Systematic derivation of boundary Lax pairs

We systematically derive the Lax pair formulation for both discrete and continuum integrable classical theories with consistent boundary conditions.

💡 Research Summary

The paper presents a unified algebraic framework for constructing Lax pairs of classical integrable systems when non‑trivial boundary conditions are present, covering both discrete lattice models and continuous field theories. Starting from the quadratic Poisson algebra originally introduced by Sklyanin, the authors generalize the Sklyanin bracket to accommodate non‑skew‑symmetric r‑matrices, as shown in equation (1.1). This broader setting allows the same formalism to be applied to models whose Lax operators depend on a discrete site index or a continuous spatial coordinate.

In the discrete case, a monodromy matrix (T_a(\lambda)=L_{aN}(\lambda)\dots L_{a1}(\lambda)) is built from local Lax matrices (L_n). The classical Yang‑Baxter equation (1.5) guarantees that (T_a) satisfies the same quadratic Poisson algebra as the local Lax operators. To incorporate boundaries, the authors introduce non‑dynamical reflection matrices (K^{\pm}(\lambda)). Two families of boundary conditions are distinguished: soliton‑preserving (SP) and soliton‑non‑preserving (SNP). These correspond respectively to the reflection algebra (\mathcal{R}) and the twisted Yangian (\mathcal{T}), whose defining Poisson brackets are compactly written in (1.11). The matrices (\hat r, r^{}, \hat r^{}) in (1.10) encode the appropriate transpositions needed for each case.

A modified monodromy matrix \