Underground Cable Fault Location Detection using Machine Learning

Project Title

Underground Cable Fault Location Detection using Machine Learning

Project Area of Specialization Electrical/Electronic EngineeringProject Summary

Since modern systems do not locate exact location of the fault in cable which results in economical and infrastructural loses.

For overcoming these shortcomings we  will design cheap system for finding exact location of the fault.

The project is combination of hardware and software.

• We will find the exact location of  fault in the cable.
• First we find distance of the fault from base station.
• Then the value of distance is fed to ML model.
• ML gives us GPS coordinates.
• Then with help of these coordinates we will show exact location of the fault on Google  map.

Project Objectives

• Since modern systems do not locate exact location of the fault in cable which results in economical and infrastructural loses. The proposed project will find the location of the fault with the help of Machine Learning techniques.

• For overcoming these shortcomings we  will design cheap system for finding exact location of the fault.

Project Implementation Method

The project is combination of hardware and software.

• We will find the exact location of  fault in the cable.
• First we find distance of the fault from base station.
• Then the value of distance is fed to ML model.
• ML gives us GPS coordinates.
• Then with help of these coordinates we will show exact location of the fault on Google  map.

• The faults created are manually by using switches.

• Cable has many types. Every cables has different resistance which mainly depends upon the material.

• The value of the resistance is depends upon the length of the cable wire.

• If any changes occur in the resistance, the value of the voltage will be changed that particular point is called Fault.

• Faults has many types & are given below.

1. Short Circuit Fault
2. Open Circuit Fault

• Current sensing part of cable represented as set of resistors & switches are used as fault creators to indicate the fault at each location.

• This part senses the change in current by sensing the voltage drop due change in length of the resistor.

• Next is controlling part which consists of internal ADC in ARDUINO which receives input from the current sensing circuit as AC input, converts this voltage into digital signal and then it feeds the microcontroller with that signal.

• The microcontroller also a part of the controlling unit and makes necessary calculations regarding the distance of the fault according to our program.

• Value of distance calculated by ARDUINO program is displayed on LCD and send to the server through ESP8266 module.

• In server side the value of distance is fed to ML model which generates GPS  coordinates.

• Then these coordinates are shown in the mobile app on googlemape.

• ML model is trained using the data which collected during laying down the cable.

  This data set consists of the following features.

  1.DISTANCE FROM BASE STATION

  2.GPS COORDINATES ASSOCIATED WITH EACH POINT ON THE CABLE.

• ML model is trained using linear regression

SOFTWARES USED.

1. Embedded C is used for programing ARDUINO using arduino IDE.

2. Python and it libraries for machine learning like

Scikit-learn

 Pandas

Matplotlib

Seaborn

3. JAVA for creating android app using android studio.

4 XML for designing mobile layout.

Benefits of the Project

Followimg are the advantages of underground cables

1. Suitable for congested urban areas.
2. Require low maintenance as damage rate is low.
3. Ensure small voltage drops.
4. Not easy to steal and damage.
5. Avoid the chances of illegal connections.
6. Protection from environmental stresses like wind, storms, and thunder.

But one of the major challenges in underground cables is to detect the location of the fault.

The major benifit of the project is to locate the fault and avoid the digging and manually finding the fault.

Technical Details of Final Deliverable

Following is the list of equipments used in the project

• GSM Module SIM900a
• Arduino Mega
• 4 Channel Relay Module
• Bread Board
• Resistor packet
• TFT LCD
• AC power suppply
• Regulators
• Transformer

The circuit diagram is as follow

SOFTWARES USED.

1. Embedded C is used for programing ARDUINO using arduino IDE.

2. Python and it libraries for machine learning like

Scikit-learn

 Pandas

Matplotlib

Seaborn

3. JAVA for creating android app using android studio.

4 XML for designing mobile layout.

Final Deliverable of the Project HW/SW integrated systemCore Industry Energy Other IndustriesCore Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Industry, Innovation and Infrastructure, Sustainable Cities and Communities, Climate ActionRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	57700
GSM Module SIM900a	Equipment	8	2500	20000
Arduino Mega	Equipment	8	1500	12000
4 Channel Relay Module	Equipment	16	400	6400
Bread Board	Equipment	20	250	5000
TFT LCD	Equipment	7	1000	7000
Power Supply unit for circuit operation	Equipment	4	500	2000
transformer	Equipment	5	900	4500
Regulator	Equipment	4	200	800