基于神经网络的地铁牵引整流器直接功率控制算法研究

Direct Power Control of Subway PWM Rectifier Based on Neural Network

现代地铁交通车辆短间隔内重复起动停车、加速减速,车辆需要的电流处于不断变化的过程,难以建立复杂的精确数学模型,且对系统在短时间内抗负载突变的能力有较高要求。将直接功率控制理论和神经网络控制结合起来,应用于地铁整流器控制,通过建模仿真可以看出此系统动态响应较快,抗负载突变能力强,仿真结果验证了此控制方法的正确性与可行性。

Subway vehicles often meet the conditions of acceleration, deceleration, start and stop in short time. It is difficult to establish an accurate model for changing current of vehicle, which needs a high ability to resist load mutation. This paper presents a direct power control method based on neural network theory, and applies it to the PWM rectifier control system in subway. This system has high respond speed and strong ability of anti-load interference, and the simulation results show that the proposed method is effective.