Sapientia Monographs
Linear Models for Nonlinear Dynamics
Foundations, representations, and finite models
How finite linear models can describe nonlinear dynamics: choose functions or distributions in which evolution is linear, decide what finite information to keep, and account for what is left out.
How can a finite linear model help us understand nonlinear motion? The book’s organizing idea is lift, then project: first choose functions or distributions in which the evolution is linear, then choose the finite information to keep and account for what it leaves out.
The opening chapters derive a finite lift line by line and build the matrix, approximation, probability and transport foundations. Later chapters add memory, fitting from data, delay coordinates, compression and bounds, then develop operator background, exactness and spectra, diffusions, distributed systems and controlled lifts. The final chapters cover predictive control, observers, integrated case studies and task-based model selection.
For the reader
- Every chapter includes exercises, an executable Python lab and selected solutions.
- Separate reading routes for prediction and identification, for control, and for stochastic and distributed systems.
- Worked examples come before the abstract qualifications, and advanced results state their assumptions and identify the sources of their proofs.
Contents
- Why finite linear models?
- Linear dynamics and the matrix tools we need
- Functions as coordinates
- Approximation, projection, and least squares
- Evolution of observables
- Polynomial lifts and Carleman hierarchies
- Probability and finite Markov chains
- Distributions, transport, and duality
- Projection, residuals, and memory
- Learning finite evolution matrices
- Better coordinates and delay coordinates
- Modeling memory
- Compressing a finite input-output model
- Bounding without reconstructing a trajectory
- From matrices to operators
- Exactness, spectra, and resonance
- Stochastic dynamics and diffusions
- Distributed systems and two projections
- Inputs and bilinear lifts
- Predictive control and robustness
- Observers and state reconstruction
- Integrated computational case studies
- Choosing a modeling strategy
- Appendix A: Laboratory practice and numerical accuracy
- Selected solutions
- Index
Details
- Series
- Sapientia Monographs
- Type
- Research monograph
- Status
- Forthcoming. Full book coming soon
- Publisher
- Sapientia Scientifica
- Readership
- Assumes calculus, basic matrix algebra and introductory ordinary differential equations
- Subjects
- Dynamical systems, Applied mathematics