Non Intrusive Load Monitoring System for Household Appliances using Deep Learning
A method of Non-Intrusive load monitoring using a Convolutional Neural Network (CNN) is proposed which will enable the consumers to monitor from a single point what appliances are ON in the whole house at any given time. The proposed method requires only the aggregated current and voltage values of
2025-06-28 16:34:16 - Adil Khan
Non Intrusive Load Monitoring System for Household Appliances using Deep Learning
Project Area of Specialization Electrical/Electronic EngineeringProject SummaryA method of Non-Intrusive load monitoring using a Convolutional Neural Network (CNN) is proposed which will enable the consumers to monitor from a single point what appliances are ON in the whole house at any given time. The proposed method requires only the aggregated current and voltage values of the appliances. The Voltage-Current (VI) trajectories of these appliances are input to the neural network, which then classifies the appliances based on their VI trajectories.
Project ObjectivesThe main objectives we need to achieve to make this project are as follows:
- Hardware construction to read the current and voltage in real-time.
- Measurement and recording of aggregated and appliance level data.
- Pre-processing of aggregated and appliance level data such as filtering the circuit noise, feature extraction, normalization, etc.
- Disaggregation of aggregated electricity consumption into individual appliance level consumption.
- Real-time identification of the appliance(s) being used.
- Restricting the usage of the appliance(s) according to the user in user-specified slots.
- Informing the user about the statistical analysis of the recorded data and appliances being used at the instance along with the Power Factor.
A NILM system consists of three main components:
- Data Acquisition
- Feature Extraction
- Classification
First, acquire the voltage and current values of different appliances from the voltage (ZMPT101B) and current (ACS712) sensors. Power Analyzer is used for the benchmarking of results so that our current and voltage values from sensors gets equal to the power analyzer values. If some error occurs, then it will be compensating through the microcontroller.
After getting the values of voltage and current, record these values. Repeat this scenario for all the appliances and form a data set for all the appliances.
Then aggregated load will be measured and recorded using current and voltage sensors. And Non-Intrusive Load Monitoring system will disaggregate this data into individual appliances for load monitoring, and display power factor, active power, reactive power, and total harmonic distortion for each appliance.
Benefits of the ProjectThe benefits of the NILM system are as follows:
- The NILM based system consists of sensors and modules which require a small area for integration and installation.
- The NILM system can easily be integrated into an existing electricity network.
- Data obtained from the NILM base system is not only utilized to calculate and monitor the appliance functionality but is also used as feedback to train the system for system training hence data analytic techniques are utilized to identify problems such as when an appliance stops working or when the appliance in question shows rogue behavior.
- NILM system can disaggregate the consumed energy using one meter rather than installing sensors to each appliance.
- Reduce installation complexity and cost.
- The system is user-friendly as the user just has to deal with the Graphical user interface to restrict appliances.
- NILM system has a positive impact on the environment not only by conserving electricity but also by making use of the eco-friendly building.
The final deliverables of the project are as follows:
- Monitoring and informing the user about the statistical analysis of the appliances being used at the instance.
- Real-time detection and classification of the appliances being used.
- Disaggregation of aggregated power profile into appliance level power consumption.
- Restricting the usage of appliances during user-specified hours.
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 6560 | |||
| ESP32-S Wifi + BT SoC Module | Equipment | 1 | 1300 | 1300 |
| Current Sensor (ACS712) | Equipment | 1 | 280 | 280 |
| Voltage Sensor (ZMPT101B) | Equipment | 1 | 350 | 350 |
| 4 Channel Relay | Equipment | 1 | 350 | 350 |
| Drawing Board | Miscellaneous | 1 | 600 | 600 |
| Wires | Miscellaneous | 1 | 300 | 300 |
| LED Bulb 12W | Equipment | 1 | 180 | 180 |
| 12V FAN | Equipment | 1 | 150 | 150 |
| Switch Board | Miscellaneous | 2 | 500 | 1000 |
| Reports and Thesis Printing | Miscellaneous | 1 | 2000 | 2000 |
| Bulb Holder | Miscellaneous | 1 | 50 | 50 |