,and optimizer,…
Which one will evolve the fastest gaits?
Bongard, J. C. and R. Pfeifer (2002) A Method for Isolating Morphological Effects on Evolved Behaviour.
Bongard, J. C. and R. Pfeifer (2002) A Method for Isolating Morphological Effects on Evolved Behaviour.
Footprint graph panel: results from one evolved controller row: reports the touch patterns for one foot black pixel: that foot touched the ground at that time white pixel: that foot was in the air at that time
Bongard, J. C. and R. Pfeifer (2002) A Method for Isolating Morphological Effects on Evolved Behaviour.
Evolutionary Robotics
Bongard, J. C. and R. Pfeifer (2002) A Method for Isolating Morphological Effects on Evolved Behaviour.
Evolutionary Robotics
Q: What is it about therobot’s morphologythat predicts theability to evolvegaits for it?
Given your experiences withBullet, what physical aspectsof the robot may make a difference?
Evolutionary Robotics
Evolutionary Robotics
Q1: Do changes toThe cognitive architecture(going from 3 to 5 hiddennodes)affect the evolutionof gaits?
Q2: Is this effect thesame across differentrobots?