فهرست نشریات علمی دسترسی آزاد ایران

This study develops a nonlinear dynamic modeling framework to analyze and predict performance behavior in industrial environments using competitive-intelligence-related variables. Four organizational resource components are formulated as elements of a discrete-time state vector, and their influence on system output is modeled through a nonlinear state-transition function. Empirical observations collected from a steel manufacturing company were used to identify the unknown dynamics through a feed-forward artificial neural network trained via a gradient-based optimization procedure. Reliability of the measurement instrument was verified using Cronbach’s alpha coefficients of 0.92 and 0.86 for the independent and dependent constructs, respectively. The identified model demonstrates stable convergence, with the minimum prediction error achieved near iteration 1500, and outperforms a linear baseline in mean-squared error and correlation accuracy. The proposed formulation provides a mathematically oriented approach for reconstructing performance-driven system behavior and establishes a foundation for future extensions involving adaptive estimation, robust analysis, and optimal control strategies in industrial systems.