Research on Error Correction Technology in Underwater SINS/DVL Integrated Positioning and Navigation

Round 1

Reviewer 1 Report

The research presented in paper are interesting, however at the moment rather of local meaning.

In my review I'd like to propose a few amendments that will improve soundness of the paper.

 

First of all more thorough biblioraphical review should be made to provide the background for your research.

 

- Introduction - more referneces hsould be given in Introduction, especcially for the integration part, where references are missing in fact

- Indroduction doesn't give sufficient evidence of the gaps in state of art.

- Indtroduction - it is not clear which aprticular algorihtms will be tested for integration and filtering and why it is important

- it is not clear what is given in eq. 1 and what is the relationship of it with eq. 2

- eq.2 - F, G, W should be presented as it is for H and V

- eq.3 - Xsins and Xdvl should be directly preented

- eq. 11, 12 - please describe covariance matrix P in details

- sec. 4.4 - DGPS doesn't work underwater - please explain thus how was the data obtained

- table 2 - please provide exact locaiton of testbed and closest DGPS station to prove DGPS acurcy in this area

editorial remarks:

- word thesis is overused, probably paper or research would be better

- automatic references are lost, at least in my version

- there are some grammar issues in the paper

Author Response

1.I have added more references in Introduction :In references 12-19 , these references mainly introduce the gaps in state of art ,by analyzing the references ,I gave the current mainstream methods and research results,in Line 66-78,the added references as follows，changed reference 5,15,18,21 for the integration part.

• ANDRLE MS. CRASSIDISJ L. Attitudeestimation employing common frame errorrepresentations[J].JournalofGuidance,Control, andDynamics，2015,38(9):1614-1624.
  • ZHU T G, LIU Y， LI W K,et al.Thequaternion-basedattitudeerrorfor thenonlinear error model of the INS[J. IEEESensors Journal,2021,21(22): 25782-25795.
  • ANDRLE MS. CRASSIDISJ L. Attitudeestimation employing common frame errorrepresentations[J].JournalofGuidance,Control, and Dynamics，2015,38(9):1614-1624.
  • WANG M s, WU W Q，ZHOU P Y, et al.State transformation extended Kalman filterfor GPS/SINS tightly coupled integration[J].GPS Solutions,2018,22: 112.
  • CHANGL B,QIN FJ,JIANG S.Strapdowninertial navigation system initial alignmentbased on modified process model[J]. IEEESensors Journal,2019,19(15):6381-6391.
  • ZHAO R J，LIK L，HU B Q, et al.SINSinitial alignment algorithm based on improvedquaternion damping error model[J]. SystemsEngineering and Electronics，2021，43(11):3330-3337.
  • ZHAO Jun-bo，GE Xi-yun,FENG Xue-lei,et al，AReviewof Underwater SINS/DVL IntegratedNavigation Technology [J]. Journal of UnmannedUndersea Systems, 2018,26(1): 2-9.
• SONG Jin-xue.Model-aidedand Sea CurrentEstimation for UUV Autonomous Navigation Research[D]. Harbin Engineering University, 2018.

 

2.By analyzing the references ,I gave the current mainstream methods and research results, as shown in Line 66-78.

• InDVL/SINS integrated positioning and navigation system, the state equation of Kalman filtering is composed of SINS error equations. According to the different expressions of attitude errors, the main SINS error equation are Euler angle error models and quaternion error models [12] [13]. Tothe problem of precisionin traditional SINS error models, ANDRLE MS proposed an error construction idea based on vector operation coordinate system consistency[14]. On this basis, WANG M s points out that the traditional definition of SINS velocity error only considers the difference between the actual velocity and the theoretical velocity, ignoring the impact of inconsistent vector coordinate systems[15]. CHANGL B and ZHAO R J defined velocity error in the computational navigation system, derived the Euler angle and quaternion forms of the GNSS damped SINS nonlinear error model, and applied them to GNSS/SINS large misalignment angle alignment [16] [17].To the instability of the DVL signal, ZHAO Jun-bo and SONGJin-xue proposed using other methods to replace the DVL signal in order to get more accurate position[18] [19].

 

3. The method I used is showed in line 82-92, using several DVL error corrections, which

can effectively improve the final accuracy of the combined positioning and navigation system.

• a) Using the sound speed measuring instrument to obtain the sound speed information of the carrier, and compensating the sound speed error in real time. After correction, the speed measurement data of DVL is closer to the carrier’s setting speed; b) Studying the correction method of the carrier attitude error, using the change data of carrier roll angle, pitch angle and heading angle obtained by the attitude sensor, DVL is compensated for the attitude error. The corrected carrier's navigation trajectory is closer to the real trajectory; c) Studying a covariance-based outlier detection and removal method for the mutation outliers in DVL data. Through simulation analysis, this method can effectively remove DVL mutation data and ensure the validity of the data.

 

4. I have reproduced and added the formula.
5. eq. 1 determines the state variables of the system ,then we can establish the state equation of the SINS/DVL integrated navigation system eq. 2.
6. F, G, W has been presented as it is for H and V,as shown in Line 249-250.

• ,      

             

1. Xsins and Xdvl have been directly preented in Line 254:

• .

 

394.  We have described covariance matrix P in details in Line 393-394.

• is the predicted value of Kalman filter variance matrix at time k, , is Kalman Gain.

 

5. Although Differential GPS doesn't work underwater ,but we can get it when AUV navigation on the water surface.

 

6. The exact locaiton of testbed is ShazhouLake,Suzhou,China. The exact longitude and latitude is 120.568061,31.892607.

 

7.I've replaced word thesis with word paper, in Line 21,79,112,196,328,337,536,566 and 571.

 

8. Automatic references are used in this paper.

Author Response File: Author Response.pdf

Reviewer 2 Report

Presented paper “Application of “Research on Error Correction Technology in Underwater 2 SINS/DVL Integrated Positioning and Navigation” exposes comparative annalise of methods in position fixing of an object, in this case underwater object. Filtering process of used data is well observed and explained with focus on position fixing. Diagrams and equations are well visible and readable. Cited references are ordered in the section as per the standard of the journal.

Despite of the above I have the following remarks to the authors reviewing this article:

1. The abstract of the manuscript doesn’t present the goal of the research. It is vital important for every scientific article, which addressing new idea to the readers.

2. No references in the text are shown. Instead of cited references a message “Error! Reference source not found” appears in many positions in the text (L35, L48, L56 and so on).

3. No literature review was found in the manuscript. The authors have taken huge space to explain already known theory – Kalman filter for example. I propose revision of this part of the manuscript in order to present the readers existing problems and original way for solution.

4. Figure 7 and figure 8 have to be replaced with better resolution pictures.

5. The authors have to explain their main idea in the research and they have to follow it consequently in the text. I couldn’t get the main goal of the research and couldn’t understand whether the goal has been reached or not.

6. The authors have to explain their experiment in more details – why sprite E200D small AUV has been chosen, what is its advantage compared with other vessels, technical data of the device.

7. Discussion and conclusion are not clear defined.

8. My general impression is that the information in the manuscript is not good ordered. I get that feelings that something interesting is done, but it is bad explained and bad presented finally.

Author Response

1.The abstract of the manuscript has presented the goal of the research. As shown in Line19-21,the goal is to correct the error of the combined positioning and navigation system. Improving the accuracy of underwater vehicle positioning and navigation always is an important issue that needs to be addressed in the development of underwater vehicles. This paper mainly takes the SINS/DVL integrated positioning and navigation system as the starting point, analyzes in detail the error sources of DVL in the system, focuses on studying DVL error correction methods.

• Therefore, error correction technology has great significance for underwater inspection and operation tasks.

 

 

2. Automatic references are used in this paper.

 

3. Literature review has added in Line 66-78.By analyzing the references ,I gave the current mainstream methods and research results, also I have adjusted the format.

• InDVL/SINS integrated positioning and navigation system, the state equation of Kalman filtering is composed of SINS error equations. According to the different expressions of attitude errors, the main SINS error equation are Euler angle error models and quaternion error models [12] [13]. Tothe problem of precisionin traditional SINS error models, ANDRLE MS proposed an error construction idea based on vector operation coordinate system consistency[14]. On this basis, WANG M s points out that the traditional definition of SINS velocity error only considers the difference between the actual velocity and the theoretical velocity, ignoring the impact of inconsistent vector coordinate systems[15]. CHANGL B and ZHAO R J defined velocity error in the computational navigation system, derived the Euler angle and quaternion forms of the GNSS damped SINS nonlinear error model, and applied them to GNSS/SINS large misalignment angle alignment [16] [17].To the instability of the DVL signal, ZHAO Jun-bo and SONGJin-xue proposed using other methods to replace the DVL signal in order to get more accurate position[18] [19].

 

4. Figure 7 and figure 8 have replaced with better resolution pictures.

 

5. We added the main idea of the research in Line 79-92.

• In this paper, mainly focus on DVL error correction to improve the accracy ofDVL/SINS integrated positioning and navigation system ,a feasible integrated positioning and navigation system is designed by studying SINS/DVL integrated positioning and navigation filtering algorithm. To the factors which affect the speed measurement accuracy of DVL, establishes the DVL error model, compensates the error by the algorithm comprehensively :a) Using the sound speed measuring instrument to obtain the sound speed information of the carrier, and compensating the sound speed error in real time. After correction, the speed measurement data of DVL is closer to the carrier’s setting speed; b) Studying the correction method of the carrier attitude error, using the change data of carrier roll angle, pitch angle and heading angle obtained by the attitude sensor, DVL is compensated for the attitude error. The corrected carrier's navigation trajectory is closer to the real trajectory; c) Studying a covariance-based outlier detection and removal method for the mutation outliers in DVL data. Through simulation analysis, this method can effectively remove DVL mutation data and ensure the validity of the data.

 

6. As shown in Table2, we think sprite E200D small AUV is fully capable of completing experiments, It has more precise parameters compared with other vessels.
7. we added the main work content of this paper

 in Line541-559 to define the discussion and conclusion more clearly.

 

• The main work content of this paper is as follows:

1. Combined with the working principle of SINS and DVL, deducing the basic equations, and establishing their respective error models. On this basis, designing the SINS/DVL integrated positioning and navigation system.
2. Studying the DVL error correction method of integrated positioning and navigation system. Firstly, analyzing the source of DVL speed measurement error, and then studying the correction and compensation methods for each error factor: a) Using the sound speed measuring instrumentto obtain the sound speed information of the carrier, and compensating the sound speed error in real time. After correction, the speed measurement data of DVL is closer to the carrier’ssetting speed; b) Studying the correction method of the carrier attitude error, using the change data of carrier roll angle, pitch angle and heading angle obtained by the attitude sensor, DVL is compensated for the attitude error. The corrected carrier's navigation trajectory is closer to the real trajectory; c) Studying a covariance-based outlier detection and removal method for the mutation outliers in DVL data. Through simulation analysis, this method can effectively remove DVL mutation data and ensure the validity of the data.
3. Applying the above-mentioned DVL error correction method for overall correction, using AUV to carry out the lake test, the lake test results verify the effectiveness of the DVL error overall correction method.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Not all of my remakrs were fully adrressed.

- eq. 1 - you should call it determination of state variables, because that's what it is

- covariance matrix - please explain how do you calculate initial covariance matrix, as it is still not clear

- DGPS issue on UAV - does your response mean that UAV was floating on the surface, acting as USV in fact? You should explain it in text directly to make it clear to readers

- reference station - again please give information about DGPS reference station and the distance to it from the test bed

moderate - should be edited

Author Response

1. We have changed Equation of state as Determination of state variables, in Line 217 and 220.

• Determination of state variables

 

2. We have explained how to calculate initial covariance matrix, as shown in Line384-396.

• In the combined system, the difference between the measured value of Kalman filter and the measured value of DVL is called Innovation, which represents the error of measurement estimation.

The discrete filtering model of SINS/DVL combined system is as follows:

                                     

Further state prediction:

State prediction mean-square error matrix:

Filtering gain:

State estimation:

                                             

State estimation mean-square error matrix:

 

3. The AUV was designed to run on the surface, acting as USV. So we can get DGPS’s outputting information. We showed in Line 520-521.

• The AUV was designed to run on the surface in order to get DGPS’s outputting information.

 

4. We added The AUV action path and nearest DGPS reference station in Figure 13, and the exact latitude and longitude of DGPS reference station is 31°89’21"N, 120°56’04"E. The distance from the test bed to it nearest is 18.23 m and farthest is 747.31m.

• The action path is shown in Figure13, the blue line is the route of the experiment and the yellow dot is the nearest DGPS reference station. The latitude and longitude of DPGS reference station is 31°89’21"N, 120°56’04"E.

 

5. We have replaced some pictures with better resolution pictures. Such as Figure 1,7,8,9,10,12,14.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

I am satisfy with authors' answers and comments.

Author Response

Thank you for your comments.