xiaojinggeorgezhang / h-obca Goto Github PK

Hierarchical Optimization-Based Collision Avoidance - a path planner for autonomous parking

License: GNU General Public License v3.0

Julia 100.00%

h-obca's Introduction

H-OBCA

Hierarchical Optimization-Based Collision Avoidance - An algorithm for generating dynamically feasible parking trajectories

Paper describing the theory can be found here.

Short Description

H-OBCA is an optimization-based approach for autonomous parking. It builds on OBCA, which is a recent method for generating obstacle-free trajectories using optimal control.

H-OBCA is able to generate high-quality kino-dynamically feasible obstacle-free trajectories. These trajectories are smooth, and can be accurately tracked by simple low-level path following controllers. A Julia-based implementation is provided.

Examples

H-OBCA for Reverse Parking

H-OBCA for Parallel Parking

How to run the Parking code:

First steps

Change to the directory
Install Julia from https://julialang.org/downloads/ (code tested on version 0.5.1 and 0.6.0)
Open Julia in terminal
Install Julia package JuMP using Pkg.add("JuMP")
Install Julia package Ipopt using Pkg.add("Ipopt")
Install Julia package PyPlot using Pkg.add("PyPlot")
Install Julia package NearestNeighbors using Pkg.add("NearestNeighbors")
Install Julia package ControlSystems using Pkg.add("ControlSystems")

Running the parking example

Start Julia in terminal
Type in terminal: include("setup.jl")
Type in terminal: include("main.jl")

modifying the code

To play with start points, change x0 in main.jl and run the code by: include("main.jl")
If you change anything in one of the collision avoidance problems, you need to activate the changes by running: include("setup.jl")
If you change the size of the car in main.jl, the change also need to be made in collision_check.jl

Note

this code has been tested on Julia 0.5.1 and 0.6.0, and might not run on any other Julia versions
For best results, run code in Julia terminal

h-obca's People

Contributors

Stargazers

Watchers

Forkers

yuzhangbit atsushisakai ihorseai321 francescoborrelli ayan1991 jasonzhou404 cyy1991 labimage weiqiao julianst yebo92 leevenlin xushenlz dongchankim05 bbrito secondage hankerbit jiang13 liqiaw trisonis fibonacci43 kunxianzheng xiaoyuyang0901 freeclouds schietau fhao2018 mounirjaouhari cookiew longshao0967 zhijiangxu kaikaishen boyali hect1995 hihixuyang longjianquan bufanping huangatlas fullchen kecaiwu trcclub mmxa mc9529 rafeemusabbir xiangchen982 zhangzhecqut xingzai426 zhang-yonglin guenterq xyf27323 sligeek100 iamweiweishi wanyinhai oridong stephensu2017 tianlongtang lawrancexcl tcjcxy30 gaows123 richardtoo c66tang mengxingshifen1218 marc514 nishijie xintianmega littleggghost pangjinlongpjl dlc1994 ingtikna kobbie-hub hujiaming11 sree-pavan-coast doche wuhan5827 wangfj161 richard-guan-dev builderqzc thkhoa huangaibao wz1938 dtbinh yyyyydelphine johnviljoen

h-obca's Issues

What if there are some dynamic obstacles?

Hi, I think your method is brilliant! However, I think it will work only in a static environment. It calculates the parking trajectory after the driver finds a parking spot and stops the car nearly. So what would happen if there are some dynamic obstacles? If the planning is quickly enough, maybe we can update the map and environment with dynamic obstacles' new pose, but this planning method usually runs for several seconds.
Hope for your help! Thanks!

fail to avoid collision

Hello, Mr. Zhang,

Nice to communicate with you. I'm an software engineer, focused on APA and other autonumous driving technics. and recently, I wanna use h-obca to optimize the parking trajectory which is planned by hybrid a star algorithm.

after computing, the optimized path may have collision with obstacles. I have no idea how to avoid the collision by tunning the parameter of constrains. could you give me some advices on this question? thanks in advance.

btw, the apollo team implement the tdr-obca algorithm base on h-obca, and their source code is written in C++. it seems h-obca is one of the potential and powerful algorithm in path optimization.

have a nice day.

Julia0.6 Pkg doesn't work

Julia's 0.6 version of Pkg.add is not available, and there is no way on the Internet to change the source of julia0.6. I can try running this program with julia 1.2, but there are a lot of bugs in the code. Is there a way to replace the source of julia 0.6 Pkg.add or fix the bug with code below 1.2?
the bug under julia 1.2 is follow
"
ERROR: LoadError: MethodError: no method matching -(::Matrix{Int64}, ::Int64)
For element-wise subtraction, use broadcasting with dot syntax: array .- scalar
Closest candidates are:
-(::T, ::T) where T<:Union{Int128, Int16, Int32, Int64, Int8, UInt128, UInt16, UInt32, UInt64, UInt8} at int.jl:86
-(::Union{MathOptInterface.ScalarAffineFunction{T}, MathOptInterface.ScalarQuadraticFunction{T}}, ::T) where T at /home/autolab/.julia/packages/MathOptInterface/GzkIy/src/Utilities/functions.jl:1804
-(::VectorizationBase.VecUnroll{N, W, T, VectorizationBase.MM{W, X, T}}, ::Union{Int16, Int32, Int64, Int8, UInt16, UInt32, UInt64, UInt8}) where {N, W, T<:Union{Int16, Int32, Int64, Int8, UInt16, UInt32, UInt64, UInt8}, X} at /home/autolab/.julia/packages/VectorizationBase/e4FnQ/src/base_defs.jl:103
...
Stacktrace:
[1] top-level scope
@ ~/parking/H-OBCA/main.jl:85
[2] include(fname::String)
@ Base.MainInclude ./client.jl:444
[3] top-level scope
@ REPL[13]:1
in expression starting at /home/autolab/parking/H-OBCA/main.jl:85
"
""