Utilizing a graph data structure to model physical effects and dependencies between different physical variables for the systematic identification of sensory effects in design elements

Abstract

Gaining accurate data from technical systems has become of interest, particularly in the context of condition monitoring and predictive maintenance. Hereby it is important to gather precise and reliable data. To accomplish this task, various sensors with different physical effects are used. Depending on the sensor's position and measurand, different models are necessary to describe the path from the desired variable of interest to the actual measured one. To support designers, a physical effect catalog was digitalized using a graph data structure, which uses the inherent properties of a graph to represent physical variables, physical effects and their relationships. This graph structure together with its applicability in a sensor selection process will be shown in this paper.

A Follow-Up on the Methodical Framework for the Identification, Analysis and Consideration of Uncertainty in the Context of the Integration of Sensory Functions by Means of Sensing Machine Elements

“…In the framework by Welzbacher et al (2022), the identification of disturbance factor induced data and model uncertainty relied on the analog effect catalog by Vorwerk-Handing (2021). To overcome the therefrom resulting limitations, the effect graph by Kraus et al (2022) was adapted and functionally extended to be applicable for the intended purpose. Therefore, the user interface of the effect graph was extended by a generalized version of the disturbance factor control list by Welzbacher et al (2021).…”

“…Against the background of an automated identification of sensory utilizable physical effects that can be used to fulfil an application-specific measurement goal, Kraus et al (2022) developed an effect graph based on the multipole-based effect catalog by Vorwerk-Handing ( 2021). The graph consists of two parts: the graphical user interface and the effect database, cf.…”

“…Figure 3. A brief description of the effect graph is given in the following; for a detailed description, the reader is referred to Kraus et al (2022). The effect database contains physical effects, their respective laws as well as their influencing variables and parameters.…”

“…Figure 5: Extended graphical user interface of the effect graph by Kraus et al (2022) To identify disturbance factor induced data and model uncertainty, the user must first select the types of energy flows caused by occurring disturbance factors (a) and the effect chain that describes the sensory function to be analyzed (b) in the user interface of the effect graph. The user then has the option to apply a filter (c) regarding the maximum permissible length of effect chains to be found that link the disturbance factor characterizing variables and the function variables as well as inputs of the sensory function.…”

“…As described in Section 2.2, the identification of disturbance factor induced data and model uncertainty in the methodical framework by Welzbacher et al (2022) relies on the analog effect catalog by Vorwerk-Handing ( 2021), which makes it laborious and prone to error. To overcome this limitation, the effect graph by Kraus et al (2022) is adapted and functionally extended to be applicable for a systematic identification of this type of uncertainty. By allowing for an automated identification, the efficiency of the identification process can be enhanced while simultaneously reducing its error-proneness.…”