Real-time Power and Control System (RPCS) Laboratory:
I have established Real-time Power and Control Laboratory within the school of EECS. The current and future activities of the lab are:
- Control of high-switching dc-dc converters
- Simulation of multicell battery models and control of vehicular battery packs.
- Integrated control of dc-dc converters acting as bi-directional charger for aged Vehicular Li-ion battery for converter efficiency and increasing the lifetime of the pack.
- Novel control methodologies for engine management systems and multi-sensory onboard calibration for particulate reduction and emission in IC petrol engines and diesel engines.
I am also collaborating on grid energy storage and renewable energy projects with the School of Engineering and Material Science (SEMS), and the Advanced Robotics Laboratory.
Physical systems are continuous but controllers are digital or discrete. Imagine we want to control an internal combustion engine of a vehicle. In the engine control unit (ECU) the data is read from all the sensors (via DAC) with a specific sampling time. Control algorithm must finish its execution within the sample time. Therefore, each “step” of the control program must be started exactly one sample time or step size apart, and have to be finished the computation of each step within the sample time, i.e. before the next step starts, to send the processed data (via ADC) to actuators. This is a real-time system.
Digital control system for an engine management system
Real-time control timing
dSPACE real-time simulator for emulation, test and optimization of a control/plant model developed in Matlab/Simulink
Real-time simulation is the de facto engineering process for the development, test and validation of control systems in a variety of applications including but not limited to control and simulation of power electronic circuits, motor control, engine control, vehicle control, and robotics. There are two important steps in real-time implementation, Hardware-in-the-loop (HIL) simulation, and rapid control prototyping (RCP).
RCP is the design of control system when you simulate a model of the controller connected to an actual physical system (or plant). The image above illustrates the development process of a control system using RCP (read more in dSPACE website).
In Hardware-in-the-loop testing, the process is the opposite where the control hardware (embedded implementation) connected to the simulated model of the physical plant. The image above shows the HIL test of a multi-cell battery pack using dSPACE system. Very useful material about real-time simulation, digital signal processing (DSP), HIL and RCP can be found in Mathworks website .
dSPACE Equipment of the laboratory
SCLALEXIO – a powerful Multi-core and versatile hardware-in-the-loop (HIL) simulator that provides highly flexible channels, and a specialized I/O hardware. Control Desk Next Generation software let real-time simulation of Matlab/Simulink models, e.g. dSPACE Electrical Power Systems Simulation Package allows the real-time simulation of electrical models developed in SimPowerSystems. Read more.
MicroLabBox – a small all-in-one development system for rapid control prototyping (RCP). It comes with over 100 channels of different I/O types and a combination of real-time processor and FPGA functionality. Close-loop and open-loop control algorithms can be developed in Matlab/Simulink and implemented on the freely programmable FPGA of the MicroLabBox to achieve the fastest possible control sampling rate. RTI E-Motor Control blockset is used in combination with FPGA programming blockset to achieve the fastest possible control sampling rate of 16MHz for motor control. Read more.
DS1104 R&D Controller Board – a single PCI board system for real-time control development and rapid control prototyping (RCP). Currently, two PCs in the lab have this board installed and can be used as a real-time hardware for smaller control development in power electronics, electrical machines, drives, and robotics. Read more.