Vehicle gear shifting quality evaluation method based on nerval net

Abstract

The present invention relates to a method for evaluating the property of vehicle gearshift based on the Neural Network, and includes four basic procedures as follows: First, the index for the evaluation of the property of a gearshift is determined. Second, the sample data of evaluation index and subjective evaluation ratings are withdrawn. Third, the Neural Network is trained. Fourth, a model for the property of the gearshift is built, and an objective evaluation is achieved. The objective evaluation method is accomplished by adopting the mode identification and the prediction capability of the Neural Network. The present invention can evaluate the property of the gearshift effectively and objectively, possess good intelligence and stability and have good consistency with the outcome of the subjective evaluation. The present invention can provide guidance and aid for increasing the property of a vehicle gearshift and establishing a dynamic comprehensive evaluation system for the property of the vehicle gearshift. Manufacturing factories and research institutes in the industry can use the system to accomplish an evaluation task of the property of the gearshift efficiently.

Description

technical field [0001] The invention relates to the content of automatic transmission in the technical field of automobile transmission, and uses a neural network model as the core to realize a method for evaluating the shifting quality of a stepped automatic transmission system, and the method is also suitable for evaluating the driving comfort of a vehicle. Background technique [0002] For a long time, the evaluation of shifting quality has mostly been carried out by various automobile manufacturers using the subjective method of driver satisfaction as the measurement standard, that is, some specially trained or experienced drivers are invited to conduct real-time evaluation, and a system is evaluated by test drive scoring. The smoothness of shifting is good or bad. Usually, the evaluation level is set as 1-10 levels for subjective scoring, as shown in Table 1. Among them, above 9 is the quality level, with superior performance; 6-8 is the commodity level, which can be a...

AI Technical Summary

Problems solved by technology

[0006] ① Different drivers (age, gender, experience, etc.) have different driving experiences, driving habits and driving intentions, and the evaluation level is limited by the subjective factors of the testers, which makes the method too subjective;

[0007] ②Because many testers are needed for subjective evaluation, the test consumes more manpower and financial resources, which prolongs the vehicle development cycle an

Images