How does a system change?
My research explores differential equations and geometry: the mathematics of change and structure. The practical thread is turning a useful way of thinking about equations into tools that solve them.
The curve z(x) = tanh(x) starts at zero and rises towards one. Its rate of change satisfies dz/dx = 1 − z². This supported Riccati equation is one example behind the calculator’s solver work.
A different view of the same equation.
On an interval where y is nonzero, define z = (dy/dx) / y. Differentiating gives dz/dx = (d²y/dx²) / y − z².
If y satisfies d²y/dx² + P(x) dy/dx + Q(x)y = 0, substitution gives dz/dx = −z² − P(x)z − Q(x). This is the classical Riccati transformation.
For dz/dx = 1 − z² with z(0) = 0, z(x) = (exp(x) − exp(−x)) / (exp(x) + exp(−x)). Differentiating verifies the equation, and substitution at zero verifies the initial condition. The denominator is positive for real x.
The transformation is local: zeros of y need separate treatment. For changing coefficients, a root of the instantaneous quadratic need not be a solution of the differential equation.
Manuscripts, checks and solver work.
The research includes Riccati equations, contraction integrals and geometric models. The current calculator development work implements a bounded set of exact solution families and reductions.
A revised Riccati manuscript and a 40-theorem Lean proof companion are in author review. Lean is a tool for checking formal mathematical statements; its results depend on the stated definitions and assumptions.
The public research repository also contains older drafts. Geometric physics models remain research hypotheses; mathematical calculations alone do not establish agreement with nature.