GNSS/INS Integrated Navigation GUI Software

English Version | 中文版

GINS-Navi is a GNSS/INS loosely coupled integrated navigation software based on the ESKF algorithm, compatible with both Windows (GUI) and Linux (CUI) environments.

[Please read the "Configuration" section carefully before using the software.]

Contact: Hengzhen Liu, Wuhan University, [email protected]

1 Configuration

The project's configuration files are located in the ./conf directory. Detailed parameters for the algorithm can be set through these configuration files.

Section	Parameter	Value	Description
Configure	navSys	nFrame	Navigation frame, especially the INS mechanism
	usegnssvel	true	Whether use GNSS velocity update or not
	useZUPT	false	Whether use ZUPT detection and update or not
	useodonhc	true	Whether use ODO/NHC update or not (ODO not completed yet)
	usesinglenhc	true	Whether to use single NHC update or not
	usesingleodo	false	Whether use single ODO update or not (ODO not completed yet)
	onlyinsmech	false	Whether use only the INS mechanism or not
IMU Settings	imudataformat	asc	IMU data format (0: imr, 1: asc, 2: txt[7 columns])
	imurawcoordinate	RFU	Raw IMU data in RFU or FRD
	imuInitStaticTime	5.0	IMU initial alignment time [min]
	imuSamplingRate	100.0	IMU sampling rate [Hz]
	imuInitialGPSTWeek	2315	IMU initial GPS week
	useAttFromRoughAlign	false	Use initial attitude from rough alignment
GNSS Settings	useInitalPosVelFromRTK	false	Use initial position and velocity from RTK solution
State Settings	estimateImuScale	true	Estimate IMU scale (21-dimension states if true, 15 if false)
Plot Settings	plotResults	true	Use matplotlib to automatically plot navigation results
File Paths	file-roverobs-renix	F:\GINS-Navi\data\WHU20240522-XWGI7660\rover.24O	Rover observation file path
	file-baseobs-renix	F:\GINS-Navi\data\WHU20240522-XWGI7660\base.24O	Base observation file path
	file-navi-renix	F:\GINS-Navi\data\WHU20240522-XWGI7660\base.24N	Navigation file path
	gnssfilepath	F:\GINS-Navi\data\WHU20240522-XWGI7660\GNSS-RTK.txt	GNSS file path
	imrfilepath	F:\GINS-Navi\data\WHU20240522-XWGI7660\rover.imr	IMR file path
	ascfilepath	F:\GINS-Navi\data\WHU20240522-XWGI7660\rover.ASC	ASC file path
	imutxtfilepath		IMU TXT file path
	odofilepath		ODO file path
	outputpath	F:\GINS-Navi\data\WHU20240522-XWGI7660\	Output path
Initial Information	startweek	2315	Start week (or you can use 'auto')
	endweek	2315	End week (or you can use 'auto')
	starttow	291724.0	Start time of week (or you can use 'auto')
	endtow	295006.0	End time of week (or you can use 'auto')
	initpos	[30.528078962, 114.355762445, 40.9652]	Initial position [deg, deg, m] (derived from RTK results)
	initvel	[-0.002, -0.838, -0.028]	Initial velocity [m/s] (derived from RTK results)
	initatt	[0.183, 0.183, 272.55]	Initial attitude [deg] (not used if useAttFromRoughAlign is true)
	initposstd	[1.0, 1.0, 2.0]	Initial position standard deviation [m] (N-E-D)
	initvelstd	[0.5, 0.5, 0.5]	Initial velocity standard deviation [m/s]
	initattstd	[0.5, 0.5, 1.0]	Initial attitude standard deviation [deg]
	initgyrbias	[0.5, -1.0, 6.0]	Initial gyroscope bias [deg/h]
	initaccbias	[-200, 700, 0]	Initial accelerometer bias [mGal]
	initgyrscale	[0, 0, 0]	Initial gyroscope scale [ppm]
	initaccscale	[0, 0, 0]	Initial accelerometer scale [ppm]
	initgyrbiasstd	[50, 50, 50]	Initial gyroscope bias standard deviation [deg/h]
	initaccbiasstd	[250, 250, 250]	Initial accelerometer bias standard deviation [mGal]
	initgyrscalestd	[1000.0, 1000.0, 1000.0]	Initial gyroscope scale standard deviation [ppm]
	initaccscalestd	[1000.0, 1000.0, 1000.0]	Initial accelerometer scale standard deviation [ppm]
	gyrarw	[0.24, 0.24, 0.24]	Gyroscope angle random walk [deg/s/sqrt(h)]
	accvrw	[0.24, 0.24, 0.24]	Accelerometer velocity random walk [m/s/sqrt(h)]
	gyrbiasstd	[50.0, 50.0, 50.0]	Gyroscope bias standard deviation [deg/h]
	accbiasstd	[250.0, 250.0, 250.0]	Accelerometer bias standard deviation [mGal]
	gyrscalestd	[1000.0, 1000.0, 1000.0]	Gyroscope scale standard deviation [ppm]
	accscalestd	[1000.0, 1000.0, 1000.0]	Accelerometer scale standard deviation [ppm]
	corrtime	1.0	Correlation time [h]
Installation Parameters	antlever	[0.164, -0.035, -0.890]	Antenna lever arm [m] (NED)
	odolever	[0.0, 0.0, 0.0]	Odometer lever arm [m]
	installangle	[0.0, 0.0, 0.0]	Installation angle [deg]
Measurement Noise	odonhc_measnoise	[0.10, 0.07, 0.07]	ODO/NHC measurement noise [m/s]
	zupt_vmeasnoise	[0.10, 0.10, 0.10]	ZUPT velocity measurement noise [m/s]
	zupt_wmeasnoise	50	ZUPT angular velocity measurement noise [deg/h]
	CodeNoise	4.0	Code noise [m]
	CPNoise	0.05	Carrier phase noise [m]
Update Frequency	odonhcupdaterate	1.0	ODO/NHC update rate [Hz]
	zuptupdaterate	1.0	ZUPT update rate [Hz]
RTK Post Options	pos1-posmode	kinematic	Positioning mode
	pos1-frequency	l1+2+3	Frequency
	pos1-soltype	combined	Solution type
	pos1-elmask	10	Elevation mask [deg]
	pos1-snrmask_r	off	SNR mask for rover
	pos1-snrmask_b	off	SNR mask for base
	pos1-snrmask_L1	0,0,0,0,0,0,0,0,0	SNR mask for L1
	pos1-snrmask_L2	0,0,0,0,0,0,0,0,0	SNR mask for L2
	pos1-snrmask_L5	0,0,0,0,0,0,0,0,0	SNR mask for L5
	pos1-dynamics	on	Dynamics mode
	pos1-tidecorr	off	Tidal correction
	pos1-ionoopt	brdc	Ionospheric option
	pos1-tropopt	saas	Tropospheric option
	pos1-sateph	brdc	Satellite ephemeris option
	pos1-posopt1	off	Positioning option 1
	pos1-posopt2	off	Positioning option 2
	pos1-posopt3	off	Positioning option 3
	pos1-posopt4	off	Positioning option 4
	pos1-posopt5	off	Positioning option 5
	pos1-posopt6	off	Positioning option 6
	pos1-exclsats		Excluded satellites
	pos1-navsys	59	Navigation systems
	pos2-armode	off	Ambiguity resolution mode
	pos2-gloarmode	on	GLONASS AR mode
	pos2-bdsarmode	on	BDS AR mode
	pos2-arthres	3	Ambiguity resolution threshold
	pos2-arthres1	0.9999	Ambiguity resolution threshold 1
	pos2-arthres2	0.25	Ambiguity resolution threshold 2
	pos2-arthres3	0.1	Ambiguity resolution threshold 3
	pos2-arthres4	0.05	Ambiguity resolution threshold 4
	pos2-arlockcnt	0	AR lock count
	pos2-arelmask	0	Elevation mask for AR [deg]
	pos2-arminfix	10	Minimum number of satellites for AR
	pos2-armaxiter	1	Maximum number of iterations for AR
	pos2-elmaskhold	0	Elevation mask for hold [deg]
	pos2-aroutcnt	5	Output count for AR
	pos2-maxage	60	Maximum age of differential [s]
	pos2-syncsol	off	Synchronized solution
	pos2-slipthres	0.05	Slip threshold [m]
	pos2-rejionno	30	Ionospheric delay rejection threshold [m]
	pos2-rejgdop	30	GDOP rejection threshold
	pos2-niter	1	Number of iterations
	pos2-baselen	0	Base station length [m]
	pos2-basesig	0	Base station sigma [m]
Output	out-solformat	llh	Solution format
	out-outhead	on	Output header
	out-outopt	on	Output option
	out-outvel	on	Output velocity
	out-timesys	gpst	Time system
	out-timeform	tow	Time format
	out-timendec	3	Time decimal places
	out-degform	deg	Degree format
	out-fieldsep		Field separator
	out-outsingle	off	Output single
	out-maxsolstd	0	Maximum solution standard deviation [m]
	out-height	geodetic	Height reference
	out-geoid	internal	Geoid model
	out-solstatic	all	Static solution
	out-nmeaintv1	0	NMEA interval 1 [s]
	out-nmeaintv2	0	NMEA interval 2 [s]
	out-outstat	off	Output status
Statistics	stats-eratio1	100	Error ratio 1
	stats-eratio2	100	Error ratio 2
	stats-errphase	0.003	Phase error [m]
	stats-errphaseel	0.003	Phase error elevation [m]
	stats-errphasebl	0	Phase error baseline [m/10km]
	stats-errdoppler	1	Doppler error [Hz]
	stats-stdbias	30	Standard bias [m]
	stats-stdiono	0.03	Standard ionospheric error [m]
	stats-stdtrop	0.3	Standard tropospheric error [m]
	stats-prnaccelh	1.0	PRN acceleration horizontal [m/s²]
	stats-prnaccelv	0.1	PRN acceleration vertical [m/s²]
	stats-prnbias	1e-05	PRN bias [m]
	stats-prniono	0.001	PRN ionospheric error [m]
	stats-prntrop	0.0001	PRN tropospheric error [m]
	stats-prnpos	0	PRN position error [m]
	stats-clkstab	5e-12	Clock stability [s/s]
Antenna 1	ant1-postype	llh	Position type
	ant1-pos1	90	Position 1 [deg
	ant1-pos2	0	Position 2 [deg
	ant1-pos3	-6335367.6285	Position 3 [m
	ant1-anttype		Antenna type
	ant1-antdele	0	Antenna delta E [m]
	ant1-antdeln	0	Antenna delta N [m]
	ant1-antdelu	0	Antenna delta U [m]
Antenna 2	ant2-postype	llh	Position type
	ant2-pos1	30.528231	Position 1 [deg
	ant2-pos2	114.356985	Position 2 [deg
	ant2-pos3	42.6548	Position 3 [m
	ant2-anttype		Antenna type
	ant2-antdele	0	Antenna delta E [m]
	ant2-antdeln	0	Antenna delta N [m]
	ant2-antdelu	0	Antenna delta U [m]
	ant2-maxaveep	0	Maximum averaging epochs
	ant2-initrst	off	Initial reset
Miscellaneous	misc-timeinterp	off	Time interpolation
	misc-sbasatsel	0	SBAS satellite selection
	misc-rnxopt1		RINEX option 1
	misc-rnxopt2		RINEX option 2
	misc-pppopt		PPP option
	file-satantfile		Satellite antenna file
	file-rcvantfile		Receiver antenna file
	file-staposfile		Station position file
	file-geoidfile		Geoid file
	file-ionofile		Ionosphere file
	file-dcbfile		DCB file
	file-eopfile		EOP file
	file-blqfile		BLQ file
	file-tempdir		Temporary directory
	file-geexefile		GEE executable file
	file-solstatfile		Solution status file
	file-tracefile		Trace file

2 Compilation and Execution

2.1 Source Code and Compilation

The project is managed using CMake. All source code is located in the ./include and ./src directories. It is recommended to use Mingw for compilation.

2.2 Dependencies

On Windows, the software relies on the WIN32 API for window management. Third-party libraries such as Eigen, MatPlotlib-cpp, tqdm-cpp, and thread-pool are included in the ./ThirdParty directory.

The CMakeLists file is fully configured, allowing users to compile the program in one simple step. The recommended IDE is CLion.

2.3 Output Results

The "Output Path" in the configuration file specifies the output paths for navigation results (including STD), IMU error information (including STD), and visualization results (2D and 3D trajectories, elevation, Euler angles [RPY], IMU bias/scale factor errors, and STD).

2.4 Visualization

This C++ GUI program uses the Matplotlib library to automatically generate result plots, including 2D and 3D trajectories, velocity/attitude, IMU errors, and STD. If this affects performance, it can be disabled as detailed in the configuration file.

The 'nav_result.pos' file in the output directory conforms to the RTKlib standard and can be directly imported into rtkplot.exe for RTK result visualization.

3 Datasets

3.1 Test Data

The ./data/ directory contains six sets of IMU data with varying accuracies, including:

• LeadorA15 (navigation grade, high accuracy)
• XWGI7680 (tactical grade, medium-high accuracy)
• CHC CGI-430 (MEMS, low accuracy)
• InvenSense ICM-20602 (MEMS, low accuracy)

The truth.nav file contains reference truth data obtained through backward smoothing.

Configuration files for these datasets are located in the ./conf/ directory.

3.2 Compatible Data Formats

The software supports NovAtel's IMR and ASC format data, as well as custom seven-column txt data (see the i2nav open-source dataset for details).

4 Inheritable Modules

The software provides several inheritable modules, including a progress bar class (tqdm), a configuration manager class (configManager), file read/write classes (fileloader/filesaver), algebra calculation class (algebra), and a plotting class (matplotlib-cpp).

5 Features

• Compatible with both Windows (GUI) and Linux (CUI)

• Detailed parameter settings for the algorithm

• Lightweight and efficient GUI design using native Windows API, similar to RTKlib style

• GUI features include mailto help, GUI progress bar, and robust configuration read/write error handling

• Integrated RTK post-processing module from RTKlib, providing high-precision algorithms and detailed computation strategies

• Object-oriented core code for loosely coupled integration (factory pattern, abstraction, function templates, etc.), with extensive error handling and clean code

• Efficient algorithm implementation using thread pool for parallel task submission

• Single-header inheritable classes for configuration, file read/write, algebra calculation, and tqdm progress bar

• GNSS information quality control (observation information), such as setting absolute and relative thresholds for GNSS position and velocity observation updates, constraining data in the x, y, and z directions to exclude poor-quality data

• Compatible with various inertial navigation data formats (asc, imr, txt)

• Week-crossing data handling (under testing)

• Navigation framework options for n-frame and e-frame (e-frame has a minor bug yet to be fixed)

• Estimation/non-estimation of IMU scale factors (ESKF state vector is 21-dimensional/15-dimensional)

• Robust algorithm tested with multiple datasets: navigation grade, tactical grade, MEMS

---

The algorithm is not perfect, lacking backward smoothing and unsmoothed results.

Further research in the field of integrated navigation is planned during graduate studies.

The current content, which has been developed during my undergraduate studies, remains at a reproduction level.

---