Control
Introduction
The Control Team at CMT - Clean Mobility & Thermofluids focuses on the development of advanced methodologies for the control, optimization, monitoring, and management of propulsion systems and energy conversion processes. Its research activities cover a wide range of topics, including powertrain control, monitoring and diagnosis, energy and thermal management, as well as driving and fleet management. By combining physical modeling, optimization techniques, and data-driven approaches, the group aims to design innovative solutions that enhance performance, reduce environmental impact, and ensure the reliability of current and future mobility systems.
Powertrain control
Main Features
- Closed-loop dual-fuel control for RCCI and DDF modes
- In-cylinder pressure feedback for real-time combustion adjustment
- Control-oriented model coupling injection timing, fuel ratio, and IMEP
Control Strategy and Objectives
- Optimize combustion stability, efficiency, and emissions
- Adapt injection and fuel parameters across operating conditions
Main Features
- Stochastic predictive control for hybrid energy management
- Data-driven estimation of power demand from past driving data
Control Strategy and Objectives
- Achieve near-optimal fuel economy without cycle preview
- Adaptively manage power split between engine and battery
- Ensure robust performance under varying traffic conditions
Main Features
- Sensor-based control of SCR + ASC systems
- Integration of NOx and NH3 sensors with adaptive feedback
- Real-time correction using cross-sensitivity factor (KCS)
Control Strategy and Objectives
- Optimize urea dosing for high NOx conversion and low NH3 slip
- Compensate sensor errors via adaptive KCS estimation
- Achieve robust and cost-effective emission control
Monitoring & Diagnosis
Main Methods
- Acoustic-based knock detection using resonance modelling
- Resonance indicators from HRR and pressure signals
- Analysis of resonance energy and auto-ignition behavior
Diagnostic Capabilities
- Classify knock intensity from mild to severe
- Distinguish normal combustion from knock in frequency domain
- Enable real-time, non-intrusive combustion monitoring
Main Methods
- ML-based combustion diagnosis using knock sensor signals
- STFT and SVD for vibration signal processing
- ANN trained to predict key combustion metrics
Diagnostic Capabilities
- Estimate combustion phasing without pressure sensors
- Correlate vibration features with combustion anomalies
- Enable real-time, cost-effective combustion monitoring
Main Methods
- Unsupervised knock detection using vibration analysis
- Time-frequency features extracted via STFT and SVD
- OC-SVM for combustion cycle classification
Diagnostic Capabilities
- Detect knock using low-cost vibration sensors
- Accurancy comparable to in-cylinder pressure methods
- Enable adaptive, sensorless knock intensity detection
Main Methods
- Model-based diagnosis of SCR+ASC using sensor fusion
- NOx and NH3 signals combined for emission estimation
- Luenberger-type observer detects injector and catalyst faults
Diagnostic Capabilities
- Identify injector failure and catalyst ageing in real driving
- Infer fault level from NH3 and NOx deviations
- Support adaptive control to ensure emission compliance
Energy & Thermal Management
Main Features
- Hybrid NLMPC-DP control framework for PHEVs
- Real-time optimization of fuel use and NOx emissions
- Integrated powertrain and aftertreatment predictive control
Applications and Benefits
- Improve energy efficiency and emission compliance
- Dynamically balance power split between ICE and EM
- Demonstrate benefits in routes with zero-emission zones
Main Features
- Predictive thermal management for long EV trips
- Anticipatory control of battery and cabin temperature
- Unified framework for thermal and comfort control
Aplications and Benefits
- Reduce charging time via battery preconditioning
- Prevent derating and improve passenger comfort
Main Features
- Physics-based model for BEV battery thermal control
- Experimental setup with active water-based cooling and heating
- DP optimization for energy-efficient temperature regulation
Aplications and Benefits
- Maintain battery temperature between 20-35 ºC
- Reduce thermal control energy in all climates
- Enhance system stability and battery lifetime
Driving & Fleet Management
Main Features
- Cooperative control combining long-horizon DP and short-horizon MPC
- V2V and V2I connectivity to anticipate traffic and signals
- Real-time coordination of speed and lane changes
Aplications and Impact
- Reduce fleet fuel use and emissions without delaying travel
- Improve safety and traffic flow through cooperation
Main Features
- DP-based optimal speed advisories using traffic and engine data
- V2X/LiDar integration to smooth driving and maintain safe gaps
Aplications and Impact
- Reduce fuel consumption and NOx emissions
- Enable smooth, robust eco-driving trajectories
- Basis for on-board eco-cruise and corridor control
Main Features
- Real-time MPC with air-quality and QP optimization models
- Adjusts traffic density using urban data
- Enforces capacity limits and demand conservation
Aplications and Impact
- Mitigate PM10 hotspots and ensure regulatory compliance
- Scalable and computationally efficient for large cities
- Support sustainable mobility and environmental policies
Facilities
Dynamic Vehicle Test Bench
Power Plant Test Cells
