Control Engineering Practice

Benchmark Control Applications

Control engineering continues to evolve in a wide range of theoretical and applied directions. The recent resurgence of interest in machine learning algorithms and their intersection with control engineering has led to an explosion of algorithms and applications. This has made it difficult to benchmark and compare algorithms. Moreover, new control algorithms developed by researchers are often tested on small and illustrative but simplified numerical application examples, limiting their practical relevance for practicing control engineers and making comparisons with state-of-the-art methods difficult. Indeed, representative benchmarks and models are paramount to design and evaluate new model- and data-based controllers and to optimize them before porting them to the practical application. Finally, the systems and control community rarely shares code, making reproduction of algorithms a time consuming task.

The objective of this Special Issue is to collect a set of challenging benchmark control applications that are high-fidelity enough to be relevant for practical/industrial applications and are suitable for the control research community. These benchmarks will include reference control/system identification methods for comparative analysis.

Potential domains of interest include, but are not limited to:

• Industrial Processes
• Aviation and Space
• Automotive
• Power and Energy systems
• Mechatronic Systems

Guest editors:

Laurent Burlion (Executive Guest Editor), Aerospace control, Rutgers University, USA (contact: [email protected])

Lars Eriksson, Automotive control, Linköping University, Sweden (contact: [email protected])

Marco Forgione, Identification/Estimation, Dalle Molle Institute for Artificial Intelligence, Switzerland (contact: [email protected])

Bhushan Golupani, Industrial process control/data analytics /machine learning, the University of British Columbia, Canada (contact: [email protected])

Peter Fogh Odgaard, Power systems, Goldwind Energy, Denmark (contact: [email protected])

Maarten Schoukens, Identification/Machine learning modeling, Eindhoven University of Technology, Netherlands (contact: [email protected])

Special issue information:

We aim for papers in this Special Issue to provide benchmarks in various formats:

• Open-Source Simulators: Fully accessible simulators (ideally in MATLAB or Python) that can be downloaded and used to implement and test control algorithms. The underlying system dynamics may be openly available and known or included as executables and thereby hidden with unknown structure and parameters. In the latter case users can for example use the executable to generate data for system identification or other training methods in the control synthesis and evaluation.

• Model Equations with Validated Data: The dynamic equations are provided along with numerical parameter values, enabling readers to construct the benchmark in their preferred language. A reference implementation may be provided to validate that the users implemented the model correctly.

• Input/Output Trajectories: The simulator or experimental setup is not available but input/output data has been collected and can be used for system identification or machine learning modeling. This data can be divided into training and test datasets.

• Remotely Accessible Simulators or Real-Time Experiments: Remote access to realtime experimental setups or complex numerical simulators under specified conditions.

Researchers can implement and submit their controllers, and receive the resulting data. Again, the system dynamics may be known or unknown, but input-output data will be available after a submitted and executed test in the latter case.

For each benchmark, it is desirable to include:

• A nominal solution: A baseline, legacy, state-of-the-art control (or system identification) method for comparison purposes.

• Clearly defined tasks and performance criteria: Specific metrics to evaluate the performance of different algorithms for a given task. This includes a description of the known challenges that are present in the task at hand for the considered system.

• Result interpretation: Post-processing tools in the form of additional functions or scripts can be provided to facilitate analysis and visualization of simulation or experimental results. Furthermore, additional guidelines in terms of the qualitative description of the results (e.g. computing platform and time), can be provided to be able to interpret the results beyond the provided performance criteria.

• Standardized data and simulation environments: Provide the data and system simulators in a well-documented way, following good practices and standardized formats, allowing for easy accessibility of the benchmark to the users.

By providing these comprehensive benchmark applications, we hope to stimulate openness and innovation in control research that can promote the development of more practical and effective control algorithms, that practicing engineers can use as inspiration and adopt in their work flow.

Manuscript submission information:

Manuscripts should be submitted via the Control Engineering Practice online submission system (https://www.editorialmanager.com/conengprac/default.aspx) by selecting the Article Type of “VSI: Benchmark Control Applications”. All submitted manuscripts will be screened by the editorial office and peer reviewed according to the usual standards of this journal, and will be evaluated on the basis of originality, quality, and relevance to this Special Issue. Please also note the IFAC publication policy: Papers submitted to IFAC journals with prior publication in any copyrighted conference proceedings must be substantially different from the conference publication. Authors should indicate in the cover letter in detail how the journal paper differs from the relevant conference paper or papers. In particular, the additional original contribution in the journal paper has to be pointed out explicitly. In the journal paper, the conference paper has to be cited and discussed as any other paper in the list of references.

Control Engineering Practice publishes papers providing application-related information, stressing the relevance of the work in a practical industrial/applications context, with solid industrial examples rather than hypothetical ones. In this light, simulation models for the benchmarks in this special issue must be representative for and validated at the real plant under consideration. The interested authors are encouraged to read other CEP papers in the similar field to learn more about CEP’s standards and relevance to your work.

Important Dates

• Submission deadline: January 31, 2026

• Acceptance deadline: May 31, 2026

Keywords:

benchmark; high-fidelity; model-based control; data-based control; system identification

Check out the FAQs on special issues.

Learn more about the benefits of publishing in a special issue.

Interested in becoming a guest editor? Discover the benefits of guest editing a special issue and the valuable contribution that you can make to your field.