This study offers a new Mixed Meta Heuristics algorithm (HGSABAT) for estimating the parameter values of each of the six categories of Non-Linear regression models examined (Misrald, Meyer4, Meyer7, Militky4, Militky2, and MGH09) by combining the Gravitational Search Algorithm and Bat Algorithm. Some models have different numbers of parameters. For example, the Misrald and Militky2 models of the Non-Linear Regression model have two parameters (Bl, B2). In contrast, the MGH09 and Militky4 models have four parameters (MGHl, MGH2, MGH3, and MGH4), in which location as the Meyer4 and Meyer7 models have three attributes (Meyerl, MGH2, and MGH3). To examine the effectiveness of the suggested Hybrid Meta Heuristics algorithm (HGSABAT), a simulation study based on Mean Square Error criteria is employed. Furthermore, using three metaheuristic algorithms, the simulation results are compared using Practical Swarm Enhancement, Gravitational Search, and BAT. The numerical results will be obtained using the Matlab program version 2020. Because the hybrid algorithm (HGSABAT) has a lower mean error than one of the other techniques, the mean error in the squares estimation techniques that were taken into consideration, the results demonstrated that it is satisfactory for parameter
