![Page 1: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/1.jpg)
1
Bayes Filter – Particle Filter and Monte Carlo Localization
Introduction toMobile Robotics
Wolfram Burgard
![Page 2: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/2.jpg)
2
§ Estimating the state of a dynamical system is a fundamental problem
§ The Recursive Bayes Filter is an effective approach to estimate the belief about the state of a dynamical system§ How to represent this belief?§ How to maximize it?
§ Particle filters are a way to efficiently represent an arbitrary (non-Gaussian) distribution
§ Basic principle§ Set of state hypotheses (“particles”)§ Survival-of-the-fittest
Motivation
![Page 3: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/3.jpg)
3
=ηP(zt | xt ) P(xt | ut, xt−1)∫ Bel(xt−1) dxt−1
Bayes Filters
=η P(zt | xt,u1, z1,…,ut ) P(xt | u1, z1,…,ut )Bayes
z = observationu = actionx = state
Bel(xt ) = P(xt | u1, z1…,ut, zt )
Markov =η P(zt | xt ) P(xt | u1, z1,…,ut )
Markov =η P(zt | xt ) P(xt | ut, xt−1)∫ P(xt−1 | u1, z1,…,ut ) dxt−1
=η P(zt | xt ) P(xt | u1, z1,…,ut, xt−1)∫P(xt−1 | u1, z1,…,ut ) dxt−1
Total prob.
Markov =ηP(zt | xt ) P(xt | ut, xt−1)∫ P(xt−1 | u1, z1,…, zt−1) dxt−1
,
![Page 4: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/4.jpg)
Probabilistic Localization
![Page 5: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/5.jpg)
5
§ Particle sets can be used to approximate functions
Function Approximation
§ The more particles fall into an interval, the higher the probability of that interval
§ How to draw samples from a function/distribution?
![Page 6: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/6.jpg)
6
§ Let us assume that f(x)< a for all x§ Sample x from a uniform distribution§ Sample c from [0,a]§ if f(x) > c keep the sample
otherwise reject the sample
Rejection Sampling
c
xf(x)
c
x’
f(x’)
OK
a
![Page 7: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/7.jpg)
7
§ We can even use a different distribution g to generate samples from f
§ Using an importance weight w, we can account for the “differences between g and f ”
§ w = f / g§ f is called target§ g is called proposal§ Pre-condition:
f(x)>0 à g(x)>0
Importance Sampling Principle
![Page 8: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/8.jpg)
8
§ Set of weighted samples
Particle Filter Representation
§ The samples represent the posterior
State hypothesis Importance weight
![Page 9: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/9.jpg)
Importance Sampling with Resampling:Landmark Detection Example
![Page 10: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/10.jpg)
Distributions
![Page 11: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/11.jpg)
11
Distributions
Wanted: samples distributed according to p(x| z1, z2, z3)
![Page 12: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/12.jpg)
This is Easy!We can draw samples from p(x|zl) by adding noise to the detection parameters.
![Page 13: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/13.jpg)
Importance Sampling
Target distribution f : p(x | z1, z2,..., zn ) =p(zk | x) p(x)
k∏p(z1, z2,..., zn )
Sampling distribution g: p(x | zl ) =p(zl | x)p(x)
p(zl )
Importance weights w: fg=p(x | z1, z2,..., zn )
p(x | zl )=p(zl ) p(zk | x)
k≠l∏
p(z1, z2,..., zn )
![Page 14: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/14.jpg)
Importance Sampling with Resampling
Weighted samples After resampling
![Page 15: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/15.jpg)
Particle Filter Localization
1. Draw from
2. Draw from
3. Importance factor for
4. Re-sample
![Page 16: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/16.jpg)
17
§ Let us assume that f(x)< a for all x§ Sample x from a uniform distribution§ Sample c from [0,a]§ if f(x) > c keep the sample
otherwise reject the sample
Rejection Sampling
c
xf(x)
c
x’
f(x’)
OK
a
2. Draw from
![Page 17: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/17.jpg)
18
§ We can even use a different distribution g to generate samples from f
§ Using an importance weight w, we can account for the “differences between g and f ”
§ w = f / g§ f is called target§ g is called proposal§ Pre-condition:
f(x)>0 à g(x)>0
Importance Sampling Principle
3. Importance factor for
![Page 18: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/18.jpg)
Particle Filters
![Page 19: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/19.jpg)
)|()(
)()|()()|()(
xzpxBel
xBelxzpw
xBelxzpxBel
aaa
=¬
¬
-
-
-
Sensor Information: Importance Sampling
![Page 20: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/20.jpg)
Bel−(x) ← p(x | u, x ') Bel(x ') d x '∫
Robot Motion
![Page 21: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/21.jpg)
Bel(x) ← α p(z | x) Bel−(x)
w ←α p(z | x) Bel−(x)
Bel−(x)= α p(z | x)
Sensor Information: Importance Sampling
![Page 22: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/22.jpg)
Robot Motion
Bel−(x) ← p(x | u, x ') Bel(x ') d x '∫
![Page 23: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/23.jpg)
24
Particle Filter Algorithm
§ Sample the next generation for particles using the proposal distribution
§ Compute the importance weights :weight = target distribution / proposal distribution
§ Resampling: “Replace unlikely samples by more likely ones”
![Page 24: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/24.jpg)
25
1. Algorithm particle_filter( St-1, ut, zt):2.
3. For Generate new samples
4. Sample index j(i) from the discrete distribution given by wt-1
5. Sample from using and
6. Compute importance weight
7. Update normalization factor
8. Add to new particle set
9. For
10. Normalize weights
Particle Filter Algorithm
0, =Æ= htSi =1,…,n
St = St∪{< xti,wt
i >}
η =η +wti
xti p(xt | xt−1,ut ) xt−1
j (i) utwti = p(zt | xt
i )
i =1,…,n
wti = wt
i /η
![Page 25: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/25.jpg)
draw xit-1 from Bel(xt-1)
draw xit from p(xt | xi
t-1,ut)
Importance factor for xit:
wti =
target distributionproposal distribution
=η p(zt | xt ) p(xt | xt−1,ut ) Bel (xt−1)
p(xt | xt−1,ut ) Bel (xt−1)∝ p(zt | xt )
Bel(xt ) = η p(zt | xt ) p(xt | xt−1,ut ) Bel(xt−1)∫ dxt−1
Particle Filter Algorithm
![Page 26: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/26.jpg)
Resampling
§ Given: Set S of weighted samples.
§ Wanted : Random sample, where the probability of drawing xi is given by wi.
§ Typically done n times with replacement to generate new sample set S’.
![Page 27: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/27.jpg)
w2
w3
w1wn
Wn-1
Resampling
w2
w3
w1wn
Wn-1
§ Roulette wheel
§ Binary search, n log n
§ Stochastic universal sampling
§ Systematic resampling
§ Linear time complexity§ Easy to implement, low variance
![Page 28: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/28.jpg)
1. Algorithm systematic_resampling(S,n):
2.3. For Generate cdf4.5. Initialize threshold
6. For Draw samples …7. While ( ) Skip until next threshold reached8.9. Insert10. Increment threshold
11. Return S’
Resampling Algorithm
11,' wcS =Æ=
ni !2=i
ii wcc += -1
1],,0]~ 11 =- inUu
nj !1=
11
-+ += nuu jj
ij cu >
{ }><È= -1,'' nxSS i1+= ii
Also called stochastic universal sampling
![Page 29: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/29.jpg)
30
Particle Filters for Mobile Robot Localization
§ Each particle is a potential pose of the robot
§ Proposal distribution is the motion model of the robot (prediction step)
§ The observation model is used to compute the importance weight (correction step)
![Page 30: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/30.jpg)
Start
Motion Model
![Page 31: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/31.jpg)
Proximity Sensor Model (Reminder)
Laser sensor Sonar sensor
![Page 32: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/32.jpg)
36
Mobile Robot Localization Using Particle Filters (1)
§ Each particle is a potential pose of the robot
§ The set of weighted particles approximates the posterior belief about the robot’s pose (target distribution)
![Page 33: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/33.jpg)
37
Mobile Robot Localization Using Particle Filters (2)
§ Particles are drawn from the motion model (proposal distribution)
§ Particles are weighted according to the observation model (sensor model)
§ Particles are resampled according to the particle weights
![Page 34: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/34.jpg)
38
Mobile Robot Localization Using Particle Filters (3)
Why is resampling needed?
§ We only have a finite number of particles
§ Without resampling: The filter is likely to loose track of the “good” hypotheses
§ Resampling ensures that particles stay in the meaningful area of the state space
![Page 35: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/35.jpg)
39
![Page 36: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/36.jpg)
40
![Page 37: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/37.jpg)
41
![Page 38: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/38.jpg)
42
![Page 39: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/39.jpg)
43
![Page 40: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/40.jpg)
44
![Page 41: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/41.jpg)
45
![Page 42: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/42.jpg)
46
![Page 43: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/43.jpg)
47
![Page 44: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/44.jpg)
48
![Page 45: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/45.jpg)
49
![Page 46: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/46.jpg)
50
![Page 47: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/47.jpg)
51
![Page 48: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/48.jpg)
52
![Page 49: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/49.jpg)
53
![Page 50: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/50.jpg)
54
![Page 51: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/51.jpg)
55
![Page 52: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/52.jpg)
56
Sample-based Localization (Sonar)
![Page 53: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/53.jpg)
Using Ceiling Maps for Localization
[Dellaert et al. 99]
![Page 54: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/54.jpg)
Vision-based Localization
P(z|x)
h(x)z
![Page 55: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/55.jpg)
Under a Light
Measurement z: P(z|x):
![Page 56: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/56.jpg)
Next to a Light
Measurement z: P(z|x):
![Page 57: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/57.jpg)
Elsewhere
Measurement z: P(z|x):
![Page 58: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/58.jpg)
Global Localization Using Vision
![Page 59: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/59.jpg)
69
Limitations
§ The approach described so far is able § to track the pose of a mobile robot and § to globally localize the robot
§ How can we deal with localization errors (i.e., the kidnapped robot problem)?
![Page 60: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/60.jpg)
70
Approaches
§ Randomly insert a fixed number of samples with randomly chosen poses
§ This corresponds to the assumption that the robot can be teleported at any point in time to an arbitrary location
§ Alternatively, insert such samples inversely proportional to the average likelihood of the observations (the lower this likelihood the higher the probability that the current estimate is wrong).
![Page 61: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/61.jpg)
71
Summary – Particle Filters
§ Particle filters are an implementation of recursive Bayesian filtering
§ They represent the posterior by a set of weighted samples
§ They can model arbitrary and thus also non-Gaussian distributions
§ Proposal to draw new samples§ Weights are computed to account for the
difference between the proposal and the target
§ Monte Carlo filter, Survival of the fittest, Condensation, Bootstrap filter
![Page 62: Introduction to Mobile Robotics Bayes Filter –Particle Filter and Monte …pabbeel/cs287-fa19/... · 2019. 10. 31. · Start. Motion Model. Proximity Sensor Model (Reminder) Laser](https://reader035.vdocuments.site/reader035/viewer/2022062609/6104671f436ae27d3e66df80/html5/thumbnails/62.jpg)
72
Summary – PF Localization
§ In the context of localization, the particles are propagated according to the motion model.
§ They are then weighted according to the likelihood model (likelihood of the observations).
§ In a re-sampling step, new particles are drawn with a probability proportional to the likelihood of the observation.
§ This leads to one of the most popular approaches to mobile robot localization