The weld pool shape control by intelligent strategy was studied. In order to improve the ability of self-learning and self-adaptation of the ordinary fuzzy control, a self-learning fuzzy neural network controller (FNN...The weld pool shape control by intelligent strategy was studied. In order to improve the ability of self-learning and self-adaptation of the ordinary fuzzy control, a self-learning fuzzy neural network controller (FNNC) for backside width of weld pool in pulsed gas tungsten arc welding (GTAW) with wire filler was designed. In FNNC, the fuzzy system was expressed by an equivalence neural network, the membership functions and inference rulers were decided through the learning of the neural network. Then, the FNNC control arithmetic was analyzed, simulating experiment was done, and the validating experiments on varied heat sink workpiece and varied gap workpiece were implemented. The maximum error between the real value and the given one was 0.39mm, the mean error was 0.014mm, and the root-mean-square was 0.14mm. The real backside width was maintained around the given value. The results show that the self-learning fuzzy neural network control strategy can achieve a perfect control effect under different set values and conditions, and is suitable for the welding process with the varied structure and coefficients of control model.展开更多
This paper describes a modified speed-sensorless control for induction motor (IM) based on space vector pulse width modulation and neural network. An Elman ANN method to identify the IM speed is proposed, with IM para...This paper describes a modified speed-sensorless control for induction motor (IM) based on space vector pulse width modulation and neural network. An Elman ANN method to identify the IM speed is proposed, with IM parameters employed as associated elements. The BP algorithm is used to provide an adaptive estimation of the motor speed. The effectiveness of the proposed method is verified by simulation results. The implementation on TMS320F240 fixed DSP is provided.展开更多
Realizing of weld penetration control in gas tungsten arc welding requires establishment of a model describing the relationship between the front-side geometrical parameters of weld pool and the back-side weld width w...Realizing of weld penetration control in gas tungsten arc welding requires establishment of a model describing the relationship between the front-side geometrical parameters of weld pool and the back-side weld width with sufficient accuracy. A neural network model is developed to attain this aim. Welding experiments are conducted to obtain the training data set （including 973 groups of geometrical parameters of the weld pool and back-side weld width） and the verifying data set （108 groups）. Two data sets are used for training and verifying the neural network, respectively. The testing results show that the model has sufficient accuracy and can meet the requirements of weld penetration control.展开更多
提出了通过视觉传感获取焊接过程中的焊接特征信息并利用神经网络模型预测焊缝背面宽度的方法。利用大功率盘形激光器焊接了低碳钢SS400焊件,在焊接过程中改变焊接功率、焊接速度和焊接路径,并利用两台高速摄像机同步获取焊件正面和侧...提出了通过视觉传感获取焊接过程中的焊接特征信息并利用神经网络模型预测焊缝背面宽度的方法。利用大功率盘形激光器焊接了低碳钢SS400焊件,在焊接过程中改变焊接功率、焊接速度和焊接路径,并利用两台高速摄像机同步获取焊件正面和侧面出现的焊接特征信息。对获取的图像进行色彩空间转换、分层、滤波去噪和空域图像处理,提取飞溅、熔池和金属蒸气等焊接特征信息,观察焊接路径对各个特征的影响。最后,建立了一个三层的LMBP(LevenbergMarquardt Back Propagation)神经网络模型,将提取的特征信息作为输入量,预测焊缝的背面宽度。结果显示:当熔透不稳定或出现未熔透状态时,LMBP神经网络拟合度大于0.83,最大训练误差均值为0.002 8mm,最大实际误差均值为0.225 6mm。试验结果表明所建立的预测模型具有良好的准确性和稳定性。展开更多
文摘The weld pool shape control by intelligent strategy was studied. In order to improve the ability of self-learning and self-adaptation of the ordinary fuzzy control, a self-learning fuzzy neural network controller (FNNC) for backside width of weld pool in pulsed gas tungsten arc welding (GTAW) with wire filler was designed. In FNNC, the fuzzy system was expressed by an equivalence neural network, the membership functions and inference rulers were decided through the learning of the neural network. Then, the FNNC control arithmetic was analyzed, simulating experiment was done, and the validating experiments on varied heat sink workpiece and varied gap workpiece were implemented. The maximum error between the real value and the given one was 0.39mm, the mean error was 0.014mm, and the root-mean-square was 0.14mm. The real backside width was maintained around the given value. The results show that the self-learning fuzzy neural network control strategy can achieve a perfect control effect under different set values and conditions, and is suitable for the welding process with the varied structure and coefficients of control model.
基金This project was supported by the National Natural Science Foundation of China (No. 69874086).
文摘This paper describes a modified speed-sensorless control for induction motor (IM) based on space vector pulse width modulation and neural network. An Elman ANN method to identify the IM speed is proposed, with IM parameters employed as associated elements. The BP algorithm is used to provide an adaptive estimation of the motor speed. The effectiveness of the proposed method is verified by simulation results. The implementation on TMS320F240 fixed DSP is provided.
基金the Shandong Provincial Natural Science Foundation of China （No. Z2003F05 ）.
文摘Realizing of weld penetration control in gas tungsten arc welding requires establishment of a model describing the relationship between the front-side geometrical parameters of weld pool and the back-side weld width with sufficient accuracy. A neural network model is developed to attain this aim. Welding experiments are conducted to obtain the training data set （including 973 groups of geometrical parameters of the weld pool and back-side weld width） and the verifying data set （108 groups）. Two data sets are used for training and verifying the neural network, respectively. The testing results show that the model has sufficient accuracy and can meet the requirements of weld penetration control.
文摘提出了通过视觉传感获取焊接过程中的焊接特征信息并利用神经网络模型预测焊缝背面宽度的方法。利用大功率盘形激光器焊接了低碳钢SS400焊件,在焊接过程中改变焊接功率、焊接速度和焊接路径,并利用两台高速摄像机同步获取焊件正面和侧面出现的焊接特征信息。对获取的图像进行色彩空间转换、分层、滤波去噪和空域图像处理,提取飞溅、熔池和金属蒸气等焊接特征信息,观察焊接路径对各个特征的影响。最后,建立了一个三层的LMBP(LevenbergMarquardt Back Propagation)神经网络模型,将提取的特征信息作为输入量,预测焊缝的背面宽度。结果显示:当熔透不稳定或出现未熔透状态时,LMBP神经网络拟合度大于0.83,最大训练误差均值为0.002 8mm,最大实际误差均值为0.225 6mm。试验结果表明所建立的预测模型具有良好的准确性和稳定性。
