Random road profile to quarter car model in simulink

Hellow everyone, I am currently doing a project which includes simulating a quarter car model on a random road profile based on ISO 8608. I have coded the road profile in matlab and wanted to input the profile which is a 1 x 2250 data as an input for my quarter car model, but the simulink model includes a derivative and integration blocks so I guess it is why I'm always receiving an error message everytime it runs. Is there anyway I could do this, what I want to do basically is replace the usual sine wave or step function with the data I have generated with matlab.

Thank you in advance, below are my simulink model and matlab code

% Parameters

k = 4; % Values for ISO Road Roughness Classification, from 3 to 9

V = 40; % Car Velocity (km/h)

L = 250; % Length of road profile (m)

N = L / (V / 3.6) * 100; % Number of data points

B = L / N; % Sampling interval (m)

dn = 1 / L; % Frequency band

n0 = 0.1; % Spatial frequency (cycles/m)

% Spatial frequency band

n = dn : dn : N * dn;

% Random phase angle

phi = 2 * pi * rand(size(n));

% Amplitude for Road Class A-B

Amp1 = sqrt(dn) * (2^k) * (1e-3) * (n0 ./ n);

% Abscissa variable from 0 to L in distance and time domain

x = 0 : B : L - B;

t = x ./ (V / 3.6);

% Initialize road profile

hx = zeros(size(x));

% Generate road profile data

for i = 1:length(x)

hx(i) = sum(Amp1 .* cos((2 * pi * n * x(i)) + phi));

end

% Compute Power Spectral Density (PSD) using external file

[q , C] = psd_1D(hx, B, 'x');

lambda = (2*pi) ./ q; % Wavelengths

% Frequency axis

f = q / (2*pi); % spatial frequency

PSD = 2 * pi * C; % power spectrum

log_f = log10(f); % Converts to Logarithmic

log_psd = log10(PSD); % Converts to logarithmic

% Plot road profile

% Plot the road length vs amplitude of the road profile

subplot(2,2,1)

plot(x, hx);

xlabel('Distance (m)');

ylabel('Amplitude (m)');

grid on

% Plot the PSD vs Spatial frequency of the road profile

subplot(2,2,2)

plot(log_f, log_psd);

xlabel('Frequency (cycles/m)');

ylabel('PSD (m^3)');

grid on

% Plot the time vs amplitude of the road profile

subplot(2,2,3)

plot(t, hx);

xlabel('Time (s)');

ylabel('Amplitude (m)');

grid on

Answer 1

[Maneet Kaur Bagga]

Hi,

As per my understanding the error you are encountering is due to the discrete data that is input directly into "derivative" and "integrator" blocks which expect continuously sampled input. To resolve the encountered steps you can refer to the following steps:

• Load the road profile data into Simulink using a "From Workspace" block.

% Save road profile as time series data for Simulink
t = (0:length(hx)-1) * B; % Adjust time based on distance and sampling
road_profile = [t' hx']; % 2-column matrix [time, road profile]

• In your Simulink model, replace the sine wave/step function block with a "From Workspace" block. Set the "Data" parameter of the block to "road_profile" which is the time based representation of the road profile. Set the "Sample Time" to "-1" so that Simulink auto-detects the sample time from the data.
• For blocks like "integrator" and "derivative" that expect continuous data, please check that your data is appropriately sampled. If your road profile "hx" is too coarse, resample it to finer intervals by interpolating the data. Refer to the following code snippet as example.

% Resample road profile to match solver needs
time_fine = linspace(min(t), max(t), finer_sample_points);
hx_fine = interp1(t, hx, time_fine, 'linear'); % Linear interpolation

• Use a variable-step solver like "ode45" or "ode15" and set the step size equal to the sampling rate of your road profile data.

Hope this helps!