Development of a system based on artificial intelligence algorithm for the optimization of vehicle clutch components

In this thesis, the vehicle diaphragm and the damper design were optimized for the vehicle clutch elements' dynamic behavior and functionality. Failure to adjust the balance between explorer and exploitation in algorithms and the inability to perform appropriate parameter settings may cause the problem of local optimum and early convergence. Therefore, it is essential to adjust the selection mechanisms and algorithm parameters in algorithms. For this purpose, an algorithm based on the Coyote Optimization algorithm, inspired by the coyote's social status and adaptation behaviors in the packs, was developed for vehicle clutch elements. An improved algorithm was obtained by improving the diversity and search mechanism of the coyote optimization algorithm. The resulting improved algorithm was tested with constrained and unconstrained engineering problems, and its validity was proven by applying it to vehicle clutch elements. Test results of the suggested algorithm were checked against other algorithms in the literatüre to measure the efficiency of the novel algorithm. The experimental results show that the novel algorithm has significantly better efficiency on constrained and unconstrained test functions. This optimization study introduced a new diaphragm model that performs better in more dynamic conditions than the specified minimum disengagement load compared to the original model of the vehicle's clutch elements diaphragm spring. In the second optimization study, a damper model with better damping characteristics was obtained than the original model of the vehicle clutch damper element. Thus, the coyote optimization algorithm developed is the first in the literature and optimized vehicle clutch elements.