A Benchmark Simulator for Advanced Control of Ethanol Steam Reforming

This repository contains the benchmark simulator for a staged-separation membrane reactor (SSMR) designed for pure hydrogen production. The simulator implements a nonlinear dynamic model of a two-stage process: ethanol steam reforming (ESR) and hydrogen separation. It is intended as a benchmark for testing control and estimation strategies.

Installation

1. Download the ZIP folder from the GitHub repository

2. Extract the contents to a folder of your choice

3. Launch MATLAB R2020a or later (recommended). Additional toolboxes are not necessary

4. Set the extracted folder as the current folder in MATLAB

5. Open the main simulation script: SSMR_simulator/SSMR_simulation.m

Instructions

1. Specify simulation conditions in SSMR_simulator/SSMR_simulation.m

   • Select the normal operating conditions: Mode 1, 2 or 3 (see Table 2 in the paper)
   • Select the disturbance scenario (see subsection 4.2 in the paper)
   • Specify the time at which the disturbance is to be applied (if any)
   • Choose the initial conditions (see more details in the supplementary material)
   • Specify the overall simulation time
   • Select the set-point profile (see Fig. 4 in the paper)
   • Choose the simulation type (open loop or with control)

2. Run SSMR_simulator/SSMR_simulation.m

3. The simulation results will be displayed in plots that appear on the screen

4. To modify or implement custom controllers, open control.m

5. In control.m, create a new case inside the switch environment

6. Design your own control law that outputs the changes to the input variable

7. In SSMR_simulation.m, change the value of control_law according to the number assigned to your case

8. Customize the benchmark to include multiple input variables if desired

Run case studies

1. PID controller to address control challenge 1

   • Mode 1
   • Disturbance = 0
   • Time for disturbance (min) = any (since it does not apply here)
   • Initial conditions = 0
   • Simulation time (min) = 30
   • Set-point profile = 1
   • Simulation type = 1
   • Control_law = 0

2. PID controller to address control challenge 3

   • Mode 2
   • Disturbance = 1.1 or 2.2
   • Time for disturbance (min) = 5
   • Initial conditions = 1
   • Simulation time (min) = 10
   • Set-point profile = 0
   • Simulation type = 1
   • Control_law = 0

License

This project is covered under the MIT License.

Cite

If you found this repository useful, please cite: https://doi.org/10.1016/j.renene.2025.124743