Manel Velasco

Managing quality-of-control in network-based control systems by controller and message scheduling co-design

In network-based control systems (NCSs), plant sensor-controller-actuator nodes in closed-loop operation drive principal network traffic. The quality-of-control (QoC) in an NCS, i.e., the performance delivered by each closed-loop operation, depends not only on the controller design but also on the message scheduling strategy. In this paper, we show that the co-design of adaptive controllers and feedback scheduling policies allows for the optimization of the overall QoC. First, we discuss the limitations of standard discrete-time control models for controllers of control loops that are closed over communication networks. Afterwards, we describe an approach to adaptive controllers for NCS that: 1) overcomes some of the previous restrictions by online adapting the control decisions according to the dynamics of both the application and executing platform and 2) offers capabilities for dynamic management of QoC through message scheduling.

Optimal state feedback based resource allocation for resource-constrained control tasks

In many application areas, including control systems, careful management of system resources is key to providing the best application performance. Most traditional resource management techniques for real-time systems with multiple control loops are based on open-loop strategies that statically allocate a constant CPU share to each controller, independent of their current resource needs. This provides average control performance with minimal overhead but in general fails to provide the best performance possible within the available resources. We show that by using feedback to dynamically allocate resources to controllers as a function of the current state of their controlled systems, control performance can be significantly improved. We present an optimal resource allocation policy that maximizes control performance within the available resources and provide experimental results showing that the optimal policy 1) significantly increases control performance compared to traditional control system implementations (by more than 20% in our experiments), 2) maximizes control performance over other feedback-based policies, 3) saves resources when perturbations occur infrequently, and 4) incurs negligible overhead.

On Lyapunov sampling for event-driven controllers

This paper investigates an event condition for event-driven controllers based on Lyapunov functions. Considering that constant values of a Lyapunov function define contour curves that form closed regions around the equilibrium point, in this paper we present a sampling mechanism that enforces job executions (sampling, control algorithm computation and actuation) each time the system trajectory reaches a given contour curve. By construction, the sequence of generated samples is stable in the discrete Lyapunov sense. However, in order to ensure that the system trajectory will tend to zero as time tends to infinity, it must be ensured that the sequence of samples is infinite. We provide conditions to ensure this property. The approach is illustrated by simulated examples.

A control approach to bandwidth management in networked control systems

Bandwidth allocation techniques for control loops closed over communication networks are based on static strategies that ensure average control performance at the expense of permanently occupying the available bandwidth. We present a dynamic approach to bandwidth management in networked control systems that allows control loops to consume band-width according to the dynamics of the controlled process while attempting to optimize overall control performance. By augmenting the original state-space representation of each controlled process with a new state variable that describes the network dynamics, 1) the allocation of bandwidth lo control loops can he done locally at run-time according to the state of each controlled process without causing overload situations and 2) control laws can be designed to account for the variations in the assigned bandwidth preventing the unexpected control performance degradation and even destabilization that would otherwise occur. Experimental data shows that this approach improves control performance with respect to the static strategy and uses less bandwidth.

Optimal Online Sampling Period Assignment: Theory and Experiments

In embedded systems, the computing resources are often scarce and several control tasks may have to share the same computer. In this brief, we assume that a set of feedback controllers should be implemented on a single-CPU platform. We study the problem of optimal sampling period assignment, where the goal is to assign sampling rates to the controllers so that the overall control performance is maximized. We derive expressions relating the expected cost over a finite horizon to the sampling period, the computational delay, and the amount of noise acting on the plant. Based on this, we develop a feedback scheduler that periodically assigns new sampling periods based on estimates of the current plant states and noise intensities. Extensive experiments show that online sampling period assignment can deliver significantly better control performance than the state-of-the-art, static period assignment.

Quality-of-Control Management in Overloaded Real-Time Systems

Transient overload conditions may cause unpredictable performance degradations in computer controlled systems if not properly handled. To prevent such problems, a common technique adopted in periodic task systems is to reduce the workload by enlarging activation periods. In a digital controller, however, the variation applied on the task period also affects the control law, which needs to be recomputed for the new activation rate. If computing a new control law requires too much time to be performed at runtime, a set of controllers has to be designed offline for different rates and the system has to switch to the proper controller in the presence of an overload condition. In this paper, we present a method for reducing the number of controllers to be designed offline, while still guaranteeing a given control performance. The proposed approach has been integrated with the elastic scheduling theory to promptly react to overload conditions. The effectiveness of the proposed approach has been verified through extensive simulation experiments performed on an inverted pendulum. In addition, the method has been implemented on a real inverted pendulum. Experimental results and implementation issues are reported and discussed

Resource management for control tasks based on the transient dynamics of closed-loop systems

This paper presents a resource management strategy for control tasks that maximizes control performance within the available resources by readjusting the task periods at runtime. A feedback scheduler is used to determine on-line the optimal task periods considering the response over a finite time horizon of the plants controlled by arbitrary linear control laws. We show how this problem can be expressed as an optimization problem, where the objective function relates the sampling periods to the transient responses of the controlled plants, and where restrictions are based on EDF schedulability constraints. For the general case, the solution of the optimization problem is computationally expensive, and thus, an approximate procedure to be executed on-line has been developed. We present simulation results that validate the presented approach

The One-Shot Task Model for Robust Real-Time Embedded Control Systems

Embedded control systems are often implemented in small microprocessors enabled with real-time technology. In this context, control laws are often designed according to discrete-time control systems theory and implemented as hard real-time periodic tasks. Standard discrete-time control theory mandates to periodically sample (input) and actuate (output). Depending on how input/output (I/O) operations are performed within the hard real-time periodic task, different control task models can be distinguished. However, existing task models present important drawbacks. They generate task executions prone to violate the periodic control demands, a problem known as sampling and latency jitter, or they impose synchronized I/O operations at each task job execution that produce a constant but artificially long I/O latency. In this paper, the one-shot task model for implementing control systems in embedded multitasking hard real-time platforms is presented. The novel control task model is built upon control theoretical results that indicate that standard control laws can be implemented considering only periodic actuation. Taking advantage of this property, the one-shot task model is shown to remove endemic problems for real-time control systems such as sampling and latency jitters. In addition, it can minimize the harmful effects that long I/O latencies have on control performance. Extensive simulations and real experiments show the feasibility and effectiveness of the novel task model, compared to previous real-time and/or control-based solutions.

Self-triggered networked control systems: An experimental case study

A self-triggered controller is characterized, in general, by a non-periodic sequence of job activations. And each job execution, apart from performing sampling, control algorithm computation and actuation, calculates the next job activation time as a function of the plant state. This paper describes the implementation of self-triggered controllers in networked control systems (NCS). The implementation corroborates that self-triggered control can be used for minimizing bandwidth utilization while providing similar control performance than periodic controllers.

Runtime Allocation of Optional Control Jobs to a Set of CAN-Based Networked Control Systems

The Controller Area Network (CAN) provides the basis for many cost-effective distributed embedded systems. In this paper, we present a novel approach to networked control systems (NCS) analysis and design that provides increased control performance for a set of control loops that share a single CAN network. This is achieved by enabling the following functionality for each control loop: first, standard periodic messaging is guaranteed to ensure stability, and second, non-periodic additional messaging is added whenever bandwidth is available in such a way that the aggregated control performance for all control loops is improved. To validate the presented approach, a proof-of-concept implementation is presented, and the extensive experimental results show the type of benefits that can be achieved as well as the resulting behavior depending on several key parameters. Moreover, it demonstrates that the approach can be implemented in practice.