Section: Application Domains
Other application fields
Besides the above mentioned axes, which constitute the project's identity, the methodological tools described in Section have a wider range of application. We currently carry on also the following research actions, in collaboration with external partners.

Modeling cell dynamics. Migration and proliferation of epithelial cell sheets are the two keystone aspects of the collective cell dynamics in most biological processes such as morphogenesis, embryogenesis, cancer and wound healing. It is then of utmost importance to understand their underlying mechanisms.
Semilinear reactiondiffusion equations are widely used to give a phenomenological description of the temporal and spatial changes occurring within cell populations that undergo scattering (moving), spreading (expanding cell surface) and proliferation. We have followed the same methodology and contributed to assess the validity of such approaches in different settings (cell sheets [122], dorsal closure [58], actin organization [57]). However, epithelial cellsheet movement is complex enough to undermine most of the mathematical approaches based on locality, that is mainly traveling wavefrontlike partial differential equations. In [109] it is shown that MadinDarby Canine Kidney (MDCK) cells extend cryptic lamellipodia to drive the migration, several rows behind the wound edge. In [149] MDCK monolayers are shown to exhibit similar nonlocal behavior (long range velocity fields, very active borderlocalized leader cells).
Our aim is to start from a mesoscopic description of cell interaction: considering cells as independent anonymous agents, we plan to investigate the use of mathematical techniques adapted from the meanfield game theory. Otherwise, looking at them as interacting particles, we will use a multiagent approach (at least for the actin dynamics). We intend also to consider approaches stemming from compartmentbased simulation in the spirit of those developed in [106], [111], [113].

Game strategies for thermoelastography. Thermoelastography is an innovative noninvasive control technology, which has numerous advantages over other techniques, notably in medical imaging [142]. Indeed, it is well known that most pathological changes are associated with changes in tissue stiffness, while remaining isoechoic, and hence difficult to detect by ultrasound techniques. Based on elastic waves and heat flux reconstruction, thermoelastography shows no destructive or aggressive medical sequel, unlike Xray and comparables techniques, making it a potentially prominent choice for patients.
Physical principles of thermoelastography originally rely on dynamical structural responses of tissues, but as a first approach, we only consider static responses of linear elastic structures.
The mathematical formulation of the thermoelasticity reconstruction is based on data completion and material identification, making it a harsh ill posed inverse problem. In previous works [123], [132], we have demonstrated that Nash game approaches are efficient to tackle illposedness. We intend to extend the results obtained for Laplace equations in [123], and the algorithms developed in Section 3.1.2.4 to the following problems (of increasing difficulty):
 Simultaneous data and parameter recovery in linear elasticity, using the socalled Kohn and Vogelius functional (ongoing work, some promising results obtained).
 Data recovery in coupled heatthermoelasticity systems.
 Data recovery in linear thermoelasticity under stochastic heat flux, where the imposed flux is stochastic.
 Data recovery in coupled heatthermoelasticity systems under stochastic heat flux, formulated as an incomplete information Nash game.

Constraint elimination in QuasiNewton methods. In singleobjective differentiable optimization, Newton's method requires the specification of both gradient and Hessian. As a result, the convergence is quadratic, and Newton's method is often considered as the target reference. However, in applications to distributed systems, the functions to be minimized are usually “functionals”, which depend on the optimization variables by the solution of an often complex set of PDE's, through a chain of computational procedures. Hence, the exact calculation of the full Hessian becomes a complex and costly computational endeavor.
This has fostered the development of quasiNewton's methods that mimic Newton's method but use only the gradient, the Hessian being iteratively constructed by successive approximations inside the algorithm itself. Among such methods, the BroydenFletcherGoldfarbShanno (BFGS) algorithm is wellknown and commonly employed. In this method, the Hessian is corrected at each new iteration by rankone matrices defined from several evaluations of the gradient only. The BFGS method has "superlinear convergence".
For constrained problems, certain authors have developed socalled Riemannian BFGS, e.g. [152], that have the desirable convergence property in constrained problems. However, in this approach, the constraints are assumed to be known formally, by explicit expressions.
In collaboration with ONERAMeudon, we are exploring the possibility of representing constraints, in successive iterations, through local approximations of the constraint surfaces, splitting the design space locally into tangent and normal subspaces, and eliminating the normal coordinates through a linearization, or more generally a finite expansion, and applying the BFGS method through dependencies on the coordinates in the tangent subspace only. Preliminary experiments on the difficult Rosenbrock testcase, although in low dimensions, demonstrate the feasibility of this approach. Ongoing research is on theorizing this method, and testing cases of higher dimensions.

Multiobjective optimization for nanotechnologies. Our team takes part in a larger collaboration with CEA/LETI (Grenoble), initiated by the Inria ProjectTeam Nachos, and related to the Maxwell equations. Our component in this activity relates to the optimization of nanophotonic devices, in particular with respect to the control of thermal loads. We have first identified a gradation of representative testcases of increasing complexity:
These cases involve from a few geometric parameters to be optimized to a functional minimization subject to a finiteelement solution involving a large number of dof's. CEA disposes of such codes, but considering the computational cost of the objective functions in the complex cases, the first part of our study is focused on the construction and validation of metamodels, typically of RBFtype. Multiobjective optimization will be carried out subsequently by MGDA, and possibly Nash games.