WP2 Virtual sensors and control

WP Leader: Kai Zenger, Aalto University & Jari Böling, Åbo Akademi University
Research organizations: Aalto University, Åbo Akademi University & University of Vaasa
Companies: Wärtsilä Finland Oy, Meyer Turku Oy & Napa Oy
International collaboration partner: National Technical University of Athens – Laboratory of Marine Engineering, Greece

WP2 develops virtual sensors and algorithms generally aimed at applications in soft computing, sensor fusion and machine learning. Control algorithms for the RCCI combustion are developed. The work focuses on the following targets:

  • Study on the above methods and algorithms theoretically in complex applications having severe disturbances and uncertainties in the models;
  • Develop and apply virtual sensing in estimating emissions in marine engines and in the purification process of emissions (Selective Catalytic Reduction, SCR);
  • Using virtual sensing in correcting uncertain measurement results, e.g. in the flow of ammonia in an SCR;
  • Using virtual sensing in the NOx and GHG emissions in the combustion process of marine engines;
  • Developing sensor fusion methodology based on virtual sensors and data analytic algorithms to be applied in engine applications;
  • Developing control algorithms for low emission combustion systems (RCCI) including multiple injections in engine cylinders;
  • Using virtual sensing in the estimation of variables in hybrid power systems;
  • Using virtual sensing in ship route planning in the presence of large uncertainties in measurements.

WP2 Tasks & Task Leaders:

T.2.1 Estimation of urea dosage in an SCR system (Jari Böling, Åbo Akademi University);
T.2.2 Estimation of NOx concentration in different parts of the SCR (Jari Böling, Åbo Akademi University);
T.2.3 Virtual sensor by PCA algorithm (Kai Zenger, Aalto University);
T.2.4 Sensor fusion and machine learning algorithms (Kai Zenger, Aalto University);
T.2.5 Control of RCCI combustion (Xiaoguo Storm, University of Vaasa);
T.2.6 Virtual sensors and digital twins for robust and optimal energy efficiency and route planning (Jari Böling, Åbo Akademi University & Kai Zenger, Aalto University).

WP2 Solutions and innovations:

S1. New estimation algorithms of input variables in dynamic systems are developed;
S2. The results are used in estimation of NOx emissions and small flow rates correction of SCR system;
S3. Virtual sensing and the related algorithms in soft sensors and sensor fusion are developed;
S4. New optimal controllers for multi-injection in RCCI combustion are developed;
S5. New route planning algorithms in operation conditions with large uncertainties are developed.