Add a QTT solver based on TensorTrainNumerics DMRG

[ohno]

Summary

• add a radial quantics tensor train solver backed by TensorTrainNumerics.jl
• construct the Hamiltonian as a compressed QTT/MPO and solve low-energy states with two-site DMRG
• support excited states through projector deflation and report energies, overlaps, residuals, and tensor ranks
• document the QTT representation, theory, usage, limitations, acknowledgments, and bibliography

Motivation

This implements the QTT method requested in #26 while keeping the solver as an additional method alongside the existing finite-difference and variational approaches. The returned energies are Rayleigh quotients of the original, undeflated Hamiltonian.

Validation

• QTT test set: 82/82 passed
• Documenter QTT page built with examples executed
• spherical oscillator example at 32 grid points: E₀ = 1.4907554064, E₁ = 3.4534502119, with off-diagonal overlap approximately 8e-14
• Pkg.test() was also attempted, but its isolated environment did not finish initial compilation within 10 minutes and produced no failing-test output

Close #26

---

This pull request was developed with assistance from Codex using GPT-5.6 Sol with High reasoning effort.