Mixture deviation self-learning method, system and readable storage medium

Abstract

In the mixed gas deviation self-learning method, system and readable storage medium provided by the present invention, first the domain controller obtains the model deviation of the pre-control model in the engine controller and the working condition data corresponding to the model deviation; then the domain controller obtains the model deviation according to the model The deviation and working condition data are divided into candidate areas, and the candidate areas are generated and the self-learning parameters corresponding to the candidate areas are obtained; then the domain controller sends the self-learning parameters to the engine controller; the engine controller uses the self-learning parameters to update The corresponding self-learning algorithm in the engine controller, and based on the self-learning algorithm, performs self-learning of the mixture deviation. In this way, the newly divided candidate area can be used as the self-learning area, so that the update of the self-learning area can be realized, and the self-learning divergence difference of the air-fuel mixture deviation can be updated, which helps to improve the air-fuel ratio control accuracy of the air-fuel mixture.

Description

technical field [0001] The invention relates to the technical field of automobiles, in particular to a method, system and readable storage medium for self-learning of mixture gas deviation. Background technique [0002] With the continuous tightening of emission and fuel consumption regulations, the requirements for vehicle emissions are becoming more and more stringent. For a vehicle's engine, the pre-control deviation of the air-fuel ratio (that is, the mass ratio between air and fuel in the mixture) has a significant impact on the accuracy of the engine's air-fuel ratio control, and then affects the emission characteristics of the vehicle's entire life cycle. In the powertrain control, the local linear model tree (lolimot) algorithm can usually be used for self-learning of the mixture deviation to improve the emission characteristics of the vehicle and improve the emission consistency. [0003] However, the lolimot algorithm currently set in the engine controller (Electr...

AI Technical Summary

Problems solved by technology

[0003] However, the lolimot algorithm currently installed in the engine controller (Electronic Control Unit, ECU) of the vehicle is based on the self-learning area demarcated offline. The self-learning area is only obtained by measuring and calculating a few experimental vehicles, and It cannot completely replace the characteristics of individual vehicles on the market; moreover, traditional engine controllers are limited by storage resources and computing resources, so they can only self-learn the self-learning parameters in the self-learning interval after dividing the self-learning interval. The self-learning of the self-learning interval division cannot be carried out, which affects the control accuracy of the air-fuel ratio of the vehicle mixture

Images