Search

In this paper, predictive control of a spark ignition engine equipped with an electronic throttle and a variable cam timing actuator is considered. The objective is to adjust the throttle angle and the engine cam timing in order to reduce the exhaust gas emissions while maintaining fast and monotonic transient engine torque response. To this end, using a mean-value nonlinear model for the air intake system of the engine, a receding horizon control scheme based on a flexible control Lyapunov function is developed. The proposed control scheme can adjust the throttle plate angle and the cam timing to provide a fast and monotonic air flow response into the engine cylinders, while explicitly taking into account constraints. Compared to a previous parameter governor controller, the simulation results show that the proposed solution provides better performance in terms of settling time. Moreover, the computational complexity of the algorithm is compatible with current electronic control units, despite using a full nonlinear model.

This paper presents the hardware configuration, modeling, and control design for a lighter-than-air wind power system being pioneered by Altaeros Energies. This unique design features a horizontal-axis turbine that is elevated to high altitudes via a buoyant shroud, which is tethered to a ground-based platform. Because the system is based on proven aerostat technology and is designed to remain substantially stationary, it circumvents many of the controls challenges faced by so-called aerodynamic (e.g. kite-based) wind energy systems. However, the need to generate energy introduces pointing, efficiency, and autonomy requirements that are not faced by conventional aerostats, thereby requiring a careful model-based control design. In this paper, we first provide a detailed description of the system and controls hardware for the Altaeros 2.4 kW proof-of-concept prototype. We provide a detailed 3-dimensional dynamic model for the Altaeros system, which is used in the design of a fully autonomous control system. This paper details the control system design and shows simulation results that substantiate the system's performance.

This paper further develops an approach for spacecraft relative motion control based on the application of linear quadratic Model Predictive Control (MPC) with dynamically reconfigurable constraints. Previous results for maneuvers confined to the orbital plane are extended to three dimensional maneuvers with three dimensional Line-of-Sight (LoS) constraint handling. The MPC controller is designed to prescribe Δv impulsive velocity changes rather than piecewise constant thrust profiles. The ability to transition between MPC guidance in the spacecraft rendezvous phase and MPC guidance in the spacecraft docking phase, with requirements, constraints, and sampling rates specific to each phase, is demonstrated. Bandwidth constraints of the spacecraft attitude control system and exhaust plume direction constraints are also addressed. The MPC controller is validated on the full nonlinear model of spacecraft orbital motion and augmented with an Extended Kalman Filter (EKF) to estimate spacecraft states based on relative angles and relative range measurements.

This paper describes the process of creating a computationally-inexpensive yet relatively accurate atmospheric environment model for use in stochastic optimal control problems for glider flight management. In such problems, estimates of transition probabilities between flight in updrafts, downdrafts and thermals of varying strength are needed. This work proposes an atmospheric environment model that predicts updraft and downdraft strengths in a given region and, when combined with existing glider flight data, estimates thermal locations and strengths. The resultant predictions can be utilized to compute the desired transition probabilities. A simple approach currently employed in flight simulator games is adapted for updraft and downdraft modeling. The method is empirical and requires the computation of a linear factor. Interestingly, when validated against actual flight data, this technique is 92.4 percent accurate on average. This paper also shows that the location and intensity of thermals can be deduced from flight data by utilizing updraft and downdraft predictions. The work then illustrates the modeling process for a sample topographical area, and utilizes the model to solve a stochastic drift counteraction optimal control problem where control policies that maximize glider flight range are generated.

In applications involving time-varying systems, the state dynamics matrix is often not known in advance. To address this problem, this paper investigates the effectiveness of a forward-in-time Riccati-based control law. This approach is motivated by the fact that the optimal state estimator is based on a forward-in-time Riccati-based solution that does not require advance knowledge of the system dynamics. In this paper we show that a forward-in-time Riccati-based control law stabilizes the system if the dynamics of the quasi-dual system are asymptotically stable. This property holds if the closed-loop dynamics are symmetric, and, for some plants, is achieved by dynamics with sufficiently fast time variation. In addition, using a separation principal type result, we guarantee closed-loop stability in the case of output feedback.

This paper considers an on-line approach to reducing engine fuel consumption in a series hybrid electric vehicle operated under power smoothing strategy, in which the generator power is slowly varying. An optimization algorithm adjusts the engine speed set-point (and other engine actuator settings as appropriate) on-line, as the vehicle is driven. The actual generator power is estimated, and the battery power output is adjusted to ensure that the wheel power and the drivability are unaffected. With this approach, operating points on the optimal operating line (OOL) can be learned during vehicle driving thus compensating for the effects of the engine aging and variability while reducing upfront calibration time and effort. In the paper, the control approach is defined and an optimization algorithm is examined. Simulation results are reported.

This paper develops further the network reference governor, which is a predictive algorithm for modifying commands sent to the remote system to satisfy state and control constraints. Due to the network communication, the governor must account for a delay that can be time-varying and unknown. The paper summarizes the results on network reference governor theory, and demonstrate its operation on a case study of a attitude control of a spacecraft with a very flexible appendage, where the commands are transmitted remotely over a network to the spacecraft, and hence delayed by a bounded, unknown delay. In this case study, the reference governor ensures that the elastic deflections of the appendage and the control signal satisfy the imposed limits while the spacecraft performs a reorientation maneuver. The paper then presents the novel theoretical construction of a less conservative network reference governor for the case when the delay is long but only slowly time-varying, with known bounds on the rate of change. A spacecraft relative motion control example with constraints on thrust and Line Of Sight (LoS) cone positioning is considered to illustrate these theoretical developments.

This paper considers a range maximization problem for a direct methanol fuel cell powered and battery powered Mini Air Vehicle (MAV) operating in an uncertain environment. The control policy is developed to prescribe the power split between the fuel cell and the battery. This problem fits naturally into the Stochastic Drift Counteraction Optimal Control (SDCOC) framework. The approach accounts for the statistics of MAV velocity and climb angle transitions but does not require their exact a priori knowledge. The control policy is generated based on the hybrid powerplant model that accounts for the dependence of fuel cell efficiency on the orientation.