Transformer Health Monitoring

The need t? m?nit?r c?nti?n?usly the transf?rmers c?nditi?ns while in ?perati?n is crucial f?r utilities due t? the imp?rtance ?f the electrical equipment present in the Grid. F?r this purp?se, a system is being de?el?ped that pr??ides transf?rmer health m?nit?ring system f?r better clas

Project Title

Transformer Health Monitoring

Project Area of Specialization

Internet of Things

Project Summary

The need t? m?nit?r c?nti?n?usly the transf?rmers c?nditi?ns while in ?perati?n
is crucial f?r utilities due t? the imp?rtance ?f the electrical equipment present
in the Grid. F?r this purp?se, a system is being de?el?ped that pr??ides
transf?rmer health m?nit?ring system f?r better classificati?n ?f faults in
transf?rmer using machine learning techniques (PCA). The System is built ?n the
t?p?l?gy ?f Arduin? and ?nline m?nit?ring ?f the system health is ?n?wn thr?ugh
GSM m?dule. This technique will ser?e the purp?se ?f transf?rmer health
m?nit?ring system and shall pr??e t? be an alternate meth?d ?f DGA testing that
is being currently used in the transf?rmer industry f?r the same purp?se.
C?nsiderable research has been d?ne ?nly ?n current signature analysis but ?ur
technique in??l?es current, ??ltage, ?ibrati?n and temperature signature analysis.
By using di?erse library data set, the system shall be able t? estimate transf?rmer
life expectancy. Additi?nally, it als? pr?mpts the user when the system needs any
pre?enti?e measures t? be ta?en with regard t? the rectificati?n and
maintenance w?r?s, a??iding the hea?y c?st ?f brea?d?wn maintenance. The
transf?rmer health m?nit?ring system will gi?e transf?rmer its necessary safety
and pr?tecti?n mechanism and at the same time is an impressi?e c?st-effecti?e
m?dule f?r the transf?rmer unit. This meth?d helps us in ad?ancing t?wards a
safer, reliable and smart electrical netw?r? system which will supp?rt in
transf?rming ?ur li?es t? a better state.
?eyw?rds- PRINCIPLE C?MP?NENET ANALYSIS, Arduin?, MAINTENANCE C?ST EFFECTI?ENESS TRANSF?RMER
HEALTH M?NIT?RING SYSTEM

Project Objectives

Under the scope of the proposed project, a system is being developed which performs the condition-based monitoring of electrical distribution transformer by finding and analyzing various, specific and unique signatures by parameters like vibration, temperature, current, and voltage. The system shall be able to indicate the faults occurring in the distribution transformer by adopting a technique based on intrusive parameters. Intrusive methodology for the classification of electrical faults in transformers. Development of a software tool for the detection of faults in transformers. Condition based monitoring of the system. 

Project Implementation Method

The methodological approach is based upon the comparative study of vibration, temperature, voltage, and current. Parameters of the transformer between the gathered reading of healthy and faulty readings database through real-time data acquisition set up that constitute multiple sensor package signal conditioning board and analog to digital converter(NI-DAQ Card). This process is broadly divided into three steps. In step number one, frequency and harmonic content are recorded and to be used as a reference for the healthy state of the transformer. In step number two, various faults like, line to line(LL), line to ground(LG), Double line to ground(LLG), three-phase fault(LLL), and in-rush current fault is introduced in transformers and again the frequency and harmonic content for each particular in each transformer to be used as reference point for the faulty transformer condition. In step number three, feature extraction is done for getting the unique signatures of each fault to be defined and every signature to be stored in the system to check the presence of fault using these signatures once it will be fed with live data incoming from the transformer.

Data acquisition with MATLAB starts by importing data into it. Then, applying a wavelet transform that will give the unique features of every dataset. After that PCA (Principal Component Analysis) has been applied that will investigate and analyze the type of fault that is present in the system, then we merge all the datasets in one excel file, then classifier (KNN) is applied on group data that give us the accuracy of our dataset in the form of confusion matrix.

Now in the testing phase, the data from the transformer, through the sensor package, will reach the DAQ card where it will be converted to digital form and will feed this data to Arduino where it will be compared with different parts of the dataset. The data which will match with the dataset will correspond to a certain result which will be displayed onto the mobile through the GSM module in the form of the message.   
Arduino Mega is used as a comparator for comparing real-time data with the dataset, and GSM Module 900A is used to display the message on mobile.

Benefits of the Project

It provides real-time monitoring of the health of the transformer. Transforms our grid system to a smart grid system which is the need of the day. minimizes our financial losses in our electrical system. 

Also a cost-effective system with a minimum cost. Operation is quick having mobile notification gives a quick response about any fault also by using PCA it provides a future prediction that how much longer it can run or require maintenance in what time.

cost of electricity can also be reduced such as by reducing line losses and other maintenance charges as well. it is an alternative of DGA system. our system will be given to factories for factory level product improvement.

Technical Details of Final Deliverable

The technical details of this project are as follows.

For data acquisition purpose NI LabVIEW circuitry is being used in coordination with NI DAQ card. For signal analysis wavelet transform is being used, Then for the system training purpose, PCA is used which shows the accuracy of our training dataset. At last, when the system is trained a new signal is fed whom the response is seen by a comparative analysis of this signal with predefined data set and notification on mobile will appear after 2 seconds.  

Final Deliverable of the Project

Hardware System

Type of Industry

Energy

Technologies

Internet of Things (IoT)

Sustainable Development Goals

Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com