A model-based development environment for rapid-prototyping of latency-sensitive automotive control software

Abstract

The innovation in the field of automotive embedded systems has been increasingly relying on software-implemented functions. The control laws of these functions typically assume deterministic sampling rates and constant delays from input to output. However, on the target processors, the execution times of the software will depend on many factors such as the amount of interferences from other tasks, resulting in varying delays from sensing to actuating. Three approaches supported by tools, namely TrueTime, T-Res, and SimEvents, have been developed to facilitate the evaluation of how timing latencies affect control performance. However, these approaches support the simulation of control algorithms, but not their actual implementation. In this paper, we present a model interpretation engine running in a co-simulation environment to study control performances while considering the run-time delays in to account. Introspection features natively available facilitate the implementation of selfadaptive and fault-tolerance strategies to mitigate and compensate the run-time latencies. A DC servo controller is used as a supporting example to illustrate our approach. Experiments on controller tasks with injected delays show that our approach is on par with the existing techniques with respect to simulation. We then discuss the main benefits of our development approach that are the support for rapid-prototyping and the re-use of the simulation model at run-time, resulting in productivity and quality gains.

A Model-Driven Co-Design Framework for Fusing Control and Scheduling Viewpoints

“…For instance, if the control requirement is to achieve a desired settling time, defined as the time taken to settle within 2% of the steady state value, equal to $0.3$ s, then the deadline should be no greater than $8.2$ ms (versus $0.44$ s with a deadline of $9.09$ ms). In our previous work [5], we have exemplified the co-simulation of CPAL in Simulink to study the control system performance for different scheduling options.…”

“…In [5], we have discussed how to integrate the timing behaviour of the controller into Simulink models. In this work, we have studied how CPAL timing-accurate interpretation in Simulink compares against TrueTime and T-Res.…”

“…Both approaches use Simulink for control design and timing extension toolboxes to include computing aspects, foremost the effects of task scheduling on the control performance. In a recent study [5], we discussed our co-design and simulation environment and compared it with these state-of-the-art tools. Our co-design technique mainly differs from these approaches by allowing the control model to be directly executed (by an interpreter engine) on the target hardware, without changing a single line of code.…”

CPAL: high-level abstractions for safe embedded systems

“…CPAL can also be executed in a co-simulation environment. For the development of control systems, the execution in Matlab/Simulink as an S-function is supported (see [16]). CPAL is also the simulation language to program high-level protocol layers in the RTaW-Pegase embedded network simulator from RTaW.…”

“…Timing annotations used in simulation to capture the execution times observed at run-time. The Matlab/Simulink port of CPAL [16] enables to re-inject these delays in the control loops for the design of timing-sensitive controllers.…”