I am trying to simulate Random Road Profile Grade-C as per ISO 8608 using Shaping Filter Method equation (Eq. 15 of attachement).
The value of "Alpha" is taken as 0.127 from Table.3 shown in attached image, "Velocity of vehicle, v = 20 m/s".
But as per Simulink Model (Fig.3) "Random Number Block" is used for generating "Zero Mean Gaussian White Noise". The Random Number block used for this requires 4 inputs i.e. "Mean", "Variance", "Initial SEED" and "Sample Time".
The "Mean" is taken as "0" , "Sample Time" is taken as Simulation fixed step time i.e. 0.001.
Now I have no clue what value will be used for "Variance" and "SEED"? for this white noise.
Will the "variance" be Square of "Sigma" from Table 3? or some thing else?
Dear Mr Pankaj Sharma
I would suggest not to use that block anymore as it's obsolete nowadays. You can better follow another approach. Which is mentioned below.
Step 1 - Insert a MATLAB Function block into the Simulink workspace.
Step 2 - Design a White Gaussian Noise following the algorithm
- signal=10 * sin(t) % your signal
- ecart_type=abs(signal * .05) % variable std 5%
- noise= mean + ecart_type .* randn(1, numel(signal));
into that block
Step 3 - Use that block as your Noise Generator.
The block 'AWGN' can not be found in the Matlab version after I think 2015 as it was removed for such complication, which you are facing presently. I hope I could help you. Good Luck!
Dear Mr Ritankar Sahu thanks for your valuable suggestion.
The data which I know for generation of white noise for my case is "Variance =0.000064" and "Mean=0".
As per your suggestion if I am not wrong sin wave with is used for generation of White signal
function W = trial(u1)
%u1 is product of v, alpha and Zr and signal to which white noise is to be added%
mean=0;
variance =0.000064 %value of variance to be used%;
W = sqrt(variance).*randn(1,size(u1,2))+mean; %Gaussian white noise W%
end
The white noise so generated is added in eq. 15 as shown in fig. in previous post.
I am attaching Simulink Model as per equation 15 prepared by me after your suggestion.
Am I doing it in right way now?
Hello Mr Pankaj Sharma
According to your explanation, I think you are doing good. The model looks great either. Did the simulation work as per your requirement? If yes then it's ok, If not, then there are other ways too. Please let me know, good luck!
Dear Mr. Ritankar Sahu
The profile obtained as per my function block varies a lot from expected one. I have attached the image of Expected Profile and The Obtained One.
Hello Mr Sharma
It looks like having 10% of the expected value. What happens when we have fractional gain. You can simply multiply the value by 10 and it will be nearby. For the rest, now I need to see your whole project. If you don't have a problem, you can send me the Simulink Model file and the *.m file. I will run and I will check what's wrong. I believe multiplication by 10 can solve the problem. Good Luck!
Hello Mr. Ritankar
I am hereby adding my Simulink Model as well as the Paper Which I am referring to prepare this model.
The Page no. 155 (Fig.3) & (Fig.4), (Fig.5), Page no. 154, Table:3 are the base for Simulink Model.
Good Afternoon Mr Pankaj
I will look through your model and code for sure during the day. Please give me few hours. I will get back to you with the results. Have a nice day!
Hello Mr. Ritankar Sahu
Have you got any idea where I am making mistake in generating random road profile in my simulink mode.
Hello Mr. Sharma,
I am sorry I couldn't look through your problem last week as I was out of the country. Now I am back and within this weekend I will definitely find a solution. Don't worry and please don't mind. Have a nice day!!
- Badges
- Science topic