In this paper an improved chaotic switching algorithm based on modified Lozi system is applied. Generally speaking, the main advantage of the switched encrypted model is its robustness to the noise, while the main drawback is the slow processing speed. In this article we propose to gain model productivity by adjusting parameter in Lozi chaotic generator. Chaos generator is sensitive to any structure, therefore the solution shouldn't influence the pseudo randomness which is required for encryption. We provide the results of switched chaotic model based on Lozi chaotic generator with changeable parameter studied for chaoticity and pseudo-randomness with NIST, largest Lyapunov exponent, auto-correlation, cross-correlation and cumulative distribution.