An Automated Approach for Execution Sequence-Driven Software and Physical Co-Design of Mechatronic Systems Based on Hybrid Functional Ontology
Introduction
Most mechatronic systems (MTSs) have been becoming software-intensive [1]. MTSs generally comprise basic systems (BSs), which is composed by multi-physics subsystems, low-level controllers, sensors, and actuators to implement regulated continuous physical processes, coordinated by high-level information processing [2]. In essence, the high-level information processing is implemented by “software”. The design of MTSs requires integrated efforts from multiple domains [3]. The synergetic design of multi-physics and low-level controllers that regulate the physical processes have been well investigated by many researchers from conceptual to detailed design [4], [5], [6]. However, how to combine the design of software with them is still an open problem [1], [7]. Many sequential design methods that develop software after the physical design has been frozen have been widely criticized [8], [9]. In fact, software and physical designs are closely correlated with each other in early design. For example, the workflow coordinated by software constrains the selection of working structures of physical subsystems, and the chosen physical concepts may introduce new functions that change the workflow of software. If these cross-domain influences are not considered in early design, design defects and unexpected changes may occur in late stages and hence have to be solved with a high cost. Therefore, it is significant to provide computational support to capture and process such cross-domain influences in early system design.
In this study, the software-intensive mechatronic systems with high demand on execution sequences, e.g. systems of manufacturing and robotics domains are focused. They are required to achieve functions in a step-by-step manner following certain sequences. Such execution sequences work as important constraints on the physical concept selection and software generation, and meanwhile, may be changed by the physical and software design processes. Therefore, a specific aspect of software-physical influences which concentrates on the execution sequences of actions of the systems can be observed and is investigated in this study. An automated approach to assist the software and physical co-design driven by execution sequences is proposed. Semantic web technologies have been adopted as the main enabling technologies since they provide unified computer-readable representation and powerful reasoning capability across domains. They have been intensively used to share and infer knowledge among different disciplines, design stages, and platforms [10], [11], [12], [13]. However, the heterogeneous and unlinked knowledge of software and physical designs hinders their co-design. To this end, a hybrid functional ontology that unifies their functional representations is proposed to correlate the software and physical designs in an ontological knowledge base. Software and physical designs are explicitly linked to the unified functions and thus execution sequences of physical processes can be captured and affect the software and physical design automatically. Specifically, physical structures can be constrained by execution sequences of functions coordinated by software, and behaviors of software can be generated and adapted to physical structures. The contributions of the proposed approach are two-folds: (1) a hybrid functional ontology unifying software and physical functions is proposed and the structure and behavior knowledge of physical and software design is linked through it; (2) reasoning on the design knowledge is enabled by semantic web technologies such that the execution sequences extracted from the hybrid functional ontology can be used to constrain and generate the physical and software design simultaneously.
This paper is organized as follows. Literature related to multidisciplinary co-design is reviewed in Section 2. Section 3 gives an overview of the proposed approach. The hybrid functional ontology is introduced in Section 4. The method of evaluating physical modules by execution sequences is discussed in Section 5. Automated software behavior generation and updating are explained in Section 6. The implementation and case study are introduced in Section 7. Section 8 discusses the advantages and limitations of this approach. Section 9 concludes this paper and summarizes future works.
Section snippets
Multidisciplinary co-conceptual design
It has been a consensus and necessity that multidisciplinary co-design of MTSs should be based on a common knowledge representation. The architecture of an MTS can be described from multiple viewpoints addressing different concerns held by the system’s stakeholders [14]. In this study, three viewpoints, i.e., functional, structural and behavioral which are commonly involved in the conceptual design process are the focuses. Therefore, the co-design approaches are classified and reviewed
Motivation analysis
The main challenge to be addressed in this study is to provide a computer-aided design approach to capture and deal with the influences between software and physical designs of MTSs. Here, a mobile robot is used as a motivational case study to illustrate how the two domains influence each other in a typical system design process.
The mobile robot is designed to fetch a colored can in a maze. The concept of the robot is shown in Fig. 1(a), and the task is sketched in Fig. 1(b). The square
Hybrid functional ontology
The common practice of existing function-based co-design methods is that functions of multidisciplinary components are simply connected through flows that represent objects transferred among them. The drawback is that it can only capture relationships between intended interfaces instead of behaviors of them. In this study, a hybrid functional model (HFM) is proposed, which embeds the continuous changes specified by physical functions in the discrete execution sequences organized by software.
Automated evaluation of physical modules by execution sequences
The main task of physical design is to choose multi-physical mechanisms and their continuous controllers to implement physical functions. When a physical module only implements a single function, it can be determined by simply comparing its provided behavior and the required behavior of the function. However, when a physical module implements multiple functions, the module should satisfy the execution sequences of these functions. Traditional physical design approaches have not considered such
Automated software behavior generation and updating
Software coordinators organize different physical modules to achieve the main function of the target system according to its workflow. Therefore, their design is closely related to physical functions. To keep consistency between software design and physical functions, the software coordinators should be correlated with the functional model in the knowledge base. Based on correlations between them, the behaviors of software coordinators can be automatically generated and updated.
Implementation and case study
An ontology-based design framework is developed to support the proposed automated co-design approach. A CNC bending machine is used to illustrate how the proposed approach is practically used in a typical system design process.
Discussion
It can be seen from the case study that interactions between software and physical design happen quite frequently when the design advances iteratively. The semantic web technologies-enabled approach proposed in this study can assist the co-design process effectively. The quality of the design can be improved from three aspects:
(1) For physical design, a new type of constraints, i.e., execution sequences of functions are considered in this study such that the unsatisfactory physical concepts can
Conclusions and future work
In this study, an automated approach to assist software-physical co-design is proposed, which is enabled by semantic web technologies. The knowledge of software and physical designs are correlated through the unified functional ontology such that design decisions can be synchronized across domains to avoid design defects in early design. The main novelty of this work is that it observed the impact between the software-physical design in the early conceptual design of mechatronic systems and
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
The authors appreciate the support from the National Key Technology Support Program, PR China (2018YFB1700905), XXX(SHWXX20171ZL01), the National Science Foundation of China (61672247, 61772247 and 61873236), and the Open Fund of Science and Technology on Thermal Energy and Power Laboratory, PR China (No. TPL 2019A03).
References (72)
- et al.
Towards automation of control software: A review of challenges in mechatronic design
Mechatronics
(2010) - et al.
An architecture model to support cooperative design for mechatronic products: A control design case
Mechatronics
(2011) - et al.
Integrated system and control design of a one DoF nano-metrology platform
Mechatronics
(2017) - et al.
Survey on mechatronic engineering: A focus on design methods and product models
Adv Eng Inform
(2014) - et al.
Semantic interoperability in building design: Methods and tools
Comput Aided Des
(2006) - et al.
A model-driven ontology approach for manufacturing system interoperability and knowledge sharing
Comput Ind
(2013) - et al.
Towards effective management of inconsistencies in model-based engineering of automated production systems
IFAC-PapersOnLine
(2015) - et al.
Enriching the semantics of variational geometric constraint data with ontology
Comput Des
(2015) - et al.
Systematic comparison of functional models in SysML for design library evaluation
Procedia CIRP
(2014) - et al.
A knowledge-based framework for creative conceptual design of multi-disciplinary systems
Comput Aided Des
(2012)