How to add minimum radius of curvature limitation to 3D spline interpolation?

I have 3D points data that a line should pass through these points. For this purpose I am using following code:

n = numel(xxx) ;
L = zeros(n,1) ;
for i=2:n     
    arc_length = sqrt((xxx(i)-xxx(i-1))^2+(yyy(i)-yyy(i-1))^2+(zzz(i)-zzz(i-1))^2);     
    L(i) = L(i-1) + arc_length; 
end
% Normalize the arc lengths
L=L./L(n);
% do the spline 
x_t = spline(L,xxx) ;
y_t = spline(L,yyy) ;
z_t = spline(L,zzz) ;
% for interpolation 
tt = linspace(0,1,500) ;
xi = ppval(x_t,tt) ;
yi = ppval(y_t,tt) ;
zi = ppval(z_t,tt) ;
plot3(xi,yi,zi,'.b') ;

However, this line mimics seabed pipeline, therefore, there is a limitation for minimum radius of curvature. I need your opinions that how can I add that limitation to interpolation.

[darova]

Please attach the data and some pictures

[Shamsi Musayev]

I have attached one set of input data but there will be a lot of this type of data. And graph illustrates result of my interpolation for uploaded set of data. I can write a code that it will check radius of curvature through determined points and if there are points that excess limitation it will remove those interpolated points for decreasing radius of curvature. However, I want to know that maybe there is a way that I can check that requirement during interpolation process.

[darova]

The idea with curveture radius is ok. I don't have better ideas