Skip to content
/ SLAM Public

Robotics Simultaneous Localization and Mapping

Notifications You must be signed in to change notification settings

malichao/SLAM

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

SLAM

Robotics Simultaneous Localization and Mapping

Current Progress:

=====
###QSLAM - Qt SLAM GUI Program
This is a GUI program for visualizing the SLAM algorithms,here are the current progress:
-Read bitmap directly as map,can easily create test map
-Use mouse to set start and target position arbitarily on the map
-Visualize the search trial route and shortest route after searching
-Driving animation

Here are some test results:
alt tag

=====

###XV11_LIDAR_Visualizer A small program to debug the XV-11 LIDAR module.Built with Qt and it's cross platform on Windows/Mac/Linux. Here's the main features:
-Visualizing the dot data with selectable range from 1-10 meters
-Options to filter out the bad point data
-Showing how many point data are valid on status bar
-Showing all the text data with format 0 - 360 : speed /rpm | point1 point2 point3 point4

alt tag

alt tag

===== ###Kalman Filter
Implemented a highly portable Kalman filter module in C.Here are some experiments of fusing gyroscope and accelerometer data.As we can see,Kalman filter does a good job fusing the sensors data and getting rid of gyro drifting.But it can not filter out the accelerometer spike noise.
Notice that at the beginning of the red curve there’s a bump,which is caused by gyro drift and filtered out by Kalman filter later on.
alt text
alt text
alt text

===== ###Searching Algorithms Following searching algorithms are implemented in this projects:
-BFS -Dynamic Programing -A* ,2D(x,y) and 3D(x,y,theta)
-RRT Using Vehicle Dynamic Model,5D(x,y,theta,speed,steering speed)

Here are the test results.
###Comparison of BFS, DP, A* In real life,situation is more complicated.For example,if we are heading to a place in rush hour and we need to go through a traffic light.Then perhaps we don't want to take a left turn because it's gonna wait for 5 minutes.Instead we may want to take a detour,which is faster then taking a left turn.This problem can be solved by adding a third dimension,which is orientation factor.Consider the following example:
start(4,2)
target(0,0)
{1, 0, 1, 1, 1},
{1, 0, 1, 0, 1},
{1, 1, 1, 1, 1},
{0, 0, 1, 0, 0},
{0, 0, 1, 0, 0},

Using three dimensional(x,y,orientation) A* we can get the following path:
Search effort: 1
Shortest path:14
alt text
Where
'T' is target
'#' is obstacle
"^,>,V,<" are trajectories
'+' is an overlap of trajectories
alt text

###RRT RRT Searching with Smooth Path and Obstacle Avoidance
![alt text](Search/GraphSearch/doc/RTT%20and%20Smooth/search_test1-result-smooth-avoid obstacle.bmp)

RRT Searching Using Vehicle Dynamic Model
alt text

alt text

=====

● Smooth algorithm

After calculating the shortest path,we want to smooth the path.
alt text

Here is another example of trajectories smooth algorithm:
alt text

===== ###PID Control and Vehicle Kinematic Simulation
Here's what have been done in this project:
-Standard PID controller
-PID twiddler(coefficient self optimization)
-One dimensional Car simulation(speed,distance,acceleration,friction,wind resistance,etc)
-Two dimensional Vehicle simulation(vehicle kinematic,dynamic model,etc)

Here are some test results:
###PID Control
alt text

###PID Coefficient Self Optimization
alt text

###2D Vehicle Simulation
A vehicle steering at a fixed angle and fixed speed,whose trajectory is exactly a circle.
alt text

About

Robotics Simultaneous Localization and Mapping

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages