Scribd
Upload
198 views15 pages

NILM Using FHMM & CO

The document describes a research project on non-intrusive load monitoring (NILM) using factorial hidden Markov models and combinational optimization algorithms implemented in neural networks. It provides an overview of the NILM toolkit, Python libraries, algorithms, and data used including: NILMTK, Anaconda, Spyder IDE, the REDD dataset, neural networks, factorial hidden Markov models, and combinational optimization. The contents section outlines topics covered in the document like the code demo and results sections.

Uploaded by

66umer66
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
198 views15 pages

NILM Using FHMM & CO

The document describes a research project on non-intrusive load monitoring (NILM) using factorial hidden Markov models and combinational optimization algorithms implemented in neural networks. It provides an overview of the NILM toolkit, Python libraries, algorithms, and data used including: NILMTK, Anaconda, Spyder IDE, the REDD dataset, neural networks, factorial hidden Markov models, and combinational optimization. The contents section outlines topics covered in the document like the code demo and results sections.

Uploaded by

66umer66
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 15

Non Intrusive Load Monitoring using

Factorial Hidden Markov Model and


Combinational Optimization Algorithms
implemented in Neural Networks
Submitted by:
Engr. Umer Ehsan SP17-REE-011
Engr. Waleed Iqbal FA17-REE-021
Engr. Rashid Ashraf SP17-REE-021

Submitted to: Prof. Dr. Engr. Mujtaba Jaffery


Contents
• Introduction of Non Intrusive Load Monitoring
• NILM Toolkit (NILMTK)
• Anaconda Distribution
• Spyder IDE
• REDD Data Set
• Libraries in Python
• Neural Networks
• Factorial Hidden Markov Models (FHMM)
• Combinational Optimization (CO)
• Code Demo
• Results
Non Intrusive Load Monitoing (NILM)
• Process of estimating energy consumed by individual
appliances

• Estimated itemized energy bill

• One of the feature of modern smart meters

• Useful for creating appliance specific “energy awareness”

• Useful in Home Energy Management Systems (HEMS) and


Ambient Assisted Living (AAL).
NILM Toolkit (NILMTK)
• Open source toolkit

• Designed to enable comparison of energy disaggregation


algorithms

• Basic idea is to provide standard data for algorithm training


and then comparing their performance with the results of
bench mark algorithms.

• Recommended to be used with Anaconda Python Distribution


ANACONDA Distribution
• Open source Anaconda Distribution

• Comes with pre installed libraries and modules.

• Helps to create and manage different environments with


different libraries for different simultaneous projects.
Spyder (IDE)
• Integrated Development Environment designed for python
and data science research.

• Program dedicated for software development

• Offers several tools and features i.e. editor, console, explorer


etc.

• Offers key features like save and reload code files, run code
from within environment, automatic code formatting
REDD Data Set
• Residential Energy Disaggregation Data Set

• Open source data set

• Contains real data of six homes in USA.

• Contains main power data as well as individual appliance data

• Recorded at both low and high frequencies


Libraries in Python
• Libraries are group of standard functions

• Allows reusability and ease in programming.

• Syntax : “from library_name import function”

• Makes python one of the preferred choice of modern data


scientists.
Neural Networks
Factorial Hidden Markov Model (FHMM)
• Applicable to probabilistic models of time series data

• History is recorded in single multinomial variable (hidden


state) which can take on ‘K’ discrete values

• FHMM is combination of HMM and Cooperative Vector


Quantizer (CVQ)
Combinational Optimization
(CO)
• Optimal solution in finite solution set.

• Given set of instances I, finite set of feasible solutions f(x),


feasible solution y

• Goal is to find the optimal solution g


Code Demo
Results

You might also like