Dual_ekf and navsat_transform for outdoor GPS based navigation

edit

Hi,

I need help for implementing GPS based outdoor navigation for my project and I have spent weeks trying to figure out how I could improve the functioning. I am attaching a video here as to how the visualization on RVIZ looks like and the description of the problem is as follows

Issue:

1)The robot position drifts as I move it in its environment and the orientation of the robot on RVIZ is also not facing the correct way when compared to the actual orientation. There is a 180 degrees shift in its orientation.To help you understand the drift, on the video , the robot is actually moving in the direction indicated by the green arrow.

2) The odometry/gps output from navsat_transform doesnt match odometry/filtered and odometry/filtered_map. While both the local and global odometry (i.e odometry/filtered and odometry/filtered_map match each other). I tried adding the magnetic declination and the yaw_offset (imu/data yaw value was not zero when facing magnetic north) in the navsat_transform parameters.

3) On RVIZ, the odom frame position and orientaion on TF is not stable, it changes.

Frame convention :

Map_frame: map

Odom_frame: odom

Base_link frame: Base_footprint

I have been going through several ros answers and I am running out of ideas and time

Here are my file contents

1) Dual_ekf_navsat.yaml

    ekf_se_odom:
      frequency: 30
      sensor_timeout: 0.1
      two_d_mode: true  #false
      transform_time_offset: 0.0
      transform_timeout: 0.0
      print_diagnostics: true
      debug: false

      map_frame: map
      odom_frame: odom
      base_link_frame: base_footprint   #base_link
      world_frame: odom

      odom0: odom    #odometry/wheel
      odom0_config: [false, false, false,
                     false, false, false,
                     true,  true,  true,
                     false, false, true,
                     false, false, false]
      odom0_queue_size: 10
      odom0_nodelay: true
      odom0_differential: false
      odom0_relative: false

      imu0: imu/data
      imu0_config: [false, false, false,
                    true,  true,  false,
                    false, false, false,
                    true,  true,  true,
                    true,  true,  true]
      imu0_nodelay: false
      imu0_differential: false
      imu0_relative: false
      imu0_queue_size: 10
      imu0_remove_gravitational_acceleration: true



    use_control: false

      process_noise_covariance: [1e-3, 0,    0,    0,    0,    0,    0,     0,     0,    0,    0,    0,    0,    0,    0,
                                 0,    1e-3, 0,    0,    0,    0,    0,     0,     0,    0,    0,    0,    0,    0,    0,
                                 0,    0,    1e-3, 0,    0,    0,    0,     0,     0,    0,    0,    0,    0,    0,    0,
                                 0,    0,    0,    0.3,  0,    0,    0,     0,     0,    0,    0,    0,    0,    0,    0,
                                 0,    0,    0,    0,    0.3,  0,    0,     0,     0,    0,    0,    0,    0,    0,    0,
                                 0,    0,    0,    0,    0,    0.01, 0,     0,     0,    0,    0,    0,    0,    0,    0,
                                 0,    0,    0,    0,    0,    0,    0.5,   0,     0,    0,    0,    0,    0,    0,    0,
                                 0,    0,    0,    0,    0,    0,    0,     0.5,   0,    0,    0,    0,    0,    0,    0,
                                 0,    0,    0,    0,    0,    0,    0,     0,     0.1,  0,    0,    0,    0,    0,    0,
                                 0,    0,    0,    0,    0,    0,    0,     0,     0,    0.3,  0,    0,    0,    0,    0,
                                 0,    0,    0,    0,    0,    0,    0,     0,     0,    0,    0.3,  0,    0,    0,    0,
                                 0,    0,    0,    0,    0,    0,    0,     0,     0,    0,    0,    0.3,  0,    0,    0,
                                 0 ...