Intelligent variable time domain model prediction energy management method for hybrid power vehicle

Abstract

The invention discloses an intelligent variable time domain model prediction energy management method for a hybrid power vehicle. The method belongs to the field of plug-in hybrid electric vehicle energy management, and mainly comprises: establishing a double-neural-network working condition prediction model, and further establishing an MPC parallel hybrid electric vehicle energy management systembased on double-neural-network working condition prediction; establishing a DDPG controller by using a depth deterministic strategy gradient (DDPG) algorithm; combining the built models, building anintelligent variable time domain model prediction energy management model of the hybrid power vehicle, and training the model; and when an automobile runs, obtaining the relevant state and traffic condition information of the automobile to act on the trained model, and achieving energy management of the intelligent variable time domain MPC of the parallel hybrid electric vehicle. By adopting the method, the problem that the MPC adopting a fixed prediction time domain is poor in working condition adaptability can be effectively solved, the control effect of an energy management strategy is effectively improved, and the fuel economy of the vehicle is further improved.

Description

technical field [0001] The invention belongs to the field of energy management of plug-in hybrid electric vehicles, in particular to an intelligent time-varying domain model prediction energy management method for hybrid electric vehicles. Background technique [0002] In recent years, the rapid development of the automobile industry has brought great convenience to people's travel, but it has also caused problems such as environmental pollution and resource shortage. In order to solve these problems, the automobile industry continues to reform and innovate, and actively promotes the development of new energy vehicle technology, especially hybrid vehicles. Hybrid electric vehicles are divided into traditional hybrid electric vehicles (HEV) and plug-in hybrid electric vehicles (PHEV). Compared with traditional hybrid electric vehicles, plug-in hybrid electric vehicles have large-capacity power batteries and can be directly connected to The advantages of grid charging and pur...

AI Technical Summary

Problems solved by technology

[0004] Based on the future driving conditions, the model predictive control method realizes the rolling optimization of the future vehicle torque demand in the prediction time domain, and approaches the global optimal control effect with the real-time optimal control strategy, but the model predictive control method has a fixed prediction time. Researchers rely on experience or conduct experiments to determine it. The MPC method using a fixed prediction time domain cannot adapt well to changes in working conditions, and it is difficult to ensure the optimal fuel economy of vehicles.

In the actual driving process of the car, the weather conditions and road traffic are complex and changeable, and the driver's behavior is uncertain and diverse. These random problems greatly increase the difficulty of adjusting the control parameters of the model predictive controller.

Images