Lernfähige Fuzzylogik-Drucksteuerung mit geschlossenem Regelungskreis für automatisches Getriebe

摘要

A self-organising fuzzy logic controller develops an adaptive pressure adder that modifies the magnitude of pressure supplied to a friction element of an automatic transmission. The controller uses a turbine speed versus engine torque table in KAM (Keep Alive Memory) to store the adaptive pressure adder for each shift. Tables in KAM represent fuzzy rules which are altered based on shift performance criteria to produce a self-organising fuzzy logic controller. The criteria, which define shift performance, include the ratio change slip time and the initial ratio change time divided by the desired slip time (called target ratio). The desired slip time is determined from a calibratable matrix. After a shifting event, an adaptive pressure adjustment is determined for the next shifting event having the same conditions. A two-input (slip time and target ratio) fuzzy logic controller is used to determine the pressure adjustment. The adaptive pressure adjustment from the fuzzy controller is then increased if stage time is greater than a calibratable value. The adaptive pressure is also multiplied by a factor determined from the deviation in the throttle position reading for the shift event. Once the adaptive pressure is calculated, the adjustment is added to the cells of the adaptive pressure matrix that correspond to the turbine speed and engine torque conditions.