A project report submitted in partial fulfillment of the requirements for the award of the Bachelor of Engineering. Honors) Electrical and Electronic Engineering. I certify that this project report entitled "IoT-based Smart Single Wall Outlet" has been prepared by Phang Yee Ren and has met the required standard of submission in partial fulfillment of the requirements for the award of the Bachelor of Engineering (Honours) Electrical and Electronic Engineering at Universiti Tunku Abdul Rahman. The copyright of this report belongs to the author under the terms of the Copyright Act 1987 as qualified by the Intellectual Property Policy of Universiti Tunku Abdul Rahman.
Appropriate acknowledgment should always be given for the use of any material contained in or derived from this report. To control demand, Tenaga Nasional Berhad (TNB) introduced the Enhanced Time of Use (ETOU) scheme to offer different rates to commercial and industrial end-users at different times of the day.
General Introduction
Problem Statement
ETOU has become a self-regulatory tool for customers to manage their electricity consumption and bills. Research shows that regularly providing data on electricity consumption patterns can successfully reduce the user's habit of wasting electricity. The meter will calculate the energy consumption and provide a sample of the daily electricity consumption.
Aim and Objectives
According to D. Vine et al. 2013), providing energy consumption information can change appliance usage patterns and reduce energy consumption by up to 20%. AMI can use the companion app, the myTNB app, to provide electricity consumption at up to 30 minute intervals (Poovenraj Kanagaraj, 2021).
Scope and Limitation of the Study
However, the push notification that reminds customers of their monthly usage is currently not applicable in the app. The data sent by each outlet is received and processed by the central console, which is then stored in the cloud. Meanwhile, the Android application accesses the cloud and displays the data in the user interface.
Introduction
Power Monitoring
Then a server based on Node JS language is set up and connected to the energy meter. A database based on MongoDB is created so that users can retrieve data from the database and design applications in the form of an Application Programming Interface (API). The presence of a human is identified and all the appliances connected to the electromechanical relay can be controlled based on the existence of humans.
Remote Controlling
After receiving the commands, the Raspberry Pi GPIO pins will respond with the command and control to and from devices.
Smart Wall Outlet in the Market
Amazon Smart Plug
With this function, electrical energy can be saved by automatically switching off the devices (Insider, 2020).
Meross Smart Plug
Xiaomi Smart Plug
Introduction
System Architecture
- Wall Outlets
- Central Console (Server)
- Android Smartphone Application
- Machine Learning and Electricity Consumption Benchmark Datasets
Wall sockets have three main functions, which are: monitoring the state of the wall socket, monitoring the power consumption of devices connected to the wall socket and communicating with the central console. The ESP32 will integrate with the ACS 712 current sensor to measure the current consumption of devices connected to a wall outlet. An optocoupler and a TRIAC are used to control the state of the wall outlet after receiving a signal from the ESP 32.
By implementing a browser-based flow editor called Node-Red, the server can form a network connecting multiple outlets. In this case, the server acts as an MQTT broker and the outlets act as MQTT clients. The server also functions to upload data from the outlets to the Firebase, allowing users to monitor the data remotely through a phone application.
Machine learning is applied to predict the daily power consumption of appliances connected to a single wall outlet. Electricity consumption benchmark datasets from the Australian government were chosen to do the machine learning. The dataset was obtained from the Department of Industry, Science, Energy and Resources of the Australian Government (Department of Industry, 2020).
The reason is that Amazon SageMaker Canvas is an industry standard service that enables business analysts to create highly accurate machine learning predictions.
Software, Technique Implementation and Services Used This section discusses the software and services used in this project
- Arduino IDE
- OTA Update
- Node-Red
- Cloud
- Android Studio
- Amazon S3 Bucket
- Amazon Sage Maker Canvas
The reason is that Android Studio offers an on-the-go development environment such as a Gradle-based system. Firebase is chosen because both Android Studio and Firebase are from Google and there will be no compatibility issues. The traditional way to upload new firmware to the microcontroller is through a data cable that connects to the computer.
In the future, if this wall outlet is available in the market, the wall outlet can continue to be updated accordingly. When the internet is restored, the local data will be uploaded to the cloud again. Amazon S3 Bucket is an Amazon Web Service (AWS) that enables data storage through a web service interface.
Datasets must be uploaded to the S3 Bucket before starting the machine learning process in Amazon SageMaker Canvas. Amazon Sage Maker Canvas is a machine learning tool that uses a visual point-and-click interface to construct a prediction model. Users can instantly connect and access data from cloud and local data sources, assemble datasets, and generate unified datasets for training machine learning models with SageMaker Canvas.
SageMaker Canvas automatically finds and corrects data errors, as well as analyzing data preparation for machine learning.
Hardware Implementation
- Raspberry Pi 4 Model B
- ESP 32 DevKit V1
- TRIAC
- Optocoupler
- ACS712 Current Sensor Module
The TRIAC is used in AC power control applications to control switching high voltages or high currents. In this project, the TRIAC is used to turn the wall outlet on and off. The criteria to consider for a TRIAC is the off-state peak repetitive voltage, VRPM,.
VRPM is the maximum peak voltage that the TRIAC can withstand, while IT(RMS) is the maximum RMS current that can pass through the TRIAC. Optocoupler is used to transfer the electrical signal between two isolated circuits using infrared light. The criteria for an optocoupler are the maximum operating isolation voltage, VIORM, and the infrared LED trigger current, IFT.
Assume the output voltage of a microcontroller is 3V; with a resistance of 330Ω, the IFT of 10mA can drive the infrared LED. ACS712 current sensor is used to measure the current consumed by the devices connected to the smart wall outlet. The integrated Hall IC in ACS 712 uses this theory by converting the magnetic field into a proportional voltage.
The ACS 712 is integrated with a monolithic Hall-effect based IC, making it reliable and low power dissipation.
Introduction
Hardware Prototype
Central Console
Smart Wall Outlet
Software and Technique Used for Smart Wall Outlet
Node-Red
- Wall Outlet Section in the Node-Red Dashboard
- Electricity Tariff Section in the Node-Red Dashboard
- Floorplan Section in the Node-Red Dashboard
The dashboard is divided into three sections, namely the wall socket section, the floor plan section and the electricity tariff section. Meanwhile, on/off status messages will be sent to the wall socket via MQTT protocol. The "Floor" node is a widget used to create scalable vector graphics (SVG) on the dashboard.
This section clearly shows the energy consumption and electricity bill for each room, and it is easy to identify the room that consumes the most energy. When the electricity bill exceeds the threshold, a warning logo will appear to inform consumers that their electricity consumption has exceeded the threshold. All the data about the smart wall socket will be uploaded to the real-time database and retrieved by this Android app.
This allows the application to control and monitor the smart socket outside the local network. The electricity bill for each room is highlighted in red when the limit limit is exceeded. By clicking on the "Room" buttons, the detailed information about individual wall outlets in each room will appear as shown in Figure 4.14.
Users can be aware of how much electricity has been used in the past and adjust their usage habits accordingly.
OTA Updates for ESP32
- Web Server Based OTA Updates
- Cloud-Based OTA Updates
The new firmware is then ready to load after the login credentials are entered and verified by the web server. After entering the correct user ID and password, users can perform in-place OTA updates, as shown in Figure 4.19. After the corresponding password is added to the firmware, the OTA update function is implemented successfully.
The status can be confirmed on the Mdash IoT website, as shown in Figure 4.21.
Program Operation
Power Consumption of the System
Electricity Cost of the System
Accuracy of ACS712 Current Sensor
This shows that the ACS 712 current sensor is not suitable for measuring low current. The reason for this is that the current sensor is exposed to the magnetic field and noise. For other devices with a higher power from 35W to 1600W, the error rate is very low and fluctuates between a maximum of 5.594% and a minimum of 0.23%.
Prediction of Power Consumption for Appliances Connected to Single Wall Outlet
The R-squared method is performed to test the fit and performance of the model. This means that the model has an average kWh difference from the actual energy consumption. The average percentage error between the actual value (daily energy consumption data from the original data set) and the predicted value (daily energy consumption predicted by the model) is 9.439%.
In short, this model can predict daily power consumption with an average percentage error of 9.439%.
Impacting Factors
Conclusion
Last but not least, a prediction model has been developed that predicts the daily power consumption of appliances connected to a single outlet, and the model can achieve an accuracy of 90%.
Recommendations for Future Work
This system is suitable for residential, commercial and industrial use in Malaysia and can calculate the electricity bill based on the ETOU schedule. Energy consumption and spatial assessment of renewable energy penetration and energy efficiency of buildings in Malaysia: an overview. Smart Plug Market Report Summaries Detailed Information by Key Players Belkin International Inc, BroadLink Technology Co.
What is an Amazon Smart Plug?': Everything you need to know about the Amazon plug that lets you turn on devices with your voice. The smart plug market size will reach $20.84 billion by 2028, driven by the growing popularity of smart homes and. CO2 emissions, energy consumption and economic growth in the ASEAN-5 countries: a cross-sectional dependence approach.
Retrieved from https://www.businesstoday.com.my smart-meters-a-progressive-system-for-better-user-experience. Energy conservation and smart billing system for domestic consumers connected to smart grid in Tamil Nadu Power System.
