vHow to find the best maintenance balance point between the potential failure point and the functional failure point is the core demand of predictive maintenance in enterprises. Based on deep technical accumulation and rich experience in fault diagnosis in the field of industrial automation, Schneider Electric predictive maintenance solutions introduce advanced AI algorithms and preset diagnosis knowledge base. We provide comprehensive solutions, including AVEVA Predictive Analytics Predictive maintenance system and EcoStruxure™ PMA Predictive Maintenance Consultant, to help industry users tap into the value of data and realize intelligent equipment operation and maintenance.
On the one hand, in response to the needs of large industrial enterprises to strengthen data governance and release the value of data, Schneider Electric’s AVEVA Predictive Analytics predictive maintenance system can build a normal operating state model of equipment through big data modeling of historical data and machine learning algorithms. After that, the established model is connected to the real-time data of large key units, continuous monitoring of equipment status, timely capture of early signs of equipment failure, early intervention diagnosis to take preventive measures, to help enterprises intelligently identify equipment status and changes in the external environment to achieve condition desensitization, reduce the false positive rate, and improve the overall operational efficiency.
On the other hand, conventional dynamic equipment such as water pumpA06B-6117-H104 s, gear boxes, and air compressors generally lack status monitoring systems, and enterprises urgently need to upgrade such equipment to break the data “island”. In this regard, EcoStruxure PMA predictive maintenance consultants obtain vibration data through intelligent sensors and edge computing gateways, and can combine the vibration mechanism of the equipment with industry know-how, mathematical models and AI algorithms to make real-time prediction and diagnosis of equipment failures, and guide users to intelligent operation and maintenance. To be specific:
Vibration temperature integrated intelligent sensor: Adopts industry-leading MEMS sensing technology, capable of three-axis (axial/radial horizontal/radial vertical) high-frequency sampling, and monitoring equipment surface temperature, can capture more than 20,000 equipment vibration data points per second, its built-in AI intelligent algorithm supports data preprocessing, can provide up to 100 kinds of characteristic value output. The body adopts precision flat and small design, suitable for installation in small space, can be close to the equipment axial vibration source, improve the accuracy of data identification; In addition, the products comply with CE certification, IP67 protection grade, and explosion-proof ExiaIICT4Ga standards, which can effectively cope with harsh working conditions.
Edge computing gateway: supports 6 vibration sensing channels, can monitor oil temperature and motor three-phase winding temperature, and embedded intelligent algorithm model for fault prediction, has high-performance edge AI computing capability, can achieve edge cloud collaboration, automatically determine equipment opeA06B-6117-H104 rating conditions and issue threshold alarm. The overall precision of the fuselage supports internal/external antennas, has CE certification and IP67 protection rating, and can be installed indoors and outdoors or in harsh working conditions.
Predictive maintenance software system: The “mathematical + mechanism model” equipment fault prediction and diagnosis system based on process data and vibration data, combined with the process mathematical model fault diagnosis tool, can build different models according to the process, and accurately locate the equipment fault location, to help users diagnose equipment faults or electrical faults caused by process changes, as well as complex mechanical aging and wear problems. Among them, the two core algorithms based on the mechanism analysis and mathematical analysis of the original vibration data have the functions of time domain analysis, spectrum analysis, frequency component analysis, degradation/trend analysis, etc. Big data analysis algorithms such as correlation analysis and cluster analysis can be used to find out the data rules of the process parameters and vibration data to improve the accuracy of model prediction. The embedded AI machine learning in the system supports the self-learning of device fault diagnosis model, and can also conduct self-iterative analysis of fault characteristic values to assist mechanism model diagnosis.
