Real-time Gear Shift Advisory System to Improve Fuel Economy
Publication Date: 2014-Mar-05
The IP.com Prior Art Database
Page 01 of 3
Real-time gear shift advisory system to improve fuel economy
Driver provides driving commands to the vehicle using Accelerator pedal, Brake pedal, Clutch pedal, Gear shifter and Steering wheel. Driving style, as a combination of the aforementioned driver control inputs has a significant effect on the fuel economy of the vehicle. In this document, we propose a method to provide feedback to the driver of a manual transmission vehicle on best times to shift and also a shifting performance score.
In prior art, gear performance is evaluated based on a lookup table of driver torque demand and current gear which does not consider the present load on the vehicle as a factor which is important when going uphill, downhill or towing. The proposed algorithm uses engine speed and engine power to evaluate shift performance which is directly related to the fuel economy and takes into account the load changes of the vehicle.
Gear shifting is a discrete event and the ideal gear shifting instant would be a fuzzy relation of its inputs. Hence, a fuzzy logic based gear shifting algorithm is the ideal choice and has the added benefit of computation efficiency, rules interpretability and ease of conversion from expert knowledge to code.
The block diagram of the algorithm is shown in Fig. 1.
Normalized Engine speed
Normalized Power used
Figure 1 Algorithm Block Diagram
Page 02 of 3
Normalized engine speed and a normalized engine power are used as input to the Fuzzy system. The normalized inputs are converted into fuzzy inputs using membership functions. The fuzzy inputs are applied to a set of rules are then defuzzified to obtain the shifting score and shifting feedback to the driver.
Fuzzification of inputs
To determine if the engine is operating at a fuel efficient region, the system uses inputs including engine speed and engine power. Each normalized input is transformed with calibrated membership functions to represent degree of similarity with predefined operating regions such as engine speed low / medium and high.
As shown in Figure 2, the system divides the engine speed into 3 regions of low, medium and high represented by 3 membership functions. For any given engine speed value, a fuzzification process will translate it into 3 [0, 1] values defining the degree of membership/similarities between given engine speed and 3 defined fuzzy regions. In this process, the sy...