Modeling and simulation of living systems as systems of systems

Duboz Raphaël, Soulie Jean-Christophe. 2017. Modeling and simulation of living systems as systems of systems. In : Guide to modeling and simulation of systems of systems. Zeigler Bernard P. (ed.), Sarjoughian Hessam S. (ed.). Cham : Springer, pp. 325-350. (Simulation foundations, methods and applications) ISBN 978-3-319-64133-1

Book chapter
[img] Published version - Anglais
Access restricted to CIRAD agents
Use under authorization by the author or CIRAD.

Télécharger (8MB) | Request a copy

Abstract : In this chapter, we have addressed some of the issues regarding modeling and simulation of living systems. We have seen that the systemic point of view of the DEVS formalism matches the systemic point of view adopted in the living sciences field. Two examples, one in animal epidemiology and the other in plant growth modeling, illustrated very different characteristics of DEVS and its extensions. The multi-formalistic abilities of DEVS increase the descriptive capacity of modelers when studying very complex systems. This capacity has not been used to its real potential in the field of living systems. Hopefully, its use will increase in the future since DEVS-based systems are very promising to help answer critical issues, as shown in this chapter, regarding natural risk management and poverty reduction. As we assume that DEVS can be used to specify a wide vatiety of living systems considered as a system of systems, it should be used for their specification. It is gene1ic enough to work as a universal formalism for living dynamical systems modeling and simulation. Of course, using a software environment such as the VLE platform greatly facilitates the model design, its implementation, and its execution. DEVS is an abstract formalism and can be hat·d to manage when the modeling effort has to focus on the application domain. That is why projects like RECORD (Bergez et al. 2012) provides an environment where DEVS is used at the simulation level and where the modeling level is composed by a set of specialized modeling components, where components are represented by approptiate fom1alisms. Doing that, the modeler can design the model using the most suitable formalism, or coupling several ones, without any knowledge of DEVS. The mappings of the main formalisms that ai·e used for living system modeling and simulation into DEVS already exist in VLE. The modeler can of course directly specify a model in DEVS as needed as well. In this chapter, we have discussed model continuity in the context of living systems. It is a difficult question since living systems follow complex interaction rules and we ai·e far from knowing all of them. Besides its usual function, model continuity can also apply here to help to the design of natural expe1iments based on viltual ones, saving time, and money. Such a very promising methodology to design experiments is still under construction. If we want to use simulation models for decision making, we need reliable simulators based on rigorous and highly communicable fom1alisms to be able to share model specifications and compare simulation results. Pairs should be able to reproduce any published works to check the results. As model communication is a hat·d task, it is important to provide techniques to facilitate it. The morphism between DEVS models and DEVS abstract simulators is a property that provides confidence in the ability to reproduce simulation results with a given model. The DEVS version of the Ecome1istem model presented in this chapter was validated comparing its simulation results with a previous, already validated, non-DEVS version. Results perfectly matched. Verification by pairs is a fundamental practice in science. As simulation models are more and more popular in scientific activity, this practice must be insured- unfortunately it is not always the case. The necessary interdisciplinary works to tackJe urgent issues such as the economical and ecological c1ises, diversity erosion, or povetty, will be based on creative compositions of shared representations. These shared models will embed heterogeneous knowledge elements at different scales, i.e., in heterogeneous formalisms. For these reasons, simulation models for living systems, viewed as systems of systems, should be fom1alized with DEVS and its extensions. While it is true that some work remains to standardize the DEVS simulators and specifications, we think that it is the most suitable fom1alism for heterogeneous model coupling. Such model coupling and cross-validation will play a critical role in the future of modeling and simulation in the field of living systems.

Classification Agris : U10 - Mathematical and statistical methods
U40 - Surveying methods

Champ stratégique Cirad : Hors axes (2014-2018)

Auteurs et affiliations

Autres liens de la publication

Source : Cirad-Agritrop (

View Item (staff only) View Item (staff only)

[ Page générée et mise en cache le 2020-10-02 ]