Eelco Scholte

Optimizing the Software Architecture for Extensibility in Hard Real-Time Distributed Systems

We consider a set of control tasks that must be executed on distributed platforms so that end-to-end latencies are within deadlines. We investigate how to allocate tasks to nodes, pack signals to messages, allocate messages to buses, and assign priorities to tasks and messages, so that the design is extensible and robust with respect to changes in task requirements. We adopt a notion of extensibility metric that measures how much the execution times of tasks can be increased without violating end-to-end deadlines. We optimize the task and message design with respect to this metric by adopting a mathematical programming front-end followed by postprocessing heuristics. The proposed algorithm as applied to industrial strength test cases shows its effectiveness in optimizing extensibility and a marked improvement in running time with respect to an approach based on randomized optimization.

Optimizing Extensibility in Hard Real-Time Distributed Systems

We consider a set of control tasks that must be executed on distributed platforms so that end-to-end latencies are within deadlines. We investigate how to allocate tasks to nodes, pack signals to messages, allocate messages to buses, and assign priorities to tasks and messages, so that the design is robust with respect to changes in task requirements. The notion of extensibility is used to measure robustness. The extensibility metric measures how much the execution times of tasks can be increased without violating end-to-end deadlines. We optimize this metric by adopting a mathematical programming front-end followed by post-processing heuristics. The proposed algorithm as applied to industrial strength test cases shows its effectiveness in optimizing extensibility and a marked improvement in running time with respect to an approach based on randomized optimization.

Online nonlinear guaranteed estimation with application to a high performance aircraft

The most commonly used filters (or estimators) for online control customization are based upon stochastic assumptions on the noise. In this paper a set-membership filter is extended to be used with nonlinear systems for which the Jacobian and Hessian are continuous over the uncertainty interval. This filter relies on assuming that the noise sources are bounded in order to obtain hard bounds on state and parameter estimates. These bounds are then compatible with robust control design so that the controller can be adequately updated in real-time. The method proposed is referred to as an extended set-membership filter and is applied to a two-state example and to the state and parameter estimations of a complex F-15 like model.

Robust Nonlinear Model Predictive Control With Partial State Information

A nonlinear controller methodology is developed based on nonlinear model predictive control and set-membership estimation, resulting in a controller which is robust to model uncertainties and bounded noise sources. The model predictive control makes direct use of the estimated bounds of the states and model parameters through integration as constraints in the optimization problem. This results in lower computation times and the ability to satisfy constraints with partial state information. Stability of the integrated estimation and control algorithm is guaranteed through the use of a contractive terminal cost. Real time implementation is developed through a unique sequential quadratic programming solver, which develops an initial feasible solution very quickly, and continues to optimize as time allows. The estimation and control algorithm is demonstrated in software simulation, hardware simulation, and in uninhabited aircraft flight tests. A formation flight example consisting of two aircraft is presented, showing how the controller can maintain a formation within given constraints and avoid collisions while in the presence of noise. Real time flight control results for an autonomous airplane demonstrate the capability of the algorithms to control a nonlinear aircraft in real time around a suddenly appearing threat, thus avoiding collisions.

Synthesis of embedded networks for building automation and control

We present a methodology and a software framework for the automatic design exploration of the communication network among sensors, actuators and controllers in building automation systems. Given 1) a set of end-to-end latency, throughput and packet error rate constraints between nodes, 2) the building geometry, and 3) a library of communication components together with their performance and cost characterization, a synthesis algorithm produces a network implementation that satisfies all end-to-end constraints and that is optimal with respect to installation and maintenance cost. The methodology is applied to the synthesis of wireless networks for an essential step in any control algorithm in a distributed environment: the estimation of control variables such as temperature and air-flow in buildings.

Simulation and Flight Test of Autonomous Aircraft Estimation, Planning, and Control Algorithms

The development, summary, and validation of real time autonomous estimation, planning, and control algorithms for nonlinear aircraft are presented. Two square root estimation algorithms are developed and used to estimate the state dynamics, parameters, and uncertainty bounds of a nonlinear aircraft in real time. A square root sigma point filter approximates the uncertain state using a finite set of points, while a square root extended set-membership filter bounds the states and parameters of the nonlinear system using ellipsoidal sets. The estimation methods are integrated into online optimized model predictive control methodology for fast reconfigurable control, and a path planner based on streamlines for fast path planning around obstacles and for target pursuit. The performances of the estimation, planning, and control techniques are tested in flight using the SeaScan autonomous aircraft. The SeaScan platform is briefly described, and real time hardware in the loop and flight data are presented for a variety of applications using each of the algorithms.

Multiscale consensus for decentralized estimation and its application to building systems

Multiscale approaches to accelerate the convergence of decentralized consensus problems are introduced. Consecutive consensus iterations are executed on several scales to achieve fast convergence for networks with poor connectivity. As an example the proposed algorithm is applied to the decentralized Kalman filtering problem for estimation of contaminants in building systems. Two conventional observers are designed and convergence is compared with respect to the number of communications necessary, which is an effective measure of system complexity. It is demonstrated that the proposed multiscale scheme substantially accelerates the decentralized consensus. Future extentions and directions are briefly summarized.

Active vibro-acoustic control of a flexible beam using distributed control

In this paper distributed control is applied to control the noise radiation of a flexible structure. The purpose is to investigate the effectiveness and ease of implementation versus current methods. As an illustrative example a simple supported beam is chosen. It is shown that the distributed controller is able to control the noise by penalizing the velocity states and it is shown that this results in less noise radiation, both in the near and far field. The distributed controller outperforms a local velocity feedback controller. Also, in comparison to local velocity feedback, it has the advantage that stability is guaranteed when the actuators are not assumed to be point actuators.