Amey Y. Karnik

Humidity and Pressure Regulation in a PEM Fuel Cell Using a Gain-Scheduled Static Feedback Controller

In this paper, the pressure difference between the anode and cathode compartments of a polymer electrolyte membrane (PEM) fuel cell stack is regulated along with the anode and cathode humidities using an anode recirculation system. The pressure regulation requirement stems from membrane safety considerations. The regulation of average humidities in the two compartments is a necessary (although not a sufficient) requirement for stack water management. Two actuators in the anode recirculation system are considered, namely the dry hydrogen flow and the anode back pressure valve. These actuators are adjusted using a static output feedback controller that relies on pressure and humidity measurements on the anode side of the fuel cell stack. As the water mass dynamics and the characteristics of the water transport through the PEM are significantly different between subsaturated conditions (water is present only in vapor phase) and saturated conditions (liquid water along with water vapor), we show that the performance of the static output feedback controller with a fixed set of gains for subsaturated condition deteriorates significantly under a saturated condition. A gain-scheduled controller is therefore developed to compensate for a water-vapor saturated cathode condition. Analysis and simulation provide insights on some of the design and implementation issues for the gain-scheduled output feedback system.

Dual-UEGO active catalyst control for emissions reduction: design and experimental validation

This paper investigates active control of an aftertreatment system for a conventional spark ignition engine equipped with one or two three-way catalysts and two oxygen sensors. The control objective is to maximize the simultaneous conversion efficiencies of oxides of nitrogen and unburned hydrocarbons. Linear exhaust gas oxygen (EGO) sensors are used to measure air-fuel (A/F) ratio upstream and downstream of each catalyst. A series controller configuration is adopted. The upstream controller provides relatively rapid response to disturbances on the basis of measured feedgas A/F ratio, while the downstream controller uses the feedgas and post-catalyst A/F ratio measurements to compensate for the bias corrupting the feedgas A/F ratio measurement. The performance and robustness of the proposed control system in the face of noise and model uncertainty are first evaluated through extensive simulations. The control strategy is then experimentally verified in a dynamometer test cell and its performance compared with an existing proprietary controller that is based on the more common switching-type A/F ratio sensors.

Control analysis of an ejector based fuel cell anode recirculation system

An ejector based recirculation system is incorporated on the anode side of the fuel cell to address anode flooding. The objective of the recirculation system is to regulate anode humidity with recirculating flow, while safeguarding the membrane and supporting the electric load. These objectives can be achieved by a control system that uses two actuators, namely the fuel supplied from the fuel source and a back pressure valve placed in the recirculation path. For these actuators, the control analysis presented in this paper identifies anode pressure, along with either return manifold pressure or recirculation flow rate, as variables to be tracked in order to facilitate meeting of the control objectives, while assuring internal stability. Feedback regulation of these variables with a minimum number of pressure sensors is addressed by identifying appropriate sensor locations. The sensor selection and corresponding observer design for the system are treated from the points of view of the observability requirements and robustness to sensor noise. The performance of various sensor combinations shows that measurement of anode pressure is adequate for a full order observer based state feedback controller

Internal model control design for linear parameter varying systems

In this paper, we address the design of internal model controllers (IMC) for linear parameter varying (LPV) systems. The design method presented here integrates the IMC and low-pass filter design problems. To this purpose, we formulate the design problem as a linear matrix inequality (LMI) optimization problem by imposing a constraint on the structure of the Lyapunov matrices involved to guarantee robust stability and performance of the closed-loop system. The LMI-based design procedure eventually provides the parameter-dependent state-space matrices corresponding to the IMC controller. Finally, simulation results obtained from a nonlinear system modeled in LPV framework demonstrate the effectiveness of the proposed design method, where we illustrate the improvements achieved by using the LMI-based IMC design over baseline IMC controllers amended with low-pass filters of different bandwidths.

Effect of Exhaust Gas Temperature Limits on the Peak Power Performance of a Turbocharged Gasoline Engine

Peak power of an engine is typically constrained by the maximum obtainable airflow. This constraint could arise directly from the airflow limitation imposed by the throttle restriction (typical for a naturally aspirated engine), or indirectly from other factors, such as various temperature limits for component protection. In this work, we evaluate the airflow limit for a turbocharged gasoline engine as dictated by the constraints on the turbine inlet temperature. Increasing the limit on the turbine inlet temperature requires the exhaust manifolds and turbine to be made out of more expensive materials that withstand higher temperatures. This expense is justifiable if operating with higher turbine inlet temperature allows noticeably higher power output, and not merely increases the allowable airflow. Experimental data show that under some conditions the increase in airflow does not increase the peak power. The effects of increasing airflow on the peak power and turbine inlet temperatures are systematically analyzed through individual accounting for the different losses affecting the engine torque. The breakdown analysis presented in this work indicates combustion phasing as a major contributing factor to whether increasing the flange temperature limit would increase the peak power.

IMC based wastegate control using a first order model for turbocharged gasoline engine

The basis of constructing an Internal Model Controller (IMC) are the plant model and its inversion. The construction of an appropriate invertible plant model is one of the challenges in implementation of an IMC on a nonlinear system. In this work we study the control of a turbocharged gasoline engine using a wastegate actuator. The turbocharged engine is typically represented using a fourth or higher order model. A first order representation of this system (or single state) is obtained via singular perturbation. The resulting model is invertible, and is cast into the standard structure of IMC. We also include the throttle actuator into consideration during the reduced order model development. The IMC is implemented on a full order nonlinear model for evaluation and the performance is compared to a PI controller with similar feed-forward action.

Performance of a PEM Fuel Cell Water Management System Using Static Output Feedback

Water management through anode water removal and cathode air inlet humidification is investigated in this paper. A state feedback controller is designed to meet both the steady state and transient requirements for water balance and pressure regulation. Performance of the controlled system with full- state feedback is analyzed, in particular when water content in the stack is measured by humidity sensors. In an attempt to eliminate the need for a state observer and to reduce the measurements required, static-output feedback is then explored. It is shown that comparable performance can be achieved under sub-saturated stack conditions with a static-output feedback using only anode side pressure and humidity measurements.

Synthesis of fault tolerant switching protocols for vehicle engine thermal management

Thermal management is very important to guarantee ideal performance of compact vehicle engines. One challenge in the vehicle engine thermal management is to control the engine temperature in a small interval while tolerating component failures and the uncertainties in complex environment and different operating conditions. We formulate this control problem as a temporal logic game for a switched affine system and solve it by synthesizing a switching protocol based on an abstraction. The existing algorithms for computing abstractions either cannot handle parametric uncertainties in the dynamics or can be computationally expensive. Besides, they usually do not deal with possible component failures. The main contribution of this work is to show: (i) how to compute an abstraction more efficiently under the assumption that the vector fields are multiaffine in constant uncertainties and affine in state variables, (ii) how to result in a graceful degradation in case of component failures.

Scavenging in a turbocharged gasoline engine

The phenomenon of air escaping the engine intake directly to the exhaust during valve overlap is commonly known as scavenging. This phenomenon is primarily observed at low-speed, high-load in engines with significant overlap between intake valve opening and exhaust valve closing. Evaluation on a turbocharged-gasoline engine shows increased low-speed torque when operating under scavenging conditions. This paper investigates the occurrence of scavenging and analyzes its consequences. A methodology is presented to infer the amount of scavenging using airflow and in-cylinder pressure measurements. Scavenging increases catalyst exotherm when operating with stoichiometric exhaust. A model is proposed to predict the additional exotherm.

Fuel cell thermal management: Modeling, specifications and correct-by-construction control synthesis

Thermal management is crucial for safe and efficient operation of fuel cells. The goal of this paper is to algorithmically synthesize a provably-correct controller for a fuel cell thermal management system. For this purpose, we start with developing a control-oriented model for the fuel cell thermal management system and list the associated requirements. Then, we identify some structural properties of the system dynamics that can be leveraged for making the abstraction-based synthesis algorithm computationally efficient. Finally, we synthesize a controller for this system and demonstrate the closed-loop system behavior via simulations.