ST3630A SEAGATE

Cogging force compensation is a fully automated application based on TwinCAT machine learning.The software independently records the cogging data in the customer application as part of a reference run over the entire linear motor track length.Using the data obtained in this way,the software trains the neural network,which will eventually be integrated into the control system for current pre-control.By adjusting the current pre-control in this way,the tracking error can be reduced by up to 7 times and the synchronization performance of the device can be improved by 5 times without any changes to the AL8000 hardware.

齿槽力补偿是基于TwinCAT机器学习的全自动化应用。软件独立记录客户应用中的齿槽数据，作为整个直线电机磁轨长度上参考运行的一部分。软件通过以这种方式获得的数据，训练神经网络，这个神经网络最终会被集成到控制系统中，用于电流预控。通过种方式调节电流预控后，跟踪误差最高可减少7倍，设备的同步性能可提升5倍，而无需对AL8000硬件进行任何改动。