<p>To explore the positive effects of unknown parameters and memristor terms as variable terms on chaotic systems in system construction and simulation applications, this paper provides a multi-level explanation from system feasibility analysis to chaotic image encryption system application. Firstly, based on the deformed Lorenz system, feasible cases for combining memristors with chaotic systems are proposed. Starting from the study of the system’s dynamic behavior, the chaotic characteristics of the system under different feasibility conditions are deeply analyzed. The system complexity under different initial conditions is also explored. Secondly, the unknown parameters and memristor terms are combined with the chaotic system, and the Kolmogorov-Sinai entropy is used to measure the system information generation rate and the information loss in the chaotic system. Finally, the chaotic system constructed using unknown parameters and memristor terms is combined with the improved parallel diffusion-obfuscation encryption algorithm to simulate and test the anti-interference, anti-differential attack and other capabilities of the chaotic image encryption system. The security of the improved chaotic image encryption system is introduced from different perspectives.</p>