【摘要】 螺杆点胶阀因点胶效率高、精度高、胶点均匀性高等优点，广泛应用于微电子封装等领域，而电机转速的控制性能是决定螺杆点胶阀点胶质量的重要因素。针对点胶阀在外部干扰和负载扰动情况下，PID速度控制效果不理想，参数难以整定的问题，采用BP神经网络-无模型自适应控制(BP-MFAC)算法，通过BP神经网络在线整定无模型自适应控制器参数来实现对电机速度的自适应控制，无需人工整定参数。仿真和实验对比结果表明，该算法相较于PID控制算法，在电机速度控制上具有更小的超调量和稳态误差，更短的调节时间；在外部干扰和负载扰动的情况下，具有更好的抗扰动能力。更多还原

【Abstract】 The screw dispensing valve has high dispensing efficiency, high precision and high uniformity, is widely used in microelectronic packaging and other fields. Motor speed control performance importantly influences screw valve dispensing quality. Under external interference and load disturbance, PID controller effect of screw dispensing valve is unsatisfactory. And the PID controller parameters are difficult to be adjusted. To solve these problems, a BP neural network-model free adaptive control(BP-MFAC) algorithm is proposed. It adopts BP neural network to adjust the model free adaptive controller parameters online. The algorithm achieves adaptive control of motor speed, without manual parameter tuning. The simulate and experimental results show that compared with PID control algorithm, this algorithm has smaller overshoot and steady-state error, less setting time in motor speed control. It has better anti-disturbance ability under external disturbance and load disturbance.更多还原