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.