An intelligent learning method for energy management and control of pure electric vehicles

Abstract

The invention discloses a pure electric car energy managing and controlling method for intelligent learning. The method comprises the steps that by means of an accelerating pedal opening curve, the purpose of a driver is learned, required power of a motor is obtained, and by means of a data bus, the temperature state of a power battery pack and the electrical charge state obtained through state estimation are obtained; according to obtained vehicle information, a pure electric car energy management model based on self-adaptive dynamic planning is set up; by means of double-network intelligentlearning in a self-adaptive dynamic planning method, the energy management module is solved, and the optimal distribution power is obtained to a power battery heat management system. According to themethod, self-adaptive dynamic planning is adopted, the power of the power battery pack of a pure electric car is distributed optimally, the problem that conventional dynamic planning cannot be controlled online in real time is solved, the problem that the optimization degree of threshold control and off-line fuzzy control is solved, and on the basis of meeting the requirement for normal driving ofthe car, the power battery pack can be protected effectively in a temperature control manner.

Description

technical field [0001] The invention belongs to the technical field of pure electric vehicle energy management, in particular to an intelligent learning energy management control method for pure electric vehicles. Background technique [0002] Electric vehicle energy management system belongs to the core of electric vehicle technology. The current research on control strategies mainly focuses on hybrid electric vehicles, because hybrid electric vehicles have relatively complex powertrain systems, and there is a large space for related optimization strategies. For example, a series hybrid electric vehicle typically employs a thermostat strategy, a power tracking strategy, and a ground rules strategy. Parallel hybrid electric vehicles usually adopt logic threshold strategy, fuzzy logic control strategy and global optimal energy management strategy. Parallel hybrid electric vehicle usually adopts the engine constant operating point strategy, the engine optimal working curve e...

AI Technical Summary

Problems solved by technology

The above method 1 ignores the thermal management of the power battery pack. Long-term high temperature will accelerate the performance aging of the battery, thereby affecting the economic performance of the vehicle.

In the second method, the threshold control strategy is commonly used for energy management. Although the structure is simple and stable, it cannot optimally allocate the energy of the power battery pack. After exceeding the threshold management range, the energy utilization rate of the power battery pack decreases.

In addition, the use of fuzzy control strategy can optimize the distribution of energy to a certain extent, but the formulation of fuzzy control rules is often affected by the designer's own subjective factors, which is prone to some loopholes. When the efficiency loss of system components changes during use Finally, the control strategy cannot adjust the corresponding rules and parameters in time, which is easy to cause local optimum, coupled with the long-term cumulative error of the system, so that the control effect is not ideal

In addition, in the fuzzy control strategy that considers dynamic programming optimization, conventional dynamic programming can only be optimized for specific driving conditions, and cannot consider all driving conditions of the car, and cannot run online in real time

Images