State Estimation with Kalman Filters in ROS 2 | Thisas Ranhiru Samaraweera | RoboticsDevDay 2024

Robotics Skill Learning Session: "How to use Kalman Filters for State Estimation in ROS 2" - by Thisas Ranhiru Samaraweera, Electrical Engineering Student.
Robotics Developers Day is an online virtual conference (*formerly ROS Developers Day) that serves as a platform to deepen your understanding of robotics and its career opportunities. Through engaging discussions and practical sessions, you will gain valuable insights into crucial aspects of the field.
Learn more about Robotics Developers Day: www.theconstru...
--
#Robotics #ROS #Robot #skilldevelopment #RoboticsDevDay

[@chakradharv5118]

frequency: 20
sensor_timeout: 0.06
two_d_mode: true
transform_time_offset: 0.0
transform_timeout: 0.0
print_diagnostics: true
publish_tf: true
map_frame: map
odom_frame: odom
base_link_frame: base_link
world_frame: map
# Odometry sensor input
odom0: /odom # incoming messages at 20hz every 0.05s
odom0_config: [true, true, false,
false, false, true,
true, false, false,
false, false, false,
false, false, false]
odom0_queue_size: 10
odom0_nodelay: false
odom0_differential: false
odom0_relative: false
# GPS sensor input
pose0: /gps_cartesian # incoming messages at 1hz
pose0_config: [true, true, false,
false, false, false,
false, false, false,
false, false, false,
false, false, false]
pose0_queue_size: 10
pose0_nodelay: false
pose0_differential: true
pose0_relative: false
i want to fuse odometery data with pose data(converted to catesian), is my odom0_config set correctly : i want to send x, y , yaw,x_dot & pose0 : i want to send x & y
but im getting weird behaviour, the filtered odometery is almost matching the odometey not effecting by the gps data at all, even after drifting away
can someone help me how i can configure the .yaml file to get things working
i want to use the odometery for prediction & gps data for correction in ekf process
(my odom data & gps data have covariances in their messgaes)