Mathematical models for collective behaviour
- Unit Coordinator: Debora Amadori
- ECTS Credits: 6
- Semester: 1
- Year: 2
- Campus: University of L'Aquila
- Language: English
- Aims:
Aim of the course is to present some mathematical models currently used in the analysis of collective phenomena, such as vehicular and pedestrian trauc, and locking phenomena. Emphasis will be given to the mathematical treatment of specific problems coming from real world applications.
- Content:
- Macroscopic trauc models. LWR model, its derivation. Fundamental diagrams. The Riemann problem, examples. Second order models for trauc low: Payne-Whitham model, description, drawbacks; Aw-Rascle model, shocks description, domains of invariance, instabilities near vacuum.
- Theory: systems of conservation laws, strict hyperbolicity, Rankine-Hugoniot conditions; Lax admissibility condition. The Riemann problem for systems: the linear case; GNL and LD fields; rarefactions and contact discontinuities. BV functions, examples and properties. A compactness theorem.
- Wave front tracking algorithm: approximate rarefactions, possible types of interactions. Bounds on number of waves and on total variation. Compactness of approximate solutions. The initial-boundary value problem on the half line: boundary Riemann problem, interactions with the boundary, control of the total variation by means of a Lyapunov-type functional. The Toll gate problem.
- Networks, basic definitions, conservation of the lux. Examples. Distributions along the roads, maximization of the lux. Riemann problem on a junction composed by 2 incoming roads and 2 outgoing roads. The case of 2 incoming roads and 1 outgoing road: the "right of way" rule. Junction between one incoming and one outgoing road, different luxes.
- Pedestrian low: normal and panic situation. Macroscopic models for evacuation, conservation of "mass", eikonal equation. The Hughes model for pedestrian low. The eikonal equation: non uniqueness, viscosity solutions, relation with vanishing viscosity approximation. The Hughes model in one space dimension. Curve of turning points, Rankine-Hugoniot conditions. The case of constant initial density and of symmetric initial data; conservation of the left and right mass; an example with mass exchange across the turning point. Macroscopic models for pedestrian low that include: knowledge of a preferred path, discomfort from walking along walls, tendency of avoiding high densities of pedestrian in a neighborhood (nonlocal term of convolution type), angle of vision, obstacle in the domain. Linearized stability around a smooth solution.
- Introduction to the theory of locking. Examples: Krause model for opinion dynamics, Cucker-Smale model, model for attraction-repulsion phenomena. The Cucker-Smale locking model: basic properties, estimates on the kinetic energy. A "locking theorem": proof by bootstrapping method (Ha and Tadmor). Some drawbacks of the model. Introduction to the kinetic limit for locking: the N-particle distribution function, Liouville equation, marginal distribution, continuity equation. The formal mean-Seld limit: a Vlasov- type equation.
- Reading list:
M.D. Rosini, Macroscopic models for vehicular lows and crowd dynamics: theory and applications. Springer. 2013.
M. Garavello, B. Piccoli, Trauc low on networks. Conservation laws models. AIMS Series on Applied Mathematics. 2006.
