Electrical Machine Condition Monitoring With Edge KPI Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing condition monitoring devices for electrical machines face limitations in processing power and memory, restricting the use of advanced analytics, and in remote analysis, where bulk data transfer reduces battery life and consumes power.
Innovation Solution
Implementing an edge computing approach with a condition monitoring device equipped with sensors, a data logger, memory, and a communication unit, which computes intermediate key performance indicators (KPIs) locally, reducing data dimensionality and enabling efficient data transmission to a remote server for further analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If data is transferred from local end to cloud for condition assessment, then advanced analytics capability is improved, but battery consumption increases and operating life is reduced
Solution Approach 1:
The system segments the data processing workflow into two parts: local preprocessing of raw sensor data into condensed condition assessment results at the device end, and remote analysis of these condensed results at the cloud. This segmentation reduces the volume of data transferred and processed remotely, thereby reducing battery consumption while preserving advanced analytics capability.
Solution Approach 2:
The system performs preliminary processing of raw sensor data locally before transmission, converting it into condensed condition assessment results. This preliminary action at the device end reduces the data volume that needs to be transmitted and processed remotely, optimizing battery usage while enabling advanced analytics at the cloud.
2Loss of energy
If computational processing is performed at local condition monitoring device, then data transfer to cloud is reduced, but processing speed is limited by device processor capabilities
Solution Approach 1:
The system segments computational tasks by processing type: resource-intensive raw data preprocessing is performed locally to reduce transfer volume, while computationally intensive advanced analytics are performed remotely at the cloud with sufficient processing power. This segmentation optimizes both energy efficiency and processing speed.
Solution Approach 2:
The system introduces an intermediary processing layer that transforms raw sensor data into condensed condition assessment results locally. This intermediary representation serves as a bridge, reducing data transfer requirements while preserving the information needed for advanced remote analytics, thus balancing local processing limitations with cloud computing power.
3Measurement precision
If bulk raw data is transferred to remote server, then comprehensive analysis is enabled, but network bandwidth consumption and communication power increase
Solution Approach 1:
The system extracts essential condition assessment information from raw sensor data locally before transmission. By taking out only the critical features and condensing the data into meaningful condition metrics, the system reduces data transfer volume while preserving the information necessary for accurate remote analysis.
Solution Approach 2:
The system performs preliminary data condensation and feature extraction locally, transforming bulk raw data into compact condition assessment results before transmission. This preliminary action reduces network bandwidth consumption and communication power while maintaining the information quality needed for comprehensive remote analysis.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This method allows for real-time monitoring and reduced data transfer, optimizing battery life and network usage while enabling advanced analytics for predicting faults and scheduling maintenance, thereby improving the health assessment of electrical machines.
Implementation Method 1
sensors to detect magnetic fields, currents, vibrations, temperatures etc.
Implementation Method 2
sensors to detect magnetic fields, currents, vibrations, temperatures etc.
Data Source
AI summary
The present invention relates to a condition monitoring device and method for monitoring an electrical machine. The method includes obtaining, at periodic instants, measurements from sensors of the condition monitoring device, where each sensor is one of a magnetometer and an accelerometer. The method also includes comparing, for one or more instants, amplitude data of the measurements with condition monitoring data, wherein the comparison is performed for the amplitude data in one or more axes and at one or more frequencies. The condition monitoring data includes a relation between a plurality of parameters, a plurality of conditions and a plurality of frequencies. The method additionally includes detecting a condition and at least one parameter associated with the condition, based on the comparison. According to the detection, the method includes utilizing the measurements of the at least one parameter for determining a health condition of the electrical machine.