Leveraging Domain Knowledge for the Efficient Design-Space Exploration of Advanced Cyber-Physical Systems

Abstract

Cyber-physical systems are becoming increasingly complex. In these advanced systems, the different engineering domains involved in the design process become more and more intertwined. In these situations, a traditional (sequential) design process becomes inefficient in finding good designs options. Instead, an integrated approach is needed where parameters in both the control and embedded domain can be chosen, evaluated and optimized to have a good solution in both domains. However, in such an approach, the combined design space becomes vast. As such, methods are needed to mitigate this problem. In this paper, we show how domain knowledge can be used to guide the design-space exploration process for an advanced control system and its deployment on embedded hardware. We use domain knowledge, captured in an ontology, to reason about the relationships between parameters in the different domains. This leads to a stepwise design space-exploration process where this domain knowledge is used to quickly reduce the design space to a subset of likely good candidates. In this process, we make use of cross-domain evaluation to find feasible design options with good system-level performance.

Gradient-free Multi-domain Optimization for Autonomous Systems

“…This synthesis is mainly based on the task the robot should fulfill and the results show that such a synthesis can outperform genetic algorithm optimizations. In the work of [11], the authors show how domain knowledge can be used to guide the designspace exploration process for an advanced control system and its deployment on embedded hardware. They choose this integrated approach because the parameters in both the control and embedded domain can be chosen, evaluated and optimized to have a performant solution in both domains although the combined design space becomes vast.…”

Sensorless Load Angle Control for Energy Optimal Sinusoidal Driven BLDC Motor Applications

“…However, here the speed is directly imposed by the user and thus always known. Therefore, a current level feedforward controller can be designed that anticipates on speed variations based on G d2 (18), see Section VI. The process dynamics of the system are characterized by the location of the poles being the roots of the denominator of (16)(17)(18).…”

“…Measurements in [17] show an energy saving potential up to 9.5% when the BLDC motor is driven with sinusoidal instead of square-wave shaped currents. With a view to final implementation, the first steps have already been taken for the design-space exploration process for the proposed advanced control technique and its optimal deployment on embedded hardware [18]. However, one missing link still needs to be tackled to arrive at a deployable energy optimal control technology for BLDC motors.…”

Automatic Generation of Workflows for Efficient Design Space Exploration for Cyber-Physical Systems

“…In previous work [7], [8], we have already shown how domain knowledge, captured in an ontology, is used to solve multi-domain optimization problems. We showed that by building an ontology of important design parameters and their interdependencies, it becomes possible to reason about the design space exploration workflow and to subsequently determine an efficient one.…”

“…In previous work, we manually derived a DSE workflow for the same example case as in this paper [7]. We did this by first building a similar ontology of the system under design.…”