Major improvements in version v1.4:

• Better visualization of GNSS outages.
• Better matching between equations in code and formulas in books of reference.
• Several minor errors were fixed.

Major improvements in version v1.3:

• Gravity is eliminated from triaxial accelerometers with a plus sign in function vel_update(), since accelerometers detect gravity with negative sign in the NED reference frame.

• Compass magnetometer function is added.

• INS/GNSS/MAG integration is supported.

Major improvements in version v1.2:

• The KF state vector is reduced from 21 states to 15 states. States for estimating static biases are dropped, since states for dynamic biases can predict both static and dynamic biases. It has been tested that the performance of a INS/GNSS system is the same using 15 states instead of 21.

• Static biases imu.gb_fix and imu.ab_fix are renamed as imu.gb_sta and imu.ab_sta, respectively.

• Dynamic biases imu.gb_drift and imu.ab_drift are renamed as imu.gb_dyn and imu.ab_dyn, respectively.

In this release several improvements were addressed.

• The most important change in this release is that the master clock of the INS/GNSS integration is the IMU time vector now. Thus, no previous synchronization is needed between IMU and GNSS time vectors. The INS system will run freely until a new GNSS observation is available.

• A new example of how to use NaveGo to post-process real data from a vehicle was added to the examples folder.

• Lever arm corrections for velocity are supported now.

This is the first release of a stable version of NaveGo.

Simulation of inertial sensors errors has been improved, i.e., random walk, rate random walk, and dynamic bias. New functions has been created and checked applying the Allan variance method.

NaveGo is mostly working properly.

Allan variance procedure to characterize inertial sensors have been added.

Simulation of inertial sensors has to be revised. In particular the simulation of the bias instability.

NaveGo is mostly working properly.

Previous release has an incorrect reference trajectory (ref.mat) in the provided example, navego_example_of_use.m. This problem has been solved.

One drawbacks are pending to be resolve:

Simulation of dynamic bias has to be revised. The idea is to simulate an inertial sensor with certain levels of correlation time and dynamic bias variance, and verify this values by using the Allan variance method.

NaveGo Release v0.7-alpha

NaveGo is mostly working properly.

Two drawbacks are pending to be resolve:

1. Simulation of dynamic bias has to be revised. The idea is to simulate an inertial sensor with certain levels of correlation time and dynamic bias variance, and verify this values by using the Allan variance method.
2. The acceleration in the Z axis is considered negative downwards. This situation does not correspond with NED coordinates, but ensures a correct velocity output.