Application of industrial Internet of things technology in fault diagnosis of food machinery equipment based on neural network

Due to the possibility of equipment failure that leads to aberrant processes and resource loss, smart manufacturing and Internet of things (IoT) devices require the highest level of security and configuration for industrial automation systems, e.g., food sector. However, customizing active learning for each parameter value in the IoT-driven industrial environment is not easy due to the continually changing variables of industrial control machinery. Therefore, this paper focuses on neural network to improve the efficacy of fault diagnosis in food machinery and equipment by addressing poor generalization ability and low fault recognition rates. The paper presents a fault signal recognition method based on external correction and periodically correcting mechanical equipment's fault features to extract signal features of equipment faults. Furthermore, an enhanced migration convolutional neural network is proposed for equipment fault diagnosis, establishing a convolutional neural network and two data classifiers to train fault data in the network source domain. The data classification loss and discriminant classifier functions minimize the feature distribution gap between the source and target domains. The experimental results show that the proposed method effectively enhances mechanical equipment's fault diagnosis and the safety of equipment operation to exhibit robust generalization ability.