Deep Learning based Multi-fault detection system for 3-Phase Induction Motor

Project Title

Deep Learning based Multi-fault detection system for 3-Phase Induction Motor

Project Area of Specialization Artificial IntelligenceProject Summary

The three phase induction motors are widely used among industry. Three Phase induction motors are preferred over others motors due to its rugged construction, low cost, and good operating characteristics. However, they pose different faults which ultimately lead to major failure, economical loss, and even threat to human safety.  Thus, we are going to build a condition monitoring system that can effectively diagnose various faults in three phase induction motor before it could cause any problem. More precisely, deep learning algorithms will be used to diagnose various faults in motor with interface of hardware board myRIO. The myRIO board will provide faster data processing as it is almost ten times faster than the standard micro-controllers and microprocessor owing to FPGA capabilities in it. Furthermore, deep learning has been in limelight owing to its capabilities such as high accuracy, upgradability, and feature learning even from raw input data.

For data acquisition, current and temperature sensors will be used for data to acquire information related to condition of induction motor. The acquired data will be stored and used for training and testing of deep learning models. In last, the developed system will be tested to detect various fault such as inter-turn short faults, supply imbalance faults and broken rotor bar(BRB) faults. The material required for this project include two 0.5 HP 3-phase squirrel cage induction motors, clamp type current sensors, LM35 Temperature sensors and myRIO 1900 development board.

Compared to the previous work, Here, we are proposing an effective and non-invasive fault detection system of induction motors with high performance processing unit. In this project, we will utilize myRIO as a solution for automated condition monitoring of industrial motors.  The ultimate goal of the project is to design a system that will be capable enough to diagnose three phase induction motor faults as precaution with high performance.

Figure 1. Block diagram of the proposed 3-Phase motor condition monitoring system.

Project Objectives

The main objectives of our project is to effectively and non-invasively detect faults of 3-phase Induction motor through deep learning algorithms. This system will diagnose faults in real-time.

The key project objectives are:

1. To develop a data acquisition system using current and temperature sensors.
2. Implementation of deep learning models for fault diagnosis.
3. System Integration using LabVIEW, Python and myRIO development board.
4. Testing of developed system and performance comparison of  deep learning models.

Project Implementation Method

Our project is based on deep learning techniques with high performance data processing unit for effective fault diagnosis of induction motors. Our project will focus on different electrical and mechanical fault of induction motor like broken rotor bar (BRB) faults, ground faults, inter turn short faults and crack rotor faults.

For the data acquisition of current and temperature, the induction motor will run under unhealthy condition. Motor conditions will be monitored by clamp type current sensors and LM35 temperature sensor.

After that the analog output of sensors will be fed to myRIO 1900 for digital data conversion and storage into computer. In this project, we will use myRIO as a solution for faster data acquisition and automated monitoring of industrial motors. Then, the stored data signals would be used for training of deep learning models. Here in our project, we will use Python programming language and its libraries. Then the python will be integrated with LabVIEW software which will provide data acquisition facility and graphical user interface to myRIO. In last, the developed system will be tested to detect various faults of induction motor.

 Figure 2. Workflow diagram of final prototype

Benefits of the Project

The more we learn about this topic, the more we appreciate the benefits of it. There are bundle of applications of condition monitoring of motors like:

• it allows one to predict possible malfunction and to schedule the motor shutdown if a prefixed rate is exceeded.
• Fault diagnosis with full automation.
• Its compact architecture configuration makes it suitable for real time detection and monitoring.
• It has capability to provide efficient training of the classifier with limited size of training dataset.
• The most appealing feature of deep learning is that it can be used for online diagnostics.
• Its accuracy is not limited to any certain point as it increases with each sample.
• It is cost effective and practical in real time hardware implementation.

Technical Details of Final Deliverable

This project is comprising of LABVIEW software and myRIO embedded processing unit for detecting the faults diagnosis by using deep learning methods. Here, we are going to design a system that would be comprising of a myRIO embedded device that will be used for detection of major faults such as broken rotor bar (BRB) faults and supply imbalance faults in a 3-phase squirrel cage induction motor. The existence of these faults in induction motors can be detected by monitoring condition motor using various signals such as current and temperature.  Deep learning methods can detect these faults at an early stage even from raw input data and thus assists avoid major damage and complete failure of the motor.

The development system would be an end-to-end fault detection system which can detect major faults of induction motor at an early stage. This system is designed to be used in industrial applications for detection of fault in the motor .

Final Deliverable of the Project HW/SW integrated systemCore Industry Energy Other Industries Transportation Core Technology Artificial Intelligence(AI)Other Technologies OthersSustainable Development Goals Industry, Innovation and Infrastructure, Responsible Consumption and ProductionRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	67739
YHDC SCT013 Current clamp	Equipment	12	850	10200
LM35 Temperature Sensor	Equipment	12	120	1440
0.5 HP 3-phase squirrel cage induction motor	Equipment	2	12500	25000
Gigabyte Radeon RX 560 4GB 128-Bit GDDR5 ATX Graphic Card (GV-RX560OC-	Equipment	1	21799	21799
Connecting vires	Miscellaneous	30	10	300
Thesis	Miscellaneous	1800	5	9000