Stability analysis of two linear accommodation and convergence models.

Abstract

PURPOSE: To perform time domain stability analysis on two widely accepted linear models of accommodation and convergence. METHODS: For each model, the governing differential equations were used to compute the system matrix. The system matrices were used to calculate the respective trace and determinant from which eigen-vectors and eigenvalues were quantified. These characteristic numbers fully identified and classified the fixed points of each model and, thus, their stability. RESULTS: Controller gains, time constants, and accommodation and convergence cross-links determined model stability. Accommodation and convergence cross-links have the greatest influence on stability. A model-specific transition between stabilities was identified as the product of these cross-links, AC.CA(CRITICAL). For each model, three types of fixed points are described and displayed graphically: stable node, line of nonisolated fixed points, and saddle. CONCLUSION: We demonstrated the stability analysis of a two-dimensional linear system, working only in the time domain. The benefit of time-domain stability analysis is that it can be extended to nonlinear systems when frequency domain analysis techniques fail. Given that we live in a fundamentally nonlinear world and that the use and application of computational models is extensive, this is a valuable and important tool.