cybergalactic/mss

The Marine Systems Simulator (MSS) is software that supplements the textbook "Handbook of Marine Craft Hydrodynamics and Motion Control," 2nd Edition, by T. I. Fossen, published in 2021 by John Wiley & Sons Ltd.

mss's Issues

Discrepency vs. book

MSS/GNC/llh2flat.m

Line 22 in 263d3fd

Rm = Rn * ( (1-e^2) / (1-e^2 * sin(mu0)^2) );

Hello again Dr. Fossen,

There is a discrepancy between this line of code and equation 2.112 in your Handbook (2nd edition 2021). The latter includes a square-root around the denominator term. I don't have access to the cited reference, (Farrell 2008), but am thinking this may be a typo in the book while the MSS code is correct (in which case, might be useful to note in the function documentation).

Compare to similar algorithm here: https://www.mathworks.com/help/aeroblks/llatoflatearth.html.

Otter in simulink

Dear Professor Fossen：

When I use the Demo of otter in simulink, I was confused in how to calculate revolutions. Firstly, in Control allocation Function, why the output is 'abs(n)*n' and square root subsequently, and why not output 'n' directly, in FIgure1. And in figure.2, what's the meaning of 1/(s+1).

Best Regards
Luo

Ship Irregular ZigZag Maneuvering Test

Dear Professor,

How can I configure the ExZagzig file you published so that the ship model can perform irregular ZigZag tests? I attempted to change only the rudder angle, but it was ineffective.

How to get hydrodynamic derivatives from vessel struct?

Hey @cybergalactic,

I am trying to use MSS with my own vessel. For the same, I want to know how to get the hydrodynamic derivatives (https://github.com/cybergalactic/MSS/blob/master/VESSELS/mariner.m#L57-L73) from the vessel struct (For instance, https://github.com/cybergalactic/MSS/blob/master/HYDRO/vessels_wamit/tanker/tanker.mat). I am asking as the vessel structure does not have variables with names associated to the hydrodynamic derivatives. Is there a way to extract the hydrodynamic derivatives from the vessel struct?

I am looking forward to your response.
Regards

about beta_c

dear professor Fossen, sorry to trouble you again, but i found a question about the ILOS guidance:
in your text book, it says:

which defines the crab angle and when can we see it as a'slowly varying unknown parameter'. But in file'demoOtterUSVPathFollowingCourseControl',

we know that the psi+beta_c =chi, but the derivative of chi 'omega_chi' uses r as input, which means the derivative of chi is r, but consider the function above, i don't know the reason why we can simplify here, please tell me the reason , and can i use

to design the controller ?
thank you

Is there any model descriptions in VESSELS folder?

Hi:
Thanks for the MSS projects. Where can I find the model description for each file in VESSELS folder? I am looking for a model with only one thruster and one rudder. Which file should I refer to?

Find a problem in mariner.m

In the file of mariner.m,line 86 .There is an equation"delta_dot = delta_c - delta;",using this equation to calculate the change rate of the rudder,but this equation doesn't take the time integration time into account.If the integation time is not equal to one sec,then the equation is wrong,am I right?

Control-Autopilost model

Hi Prof. Fossen,

I use the "Course autopilot" module in the MSS Simulink Library /Control/Autopilots. Can you help me to explain how the module work? Besides, where is the source code of the module?

Thank you very much.

Question about the difference between zero-frequency and infinite-frequency added mass matrix

Hi @cybergalactic,
I'm really confused about the physics meaning of the zero-frequency added mass matrix $A(0)$ and infinite-frequency added mass matrix$A(\infty)$ . Also, the physics meaning of the zero-frequency potential damping $B(0)$ and infinite-frequency potential damping $B(\infty))$.

In my imagination, the zero-frequency added mass matrix $A(0)$ and the zero-frequency potential damping $B(0)$ are the added mass matrix and potential damping when the vehicle in calm water, right? Then, what's the physics meaning of the infinite-frequency?

Best regards

I can not find the simulink file?why?(i am chinese,i am not good at english.)

about the payload mass and heave velocity

Dear professor Fossen. I have a question about the linear speed. You know, I used to discuss the model 'Otter.m' for quite a few times, however, I found a question about the payload mass 'mp', that is , when I input the mp, which is not zero, then the ship began to have a heave speed which is not converged to 0, also ,when I increase the mp,the heave speed began to decrease and the surge speed increased, and I don't really know the reason, even if I check the codes for a while and the text book you wrote to check the equations ,i still dont find the answer. ps: the model is created based on ‘otter.m' file.Here is the simulation results:

in the past, I don't consider the payload mass, so i miss that porblem, that is the reason why i come here again.

m2c.m matrix shape error

from line 37 to line 43

M21 = M12';
M22 = M(4:6,4:6);

nu1 = nu(1:3);
nu2 = nu(4:6);
dt_dnu1 = M11*nu1 + M12*nu2;
dt_dnu2 = M21*nu1 + M22*nu2;

＞＞＞ I think it should be

M21 = M12.';
M22 = M(4:6,4:6);

nu1 = nu(1:3);
nu2 = nu(4:6);
dt_dnu1 = M11*nu1.' + M12*nu2.';
dt_dnu2 = M21*nu1.' + M22*nu2.';

GNU Octave Support

GNU Octave is free software made to be compatible with Matlab scripts.

https://www.gnu.org/software/octave/

Hence, most MSS functionality should be compatible out of the box. However, there are suble differences between the two languages that might cause problems. I think MSS should be tested for Octave compatability. This will make the software usable by a wider audience, as Matlab is quite expensive.

At least the repo should be tagged with the Octave language tag.

path following

Excuse me, can you upload the code about LOS? There is very little information on the whole network.

About the Hoerner function in the simulation of Otter.m

Dear professor, when i compute the cross-flow drag in the otter.m while using the 'crossFlowDrag' function, i found something uncommon, that is:
'
rho = 1025; % density of water
n = 20; % number of strips

dx = L/n;
Cd_2D = Hoerner(B,T); % 2D drag coefficient based on Hoerner's curve
'
in this part, we need to calculate the result of Cd_2D, which is using the Hoerner function to interpolate the Hoerner curve to get the Cd_2D. However, in this part, if we do not apply payload mass, then the B and T (which i have calculted) are:
'B=1.08(m), T=11/82=0.1341463415(m)'
so that we can get B/(2T)=4.0254545444>4.00309.
but in the function Hoerner ,we only get：
‘% CD_DATA = [B/2T C_D]
CD_DATA = [...
0.0108623 1.96608
0.176606 1.96573
0.353025 1.89756
0.451863 1.78718
0.472838 1.58374
0.492877 1.27862
0.493252 1.21082
0.558473 1.08356
0.646401 0.998631
0.833589 0.87959
0.988002 0.828415
1.30807 0.759941
1.63918 0.691442
1.85998 0.657076
2.31288 0.630693
2.59998 0.596186
3.00877 0.586846
3.45075 0.585909
3.7379 0.559877
4.00309 0.559315];’
which may get a NaN result when calculating, and i have examined the original reference paper, and i found the curve hasn't the part that over 4.00309, so i wonder how to solve this dilemma while using the 'Hoerner' function.
thanks!

Accuracy of simulation results

Hi Prof. Fossen,

I want to use the motion data, such as velocity、acceleration and position, which caculated from the "otter.m". I'm wondering these results are accurate enough to show the otter USV navigation in the reat environment.

Thank you very much.

How to build my own model of vessel

Thanks for the toolbox, It is amazing!
I'm interested in building a model of my vessel. I'm a student and want simulate my vessel RC. In toolbox there is a model of Mariner and other ship, there is some way to build a 3 dof model of any ship? How compute the values?, there is an script or something else that help me to build a model for my vessel? Thanks @cybergalactic

For example in mariner.m is the model of this kind of ship, there a way to build some similar but with other ship?

Question regarding ExPathGen.m

Hi Prof. Fossen,
I'm following the example ExPathGen.m (described in Example 10.4 of the 1st edition of the handbook).
I wonder how to use the formulation to perform a deceleration trajectory, i.e., having initial speed U=5 at waypoint 0 (th=0) and Uref = 0 at waypoint 1 (th=1).
The issue I'm having is determining the parameter th, at which the deceleration starts.
Any guidance on how to properly formulate this case?

Vertical velocity used in cross-flow drag calculation for AUVs

Hi Prof. Fossen,

I noticed that in this commit you included heave force and pitch moment to crossFlowDrag which are relevant for AUVs. Regarding this point, I have two questions:

1. Shouldn't U_v use the heave relative velocity w_r = nu_r(3) and pitch rate q = nu_r(5), i.e.: U_v = abs(w_r + xL * q) * (w_r + xL * q);, or am I missing something?

2. Considering that a low aspect-ratio wing approximation that takes the vehicle's angle of attack as an input is being used to compute lift/drag/pitch moment for the hull, would such effects (extra heave force and pitch moment) already be accounted for, or do they have a different nature?

Thank you in advance.
Best regards.

wamit2vessel.m mirrors RAO incorrectly to 190 to 350

Dear Professor,

This issue is related to: wamit2vessel.m

WAMIT deliveres RAOs from 0 to 180 deg. I noticed that the motionRAO, forceRAO en driftfrcRAOs are mirrored incorrectly to 190 to 350.
The heading, sign and phase are all incorrect in some cases.

I solved this by the following code from line 669:
%-------------------------------------------------------------------------%
%% map data from 0-180 deg to 180-360 deg (symmetry about x-axes)
%-------------------------------------------------------------------------
vessel.headings = (pi/180)*[0:10:350];

% To change sign phase is shifted 180deg
rao_sign = [0 1 1 1 0 0];
for i = 1:6
vessel.motionRAO.amp{i}(:,20:36,1) = vessel.motionRAO.amp{i}(:,flip(2:18),1);
vessel.motionRAO.phase{i}(:,20:36,1) = rao_sign(i)*pi+vessel.motionRAO.phase{i}(:,flip(2:18),1);
vessel.forceRAO.amp{i}(:,20:36,1) = vessel.forceRAO.amp{i}(:,flip(2:18),1);
vessel.forceRAO.phase{i}(:,20:36,1) = rao_sign(i)*pi+vessel.forceRAO.phase{i}(:,flip(2:18),1);
end

rao_sign_WD = [1 -1 -1];
for i = 1:3
vessel.driftfrc.amp{i}(:,20:36,1) = rao_sign_WD(i)*vessel.driftfrc.amp{i}(:,flip(2:18),1);
end
clear rao_sign_WD

Question regarding reference for compass linearization

In ins_mekf_psi lines 79-86:

% Linearization of compass measurement
a = (2/q_ins(1)) * [q_ins(2) q_ins(3) q_ins(4)]'; % 2 x Gibbs vector
u = 2 * ( a(1)*a(2) + 2*a(3) ) / ( 4 + a(1)^2 - a(2)^2 - a(3)^2 );
du = 1 / (1 + u^2);
c_psi = du * (1 / ( 4 + a(1)^2 - a(2)^2 - a(3)^2 )^2 ) * ...
        [ -2*((a(1)^2 + a(3)^2 - 4)*a(2) + a(2)^3 + 4*a(1)*a(3)) 
           2*((a(2)^2 - a(3)^2 + 4)*a(1) + a(1)^3 + 4*a(2)*a(3)) 
           4*((a(3)^2 + a(1)*a(2)*a(3) + a(1)^2 - a(2)^2 + 4)) ];  

Do you have a reference for this transformation to the "2x Gibbs vector" and the following linearization? I looked through "Handbook of Marine Craft Hydrodynamics and Motion Control" but couldn't seem to find anything in the quaternion sections.

The orgin point of the "Otter. m"is not clear

Sorry for bothering again, professor Fossen.
As far as I know , the orgin point of a vessel is usually set in the center of gravity or the center of the water plane, but the rp is set as' rg = [0.2 0 -0.2]'; % CG for hull only (m)'. So I wonder where it is set since i don not the exact data even in the product introduction in www.maritimerobotics.com. It is siginificant cause I want to know some other coordinates in the vessel. Thanks.

About the motion response ampliotude operators

Thanks for the convenience you offerd.
I'm a doctoral student and I'm reading a book written by your team - 'Ship Motion Control'.
In chapter 4, the book mentioned 'Motion RAO' were given as a function of the wave frequency in the toolbox, I want to study RAO by the function but I can't find it.
Do I omit it or it has been removed ?

Ability to simulate a manipulator attached to a USV

Hi Prof. Fossen,

Would it be possible to simulate a serial manipulator attached underneath a USV with this library in such a way that the hydrodynamics is applied to the manipulator as well?

Thank you very much.

Branch(es) to support previous releases of The MathWorks toolchain

Dr. Fossen,

Thanks for the repository.

FYI, I used portions of this over 10 years ago and am thrilled to see it is still actively being developed!

I have some project needs that require prior release (i.e., r2015a, r2018b, etc.).

FWIW, I have written scripts to search the directory structure, identify mdl/slx models/libraries, do a load/save as, etc.

If it will be of use, I can provide some scripts for you. We can iterate as necessary to get to a point where a clean checkout of a branch works for my purposes (e.g., r2015a, r2018b) and that should give some assurances that other TMW release branches should be easy to facilitate.

With gratitude.

Mike Schaefer

about the SIMotter,m file

Question about mariner.m

When I read the annotation about the function mariner(x,ui,U0).I found "time derivative of the state vector: x = [ u v r x y psi delta n ]'". But there is only seven parameters in x. Where is the parameter "n"? Is there a mistake?

about 'otter.m'

Hi there，professor Fossen, I found that the motion equation of otter was using the function of follows:

so i got 2 questions about it , because i wanna use it to simulate and finish my paper.
The FIRST is, in your book , it says:

therefore, could i say that, the model you use in the otter.m can only be used in motion of line motion? what if the USV follows a curve? Can i use it to follow the Polyline path？Or i have to use the following equation in your book :

and reconstruct the model?
The SECOND is , i didn't know much about the crossflow drag and strip theory:

but i had looked into some paper and found the strip theory is often used in the calculation of wave disturbance, so i wonder if the crossflow drag the same as wave disturbance? Also, i didn't know where the origin of the equations here:

though i know the Cd_2d coefficient is based on a curve of a paper.
Besides, i have some other questions about the model, can you reply me the answer even though i know your time is limited and precious?
THIRD, you know the otter model is a 6DOF model, but usually we control the model of 3 degree, so it needs to simplify the model and transform it into 3 DOF, but i wonder how to decouple the 6DOF model into 3DOF, the only way i could do is to delete the rows and columns of w\p\q ， i wonder if there is better way to do that?
FORTH, about the hydrodynamic coefficients in this model.
i know all of the hydrodynamic coefficients you use have their experiments and statistics support, however, if i have to consider the variable mass of mp,which may change the draft , as a result , the hydrodynamic coefficients as well as the draft may be variable all the time, so i wonder if these equations effective while the mass is changing?If it is not effective, how can i consider the hydrodynamic coefficients of otter?

Thanks !

IDENTIFICATION OF DYNAMICALLY POSITIONED SHIPS paper

I am MSc student in Control engineering
I have recently came across your quite interesting papers “ IDENTIFICATION OF DYNAMICALLY POSITIONED SHIPS ” I wonder if there exists any source code or other useful documents . And if so, I appreciate if you kindly let me know how I can have access to them I need those information to be able to validate my method . I need to compare my results to the correct values.

Munk moments

Hi Prof. Fossen,

While having a look at Munk moments and it's destabilizing effect on simulation of UUVs, I came across this simplification for the Remus. I believe the term CA(5,4) was supposed to be CA(5,3), referring to term -X_udot * u_r in equation 6.54 (1st edition) or 6.60 (2nd).

For my case, I'll need to do a similar approach to compensate for the unmodeled nonlinear damping, and I was wondering if it could be generalized also for the roll part (i.e. CA(4,2) and CA(4,3)), zeroing the whole bottom left part of the CA matrix. Any though on that?

Best regards

run "demoOtherUSVHeadingControl" error,maybe exist problem

Reference request for navigation files.

Thank you for sharing the great resources.

Is there any paper or books for the following navigation filters, including ins_ahrs, ins_euler, ins_mekf, and ins_mekf_psi?

I would like to know the detail about the Jacobian matrix of the aforementioned filters.

• Navigation (m-files)
• acc2rollpitch static roll and pitch angles from specific force
• gravity acceleration of gravity as a function of latitude using the WGS-84 ellipsoid parameters
• ins_ahrs Error-state Kalman filter for INS aided by position and AHRS measurements
• ins_euler Error-state Kalman filter for INS aided by position and yaw angle measurements
• ins_mekf Error-state Kalman filter for INS aided by position and magnetic field measurements
• ins_mekf_psi Error-state Kalman filter for INS aided by position and yaw angle measurements
• insSignal basic INS signal generator

Undesirable overloading of run.m

Hi,
HYDRO/vessels_wamit/fpso/run.m overloads the built-in matlab function run (ref). Could this be fixed by e.g. renaming?

>> run my_own_script_not_at_all_related_to_mss.m
Attempt to execute SCRIPT run as a function:
/path/to/mss/HYDRO/vessels_wamit/fpso/run.m

Thanks,
Kristoffer

Find some questions in the reference book

Hi professor fossen
In your book **Handbook of Marine Craft Hydrodynamics and Motion Control (2011) ** I have some questions

1. P125 equation (6.76), (6.77) and (6.81), the coefficient maybe should be $2 \pi$ not $8 \pi$ ?
2. P125 equation (6.78)-(6.80), the coefficient 2 maybe should delete?
3. P126 figure 6.4, the second figure's legend equation should change into $C^{new}_f=(A_x/S)C_x$ or something. The equation $C^{new}_f=(A_x/S)C^{new}_f$ doesn't make any sense.
4. P208 equation (8.66), I just want to know how to choose the value of $\delta\omega$ to compute $A_k$?
   Looking forward to your reply.

Can I use the Remus100.m file to perform zigzag and turn circle maneuvers on a horizontal plane?

Dear Professor Fossen：

I want to study the characteristics of horizontal motion of the Remus100 underwater vehicle. Can I use the Remus100 file to perform zigzag and turn circle simulation experiments?
If possible, how should I modify the existing ship model file? Could you provide some guidance or ideas?

Best Regards
Heng Wang

A misspell in the README.md

There is A misspell in the explanation of Hydro Library the README.md. The word templastes should be changed into templates.

find a problem in container.m

In container.m line132 there is an equation "uP = cos(v)((1 - wp) + tau((v + xpr)^2 + cpvv + cpr*r));" .But I found the same equation in the reference "Son og Nomoto (1982). On the Coupled Motion of Steering and Rolling of a High Speed Container Ship, Naval Architect of Ocean Engineering,20: 73-83. From J.S.N.A. , Japan, Vol. 150, 1981.",but the equation is

,so "cos(v)" should change into "u".

The Otter model's Yv is 0, which may cause a non-converged roll curve when i add an addtional force in Y direction

Sorry for trouble again, but when i checked the 'Otter.m' model this evening, i put an load in the Y direction in the body-fixed coordinate system the roll angel and sway velocity is going to be larger and larger, so i looked over the code and find that in the 'Linear damping terms', the Yv is set to 0, i guess the reason behind the phenomenon is the damping coefficient is set to 0, so i try to copy the Xu 's expression to Yv and run the simulation,then i found the roll angle and the sway velocity is converged to 0 but slowly.
Accordingly, i want to know if i am right. If so, i also want to know how to set the Yv(Yv's expression) or how to deal the situation.

q2euler.m: 'theta' is not assigned on some execution paths.

Hi,
The addition of the if-statement in q2euler.m to handle NaN due to round-off errors causes Simulink to give an error: "Output argument 'theta' is not assigned on some execution paths.".

Regards,
Sif Högnadottir

C_A matrix

MSS/VESSELS/remus100.m

Line 106 in 04282b4

[MA,CA] = imlay61(a, b, nu, r44);

Hello Dr. Fossen,
First off, thanks very much for your excellent textbook and this accompanying toolkit. I've been using them as my primary reference for a model of dynamics and control of a small AUV, similar in concept to the Remus 100. I'm wondering, should this call to imlay61 pass in the relative velocity vector nu_r instead of nu? I'm looking at equation 6.59 and 6.60 from your text, which express MA and CA purely in terms of nu_r, with MA assumed to be a diagonal matrix as it is in imlay61.m

The reason I ask is that when CA is calculated this way using the absolute velocity, the terms in the lower left quadrant of CA cause angular moments that are dependent on water current. For example, if the vessel were drifting with a fast water current (zero relative velocity) that had non-zero u,v,and w components in the body frame, substantial angular moments arise from CA which doesn't seem intuitively correct.

Thanks again,
DC

something wrong

In the LOSchi.m ,at the begin of this program, the "compute" is wrong and you write it in "ompute"

ask for help

hello , could you tell how me how to caculate wave force and moments, current force and moments,I have some problems when I do it. If proper, give programs . Thanks very much!

U problem in "otter.m" and "SIMotter.m"

I’d like to point out a question in "otter.m"and "SIMotter.m"
'U = sqrt(nu(1)^2 + nu(2)^2 + nu(3)^2); % speed'
if this is used for LOS guidance law or etc., then the U means the total speed of horizontal motion ,so I think this should be replaced by:
'U = sqrt(nu(1)^2 + nu(2)^2 ); % speed'
cause in the LOS, the U defines as :

A request for a documentation resource

Dear Prof.Fossen ：
I ’m a Ph.D. in the direction of unmanned vehicles. I'm very interested in your project. When I was learning the file named clarke83.m. I found you give a reference "Clarke et al. (1983)". But I can't find the arcticle anywhere. Could you please give me more detailed information about it so I can find the arcticle?
Thanks for your time.
Sincerely yours,wtllll.

Model Otter has a pitch angle that is not 0.

Hi there , professor Fossen, i have a problem that the model Otter has a itch angle that is not converged to 0，even its thrust power is se t to zero, is something went wrong or is just normal?

Question: Hull lifting forces

Hi @cybergalactic,

I'm working on a 6-DOF simulation model for a flatfish-shaped underwater vehicle and using your textbook as a reference. I'm able to obtain fair estimates for the added mass matrix coefficients using hydrodynamic software and I can also determine rough estimates for linear viscous damping coefficients based on existing data of similar vehicles - being an underwater vehicle, linear potential damping is being neglected. To determine non-linear damping effects though, I'm using surge resistance and cross-flow drag principles.

However, as far as I could understand, in many modeling approximations of D(v) = D + Dn(v) the effects of lift on the vehicle's hull are not explicitly accounted for. From your textbook:

Lifting Forces: Hydrodynamic lift forces arise from two physical mechanisms. The first is due to the linear circulation of water around the hull. The second is a nonlinear effect, commonly called cross-flow drag, which acts from a momentum transfer from the body to the fluid. This secondary effect is closely linked to vortex shedding.

What I mean is that I don't see where is the lift generated from the hull moving in the fluid with a certain angle of attack being considered in the models, and I would assume that they play an important role for survey AUVs (moving the vehicle up-down). To my understanding, the elevators would be responsible for generating a pitch moment, but the most representative lift would be generated by the hull; which would have a way bigger area than the elevators of a flatfish-shaped AUV.

The following line is the only reference I could find so far that actually models the vehicle's hull as a "wing profile" and considers its lift-drag as a function of the vehicle's angle of attack.

MSS/VESSELS/remus100.m

Line 132 in d03cfc3

tau_liftdrag = forceLiftDrag(D_auv,S,CD_0,alpha,U_r);

Q: Is my understanding correct thus far? If so, why are aoa-dependent lifting forces not being considered using this for other vessels? Are these effects being accounted for elsewhere in other vessels models, such as in the NPS AUV which does consider cross-flow drag?

I really appreciate the help in advance!

EditAB.m

Thank you for sharing the MSS. After clicking on the plot twice to define the low and high frequency, there is no RETURN button in the figure.

A request for a reference paper

Dear professor Fossen. I'm a Phd student. I'm really interested in the project. But I can't find any website that can download the reference paper "Van Berlekom, W.B. and Goddard, T.A. (1972). Maneuvering of Large Tankers, Tansaction of SNAME, 80:264-298" in tanker.m. Could you please provide me any website that can download this paper?
Anyway, thanks a lot !

Can I use MSS for simulating free falling objects under water?

Dear Prof. Fossen,

The free falling of an object underwater should be much easier, no power and no target. Isn't it?
Can I make the simulation by some 'simple' modification to the code?

Regards
Huisky

I can not find the