Real-time Simulation Aircraft Aileron Actuator

This blog is Part 4 of a blog series that demonstrates how the Modelon Hydraulics Library can be used for the model-based development of an Aircraft Hydraulic Actuation System (AHAS). In this series, basic steps of the actuation system development workflow will be covered.

Previously, Part 2 and Part 3 focused on the modeling and simulation of electro-hydraulic actuator with datasheet-based parametrization. Part 4 – below – is dedicated to simulating the simplified aileron model in a real-time target.

The Importance of Real-Time Simulation

During the development of a Fly-by-Wire actuator and its control electronics, the trimming of the actuator control laws is of particular importance. While most can be performed with simulation on a standard computer, one relevant step is to include the control laws on a real-time target that controls the physical model (plant model) inputs. These two models are usually not coming from the same software.

This plant model shall react in real-time – i.e. the computation of every simulated time step takes less time than the simulated time itself – to avoid that the model would be delayed compared with real-time orders (this would be a task overrun). This real-time integration enables validating the actuator control laws with more accuracy.

In this blog post, we will illustrate how the simple actuator model developed in the blog post 2 is integrated in a Speedgoat target and runs real-time.

Minor Model Adaptations

The changes made on the model are the following:

  • Isolation of the plant model (the simple control logic would be in the real-time target implemented in Simulink®).
  • Addition of a few outputs monitoring the pressures, flow rate and force of the plant model – to have an easier access to and plotting of these.

Figure 1 below displays the final version of the model.

Figure 1 A320 aileron actuator-like plant model – ready for export

Integration on the Real-Time Target

The integration in the Real-Time target consists of a few steps:

  • Coding in Simulink® the control/command blocks of the actuators.
  • Including the Modelica model made from Modelon Hydraulics, exported as an FMU, using FMI Toolbox developed by Modelon.
  • Adding Speedgoat specific blocks to monitor the variables of interest.
  • Build the real time executable using the Simulink Real-Time™ target.
Figure 2 actuator plant model integration in Simulink for import on Speedgoat real-time target

In this application, the plant model has been exported with its own solver (co-simulation) so that both the control and the plant model have separate solvers and the models communicate at each time step.

Real-Time Simulation of the Integrated Systems

Once the model is integrated and imported on the real-time target, it is run satisfying hard real-time requirements. The main variables set as output previously are plotted here:

  • Top left: actuator command and response. We observe an initial position of the actuator on-purpose different from the command leading to a fast transient at the beginning of the simulation. After reaching the setpoint, the actuator reacts in a satisfactory dynamic behavior, slightly influenced by the antagonist loads.
  • Bottom left: the flow exiting the ideal pressure source is plotted. The dynamic observed is a redressed sine similar to the actuator position response. The frequency is doubled as the flow is provided in both positive and negative speeds of the actuator.
  • Top right: the pressure in both chambers are observed. These are oscillating around an average point, in phase with the force oscillation shown at the bottom right.
Figure 3 plotting of main variables of interests in the Speedgoat GUI

Conclusion

In this short blog post, we have illustrated the capabilities of exporting a Modelon Hydraulics Library model and integrated into Simulink® for running the control and plant models in co-simulation on a real-time target.

This demonstrator also illustrates that the same model can be used for different use cases – here, standard simulation and real-time simulation on a HiL target.

Stay tuned to find out the remaining interesting parts of the series covering full system performance analysis and component detailed design.