Power switching coordination control method of coaxial parallel hybrid electric vehicle with engine torque observer

Abstract

The invention discloses a power switching coordination control method of a coaxial parallel hybrid electric vehicle with an engine torque observer. Based on the method of a least square support vector machine, the testing sampledatum of the engine rack performance aretrained; the distribution estimation algorithm is adopted to optimize the parameter C and Lambda of the least square support vector machine; and through off-line training and optimization to an engine toque module and based on the optimum parameters, the engine torque observer is built. The current throttle value percentage and rotary speed are inputted to acquire a real-time torque value, thus realizing on-line observation to the engine torque. During the switching process of switching a pure electric driving model to a pure engine driving model (different power sources) of the vehicle provided herein, section coordination control is carried out aiming to different power sources, thus realizing stable transition of the whole switching process and stable torque of power system output without fluctuation, and satisfying the requirement of vehicle driving total torque.

Description

technical field [0001] The invention relates to a coordinated control method for power mode switching of a hybrid electric vehicle, in particular to a coordinated control method for power switching of a coaxial parallel hybrid electric vehicle with an engine torque observer. Background technique [0002] Hybrid electric vehicles can realize low-speed pure electric driving and parallel hybrid driving under medium-high speed and heavy load conditions, so that the control of vehicle energy flow and the optimization of energy consumption have greater flexibility, and it is easier to achieve low fuel consumption and low emission goals. However, due to the significant difference in the transient response characteristics of different power sources, switching between vehicle power sources often leads to large changes or sudden changes in the output torque of the power system, resulting in unstable power output and large longitudinal impact on the vehicle. impact, thereby reducing t...

AI Technical Summary

Problems solved by technology

However, this method has poor versatility, and different models need to be established for different types of engines, and a large amount of test data is required

(2) The neural network model method based on the engine dynamic test, which is based on the engine experimental data, and uses the characteristics of the strong nonlinear approximation ability of the neural network to estimate the dynamic torque of the engine, but the selection of the neural network type and network parameters There is no precise theoretical basis for

This algorithm needs to perform Fourier transform after obtaining the instantaneous speed sequence of the crankshaft in each cycle. The algorithm is complex, the amount of calculation is large, the hardware requirements of the control system are extremely high, and the real-time performance is not easy to guarantee.

Images