Engine torque self-learning method, device, vehicle and computer readable storage medium

By constructing a multi-dimensional basic database in engine bench tests and combining it with vehicle self-learning, the torque model is dynamically corrected, solving the problem of accuracy reduction caused by wear and changes in operating conditions in traditional models, and achieving high-precision torque control and improved ride comfort.

CN122345985APending Publication Date: 2026-07-07CHONGQING CHANGAN AUTOMOBILE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
CHONGQING CHANGAN AUTOMOBILE CO LTD
Filing Date
2026-04-10
Publication Date
2026-07-07

AI Technical Summary

Technical Problem

Traditional engine torque models suffer from decreased calculation accuracy due to mechanical wear, changes in operating conditions, and environmental differences during long-term operation, resulting in problems such as sluggish power response, increased fuel consumption, and jerky mode switching. Furthermore, self-learning methods cannot distinguish the source of error, affecting the accuracy of correction.

Method used

By constructing a multi-dimensional basic database including friction torque and pumping loss in engine bench tests, and combining it with vehicle self-learning, deviation learning values ​​are obtained and dynamically corrected to ensure that the torque model is consistent with the actual state of the engine.

Benefits of technology

It improves the control precision of the torque model, reduces power interruption and jerking, and enhances the driving experience and the smoothness of engine control.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122345985A_ABST
    Figure CN122345985A_ABST
Patent Text Reader

Abstract

The application relates to the technical field of engine control, for example to an engine torque self-learning method, device, vehicle and computer readable storage medium. The engine torque self-learning method comprises the following steps: executing a calibration program in an engine bench test to construct a basic database of a torque model; when a vehicle operating condition meets a preset condition during vehicle operation, performing engine torque self-learning, and combining the basic database to obtain a deviation learning value for engine operating parameters; and correcting the torque model according to the deviation learning value. The bench calibration provides accurate reference data, the vehicle self-learning realizes online capture of the deviation, finally the torque model is consistent with the actual state of the engine through correction, the torque output is highly matched with the actual demand, the deviation is reduced, the control precision of the torque model is improved, and then the power interruption and jerk are eliminated, the smoothness is improved, and the driving experience is improved.
Need to check novelty before this filing date? Find Prior Art