Fuzzy logic ratio control for a CVT hydraulic module

Compared with a conventional automatic transmission, which shifts among up to five gear ratios, a continuous variable transmission (CVT) uses the entire range of ratios between low and high gears. It achieves better fuel economy and drivability by constantly changing ratios to keep the engine running in its most efficient r.p.m. range based on driver demands. In this paper, a fuzzy logic controller is designed to control the primary pressure of the CVT. The extended dynamic range of the CVT and its nonlinear character dictate a dual global-local approach to the problem. The operating envelope of the CVT is viewed in state space as consisting of a finite number of cells, where the application of a control input transitions the controlled system from one cell to the next. Invariant and switching trajectories are effected in this global sense via a fuzzy logic controller. Within a particular cell, the CVT dynamics may be approximated via linearized equations, and a linear control law may be designed to accomplish this task. The structure and parameters of the global fuzzy logic controller are designed via clustering techniques and through genetic algorithms (GA) to be optimized. Using the proposed method, the controller is robust against parameter variations, simple to implement, easy to augment, and well optimized. Also, by choosing initial values from a suitable fuzzy model, the convergence time of the genetic algorithm is improved substantially.