Typhoon HIL Forum - Recent questions and answers in Helper

PSO optimization technique how to adapt this code properly for execution in the Typhoon HIL C-block.

Sat, 27 Sep 2025 13:29:01 +0000

Dear Typhoon HIL Team,

I am working on a research project titled “Photovoltaic MPPT Performance Adaptability to Partial Shading Resilience and Load Variations with PSO.” In MATLAB/Simulink, I have implemented and validated the PSO-based MPPT algorithm using a C-function block, and I now wish to port this into Typhoon HIL Control Center for hardware-in-the-loop testing and real-time verification.

My MATLAB code (attached below) takes PV voltage and current (Vpv, Ipv) as inputs and generates the duty cycle (D_out) as output. It uses persistent variables for swarm initialization, particle updates, personal/global best tracking, and smoothing of the duty cycle output.

I would like guidance on:

Code Structure: How to adapt my MATLAB C-function for Typhoon HIL’s C-block (syntax, supported functions, use of init() and step()).
Inputs/Outputs: Correct way to declare Vpv, Ipv as inputs and D_out as output, and how to connect the duty cycle output to PWM blocks.
Implementation Steps: The procedure to load and compile this code in the Schematic Editor, handle unsupported functions (e.g., rand, linspace), and properly manage persistent variables.
Verification: Best practices to confirm that the Typhoon C-block produces results consistent with MATLAB under partial shading and load variation tests.

My main goal is to compare MATLAB simulations with Typhoon HIL results under identical conditions for benchmarking MPPT performance. Any step-by-step instructions, references, or example templates would be greatly appreciated.

Thank you for your support.

Below, I have provided a detailed explanation of the MATLAB C-function block along with the points where I am seeking clarification:

1. Code Structure

In MATLAB, my MPPT logic is encapsulated in a C-function block with persistent variables for maintaining PSO states such as swarm positions, velocities, particle bests, and global best. Below is the complete implementation:

%% ------------------ PSO Parameters ------------------

persistent swarm iter stepCounter P_hist D_out_prev

% PSO constants

N = 30; % number of particles

maxIter = 300; % max iterations

c1 = 1.8; % cognitive factor

c2 = 1.8; % social factor

w_max = 0.9; % max inertia

w_min = 0.4; % min inertia

D_min = 0.05; % min duty cycle

D_max = 0.95; % max duty cycle

% Moving average for stability

avgWindow = 7;

updatePeriod = 1; % update every step

perturbMagnitude = 0.02; % small perturbation

%% ------------------ Initialization ------------------

if isempty(swarm)

D_sweep = linspace(D_min, D_max, N);

swarm.x = D_sweep; % particle positions (duty cycles)

swarm.v = zeros(1,N); % particle velocities

swarm.pBest = D_sweep; % personal bests

swarm.pBestVal = -inf(1,N); % personal best power

swarm.gBest = D_min; % global best duty cycle

swarm.gBestVal = -inf; % global best power

swarm.k = 1; % current particle index

iter = 1;

stepCounter = 0;

P_hist = zeros(1, avgWindow);

D_out_prev = swarm.gBest;

D_out = swarm.gBest;

return;

end

%% ------------------ Slow Update ------------------

stepCounter = stepCounter + 1;

if stepCounter < updatePeriod

D_out = D_out_prev;

return;

else

stepCounter = 0;

end

%% ------------------ Moving Average Power ------------------

P_hist = [P_hist(2:end), Vpv*Ipv];

P = mean(P_hist);

%% ------------------ Update Personal Best ------------------

if P > swarm.pBestVal(swarm.k)

swarm.pBest(swarm.k) = swarm.x(swarm.k);

swarm.pBestVal(swarm.k) = P;

end

%% ------------------ PSO Update ------------------

w = w_max - (w_max - w_min)*(iter/maxIter);

r1 = rand; r2 = rand;

swarm.v(swarm.k) = w*swarm.v(swarm.k) ...

+ c1*r1*(swarm.pBest(swarm.k)-swarm.x(swarm.k)) ...

+ c2*r2*(swarm.gBest - swarm.x(swarm.k));

swarm.x(swarm.k) = swarm.x(swarm.k) + swarm.v(swarm.k);

%% ------------------ Clamp Duty Cycle ------------------

swarm.x(swarm.k) = max(min(swarm.x(swarm.k), D_max), D_min);

%% ------------------ Move to Next Particle ------------------

swarm.k = swarm.k + 1;

if swarm.k > N

swarm.k = 1;

iter = iter + 1;

% ------------------ Update Global Best after all particles ------------------

[bestP, idx] = max(swarm.pBestVal);

swarm.gBest = swarm.pBest(idx);

swarm.gBestVal = bestP;

% ------------------ Small Random Perturbation ------------------

if mod(iter,20)==0

swarm.x = swarm.x + perturbMagnitude*(rand(1,N)-0.5);

swarm.x = max(min(swarm.x, D_max), D_min);

end

%% ------------------ Reset if maxIter ------------------

if iter > maxIter

D_sweep = linspace(D_min,D_max,N);

swarm.x = D_sweep;

swarm.v = zeros(1,N);

swarm.pBest = D_sweep;

swarm.pBestVal = -inf(1,N);

swarm.gBest = D_min;

swarm.gBestVal = -inf;

iter = 1;

end

%% ------------------ Smooth Duty Cycle Output ------------------

alpha = 0.2; % smoothing factor

D_out = alpha*swarm.gBest + (1-alpha)*D_out_prev;

D_out_prev = D_out;

end

function D_out = PSO_MPPT_PV_Stable(Vpv, Ipv)

Answered: Unexpected Model Behavior with Induction Motor Parameters

Thu, 24 Jul 2025 08:22:18 +0000

Hi Gordon,

It would be helpful if you could send your code for the MPC as it would allow as to better solve this issue. (The whole model would be helpful too).

In the meantime you could try using a different discretization method for your models equations because, depending on what method you use, they could introduce some stability issues (E.g. Euler forward could make a stable system unstable). I would suggest using trapezoidal integration (If it's fast enough for you) as most of the characteristics transfer over.

Sampling time (Ts) would also affect stability of your system in your case, so you could try adjusting it and see if it fixes your issue.

Also, You could get more familiar with the three-phase squirrel cage induction machine you're using here (If you haven't done so already) : Three phase squirrel cage Induction Machine.

Best regards,

Predrag.

Answered: Create a single-phase microgrid

Tue, 03 Jun 2025 11:36:19 +0000

Hi Emanuel,

As a starting point, you can check out microgrid examples within the Examples library. Some of them have also corresponding application notes.

Python code explanation can easily be obtained utilizing some of the publicly available AI models. You can also seek help from Typhoon HIL AI Chatbot Relphie embedded at our official website page in the right bottom corner of the screen.

The main idea behind Modbus API is to create python instance which is acting as a SCADA based Modbus client. You can make more than one instance of TCPModbusClient and for each of them, you can define different server (host) addresses. In the application, Modbus server can be a HIL device, any other device, or both of them.

Hope this helps.

Answered: DC microgrid

Thu, 29 May 2025 08:55:10 +0000

Hi Mohannad,

I would highly suggest (If you haven't already) check out the HIL for microgrids course.

Regarding your question, here are links to some examples that could help you. Specifically the Generic PV Plant, Generic battery and the Microgrid powered by a wind farm examples could be of good use to you. These are not necessarily low voltage DC microgrids, but you could tweak them to your liking and needs.

Best regards,

Predrag

Answered: IEC61850 MMS server: stop simulation not disconnecting/stopping iec61850 mms server

Mon, 12 May 2025 14:07:19 +0000

Hi Shubham,

Actually, the MMS Server is implemented in a way that it keeps the connection with the MMS client even after stopping the HIL simulation. If you want to forcibly disconnect from the client you can call again hil.load_model(file=cpd, vhil_device=False) function after stopping the simulation.

Best regards,
Dusan

Answered: Issue with PV Panels Generic not responding to Irradiation profile

Mon, 12 May 2025 13:23:21 +0000

Hi Leonardo,

If you can see that the SCADA input corresponding to irradiation is changing according to the desired readings from the text file, the issue is not related to component not responding to irradiation profile.

It is hard to guess exactly what is causing the problem in your model without seeing it. Here are some common mistakes:

1. You forgot to enable the PV Power Plant
2. The parameters of the grid (voltage and frequency) are not set according to nominal parameters of the PV power Plant.

The easiest way to check this is to open PV Power Plant Interface Subpanel in HIL SCADA and check the DER state and whether there is any alarm message:

Best regards,
Dusan

Symbolic Calculation in C funtion

Fri, 07 Feb 2025 08:58:22 +0000

Hello everyone
How to perform symbolic calculation using C-funtion in typhoon?

Answered: Solving a system of quadratic equation

Thu, 06 Feb 2025 16:02:27 +0000

Your link goes to the University website and therefore I can't see the content.

W.r.t. the system of equations, I would say it is possible to solve in C function, but you need to implement the code for the solver in C.

Regards,

Ricardo

Answered: Unable to Initialize Analog Signals VHIL

Thu, 05 Dec 2024 10:38:04 +0000

Greetings,

Usually, the model consists of SCADA Outputs and FMUs should be easily loaded to VHIL without any issues. If you want, feel free to attach the model here for evaluation.

Answered: What will I be helping with?

Mon, 04 Nov 2024 16:37:08 +0000

The questions in Helper department come from various sources/user groups. Questions from all the forum will be selectively re-categorized under Helper department by program mentors. This helps gain visibility and gives you opportunity to prove yourself as Helper. Here are some questions types generated from real life situations:

Users whose license doesn’t include official (subscription-based) support.
Out-of-scope – stated problem exceeds the terms of official support. The question was redirected to the forum. For example, question asked does not directly concern the usage of Typhoon HIL toolchain. Such problems are either fundamental engineering problems or coding problems.
3^rd party tools – questions on usage of tools that are not part of Typhoon HIL toolchain (Code Composer Studio, MATLAB/Simulink, etc.) get redirected to the forum.
Sponsored – questions asked by Typhoon HIL and its hiring partners as part of public talent acquisition campaigns.