当前位置:网站首页>Parameter curve notes of coursera self driving car Part4 motion planning
Parameter curve notes of coursera self driving car Part4 motion planning
2022-06-22 07:09:00 【wujiangzhu_ xjtu】
As Coursera Chapter four FinalProject Theoretical preparation for , Tidy up Coursera Driverless courses 4.7.1 parametric curves The notes .
Coursera Driverless course video link
Catalog
Quintic spline curvature calculation
Polynomial helix x,y Coordinate calculation
Learning goals
- Understand the path planning problem and its constraints and boundary conditions
- Know what the parametric curve is
- Describe the advantages and disadvantages of using spline and helix in path planning

Kinematic constraints
Maximum curvature limit : Because the vehicle has a minimum turning radius , It means that the curvature of the path that the vehicle can walk cannot exceed the maximum value .
Usually for the sake of simple calculation , By limiting the curvature of several points on the curve, the whole curve is constrained .

Parametric curve
Original y=y(x) The curve of is written as x(u),y(u) In the form of ,x,y The value of is determined by the parameter u To make sure , adopt u To traverse the entire curve .
Parameters u Can be arc length , It can also be dimensionless [0,1] 0 Represents the starting point ,1 It's for the end .

Path optimization
Through the cost function f To optimize the path , Parametric curves allow optimization in the parameter space , Can simplify optimized expressions .
They are constraints c(r(u))≤α And boundary conditions r(0)=β0, r(uf)=βf

Path parameterization example
Two commonly used parametric curves are quintic polynomial and cubic helix

Quintic spline
x,y All are u Of 5 Sub polynomial , There is an analytical solution for the boundary condition (closed form solution)

Quintic spline curvature calculation
For quintic splines , The challenge is the curvature constraint , Because of its potential curvature discontinuity , The denominator may be 0 And the curvature derivative may be 0 You can't 1 Order continuity . Curvature calculation is complicated , It is difficult to calculate constraints .

Cubic helix
Helix defines curve curvature as a function of arc length , Definition makes curvature constraint checking simple , Curvature can be constrained by simply sampling some points on the curve to obtain good curvature performance , Because of the properties of polynomials .
k(s) Is the curvature equation
θ(s) It can be understood as the increased heading angle relative to the starting point delta heading.

Polynomial helix x,y Coordinate calculation
There is no exact analytical solution for the position of the helix , Only by numerical methods - Simpson's law to approximate Fresnel integral .

summary
The boundary conditions and constraints in path planning are discussed
The parameter curve is introduced
The difference between spline and helix in path planning is discussed

边栏推荐
- 6. 安装ssh连接工具(用于我们连接实验室的服务器)
- RT-Thread临界段的保护
- Cesium loading 3D tiles model
- Introduction to 51 Single Chip Microcomputer -- the use of Proteus 8 professional
- Tpflow V6.0.6 正式版发布
- Correspondence between pytorch and torchvision
- [meta learning] classic work MAML and reply (Demo understands meta learning mechanism)
- 自然语言处理理论和应用
- 从暴力递归到动态规划
- 2022年毕业生求职找工作青睐哪个行业?
猜你喜欢

5g NR PWS system

Self supervised learning for general out of distribution detection AAAI '20
![[Gan] Introduction to Gan basics and dcgan](/img/93/f0287f93283707b7082630cb89daf1.jpg)
[Gan] Introduction to Gan basics and dcgan

Anaconda introduction, installation and use nanny level tutorial

Theory and application of naturallanguageprocessing

Introduction to 51 Single Chip Microcomputer -- minimum system of single chip microcomputer

Use of sessionstorage and localstorage

Difference between grail layout and twin wing layout

Up sampling and down sampling (notes, for personal use)
![[fundamentals of machine learning 04] matrix factorization](/img/f5/373bfe68f1a3422e907056c20a0db3.jpg)
[fundamentals of machine learning 04] matrix factorization
随机推荐
Self attention (notes, for personal use)
Error when connecting MySQL with dbeaver for the first time
Iframe framework, native JS routing
微信小程序伪装页面(定时改变页面内容)
[meta learning] classic work MAML and reply (Demo understands meta learning mechanism)
Data security practice guide - data collection security management
Introduction to 51 single chip microcomputer - 8x8 dot matrix LED
Introduction to 51 single chip microcomputer - matrix key
Cesium加载3D Tiles模型
(multithreaded: producer consumer mode) blocking queue
MySQL面试真题(十八)——经营分析实战
Tikz learning notes (III) marking and intersection of graphics
【GCN-RS】UltraGCN: Ultra Simplification of Graph Convolutional Networks for Recommendation (CIKM‘21)
MySQL面试真题(十九)——抖音-选出每个月有连续登录2天的用户名单
代理模式与装饰模式到底哪家强
从暴力递归到动态规划
Leetcode--- search insertion location
Fundamentals of neural network (notes, for personal use)
June training (day 22) - orderly gathering
Sharing the strongest summer vacation plan of ape tutoring: the summer vacation plan is the same as learning and playing