基于神经网络控制的快速充电方法探究

Research of fast-charging method based on neural network control

随着蓄电池的广泛应用,对其充电方法的研究也不断深入,并取得了较多的科研成果,但是快速充电技术仍受到多方面的制约。提出了以神经网络算法和传统控制器相结合的智能充电技术,以BP网络算法训练神经网络。由于神经网络的自适应性,控制过程不依赖于模型,因此能对复杂的非线性、不确知的充电过程进行实时修正,并将此种方法应用于阶段充电法,通过仿真对比,证明了智能充电的有效性。

With the extensive use of storage battery, more and more deeply research of its charging methods were acquired up. Many scientific research achievements were achieved. But fast charging technology was still restricted by many aspects. An intelligent charging technology by combine Neural Network Algorithm with the traditional controller was put forward. Because NNC didn＇t depends on the mathematical models, it could do the real-time correction in the complicated, nonlinear uncertain charging process, so this new method was put into section of charging. Via compare the simulation results, the efficient, rapid and non-loss battery charging were realized.