Hi I need help with a research paper
thecameragirl
07992131.pdf
Design and Implementation of a Smart Home Energy Saving System with Active Loading Feature
Identification and Power Management
C.M. Lin
Department of Electrical Engineering (EE)
National Kaohsiung University of Applied Sciences
Kaohsiung, Taiwan, R. O. C.
M.T. Chen
Department of Electrical Engineering (EE)
National Kaohsiung University of Applied Sciences
Kaohsiung, Taiwan, R. O. C.
Abstract—In this paper, a smart home electric energy saving
system is implemented by combining smart meter, smart plug, smart mobile devices, and database server. The smart meter consists of a power metering unit, a data storage unit, a meter interface unit and a ZigBee module. The smart plug is composed of a core control unit and a remote monitoring module. User can use smart phone to check and control the operation of appliance, and the power consuming information can be remotely monitored by connecting smart plug to the internet via Wi-Fi media. Besides, the load characteristics in the database server can be employed to identify appliance operation mode by Support Vector Machines (SVM) method, which provides effective message for home electric energy saving application. Finally, a prototype was built up and tested, the test results validate the feasibility of the proposed smart home electricity saving system.
Keywords—Smart Meter, Smart Plug, Support Vector Machines, Remote Monitoring, Home Electric Power Saving.
I. INTRODUCTION For global energy shortage and price rising, energy-saving
has become a serious issue that brings smart grid concept and development in recent years. Regarding as smart meter would base of the smart grid architecture, most of nations started to develop home electric energy management system with Advanced Metering Infrastructure (AMI). It measured and transmitted electric parameters to data center automatically, also provided remote control function to users, so that they could change their energy usage behavior based on energy information [1].
There are many different approaches among classification algorithms, machine learning and support vector machine (SVM) has recently achieved effectively methods for data recognition [2-3]. However, considering the uncertainty of user’s energy consumption behavior, it is not yet completely clear for applying SVM method to identify appliance operation mode and home electric energy saving application.
Besides, the handheld mobile devices are becoming increasingly popular in recent years, APPs can provide convenient interface for remote real-time monitoring and
operation, allowing users to communicate with the controlled device. This is the effectively way for users to realize home electric energy saving system, smart plug information, energy consumption history and power management [4-5].
This thesis proposed a home energy saving network frame that applied SVM with historical data to identify appliance operation mode. The smart home electric energy saving system would be shown and demonstrated.
II. HOME ENERGY SAVING SYSTEM STRUCTURE
Fig. 1. Simplified structure of the proposed home energy saving system
Fig.1 shows the home energy-saving system, it can be seen
that there are three major units, smart meter, smart plug, and
smart phone [1]. For the smart meter, the digital power meter is
used to capture raw voltage and current signals separately, and
the electric power energy consumption can be computed from
the sampled discrete signals. Then, the real-time accumulative
electric energy information is stored into EEPROM module and
delivered to Meter Interface Unit (MIU) through the Dual Port
RAM (DPR) modules. By smart plug, we can measure and
analyze the power data for different loads, which is displaying
on LCD and transferred to MIU by ZigBee. The power
978-1-5090-5157-1/17/$31.00 ©2017 IEEE 739
information will be also transmitted to the smart phone via Wi-
Fi module (CC3100). Both of the electric energy consumption
and power information can be passed to a remote Data Center
with a client-server platform.
A. Smart Meter Unit
Fig. 2. Flow chart for DPM to collect power parameters and Overcurrent information
The smart meter consists of Digital Power Meter (DPM) and Metering Interface Unit (MIU). The core components of DPM and MIU are digital signal processors TMS320-F28335 and TMS320-F449 respectively. First, the power system voltage and current signals are respectively attenuated and isolated by Voltage Probe (VP) and Current Probe (CP); and the output signals of VP and CP are then inputted to a pre-sampling circuit of the power IC (ADE7953) which is used to measure the power parameters including real power, reactive power, power factor, rms voltage and rms current. All of these power parameter information are then processed by the microprocessor, TMS320- F28335, according to the flow chart shown in Fig. 2. The microprocessor MSP430-F449 of MIU reads the stored data from Dual Port Ram (DPR) for checking emergency event and uploading the real-time power parameters to data center.
B. Smart Plug Unit
Fig. 3. Simplified structure of smart plug
The hardware of smart plug is shown in Fig.3, the microprocessor, MSP430-F5529, is the kernel of the module. The Wi-Fi interface module, CC3100, can transmit power parameters and information to user’s smart phone wirelessly. The smart plug can also upload data to MIU through ZigBee communication module. The power supply to the appliances can be switched on or off by the Solid State Relay (SSR) remotely.
C. Mobile Device Application Development
Fig. 4. The development of home energy management of mobile
The mobile device application is implemented by Android platform with software suit which are ADT, Android SDK, and Eclipse SDK. The procedure to develop home energy management by application programming interface (API) as shown in Fig. 4.
D. Remote Control by LAN/WAN
Fig. 5. Overview of LAN and WAN communication route between mobile device and smart plug
Fig. 5 shows two communication medium of LAN and WAN between mobile device and smart plug. For LAN, user control and access from router to smart plug via Wi-Fi. In WAN which using 3G or 4G mobile communication, for its protective restriction of wireless to local network, therefore, introduced Network Address Translation (NAT) technology to convert and simplify the IP address between internal and external network. This paper used the restricted cone NAT for configuring the IP address definition, checking the transmission information and ensuring the security of data transmission.
740
E. Loading Operation Mode Identification This paper proposed a identification method by SVM, that
would classify the loading operation mode by using electrical energy information as feature.
We selected the proper period for identification model, for each classifying operation mode by five samples including the moment of before 10 minutes, an hour, yesterday, last week and last month respectively. In other words, the proposed method need at least eight weeks to establish the model. The warning message will come up to notify user when identical result was different to current state after checking twice within 10 minutes period. Table I was the experimental example, illustrated the proposed rule.
TABLE I. EXAMPLE OF IDENTIFICATION FEATURE SELECTION
III. SYSTEM TEST RESULTS
A. Electric Power Parameters Measurement Fig. 6 shows the test results on the MIU LCD display real
time information for a low voltage user’s load, including rms voltage and current (Vrms and Irms), apparent power (S), real power (P), reactive power (Q), power factor (PF), electric energy and billing.
Fig. 6. Power parameters displayed on LCD monitor of smart meter
B. Home Energy Management by Smart Plug In this study, we choose the living room as a experimental
test object. In Fig. 7, the living room is divided into three
regions (A1, A2, and A3). According to the corresponding areas,
that classified different of appliances to monitor the power
energy consumption state. The test schedule planning for A1,
A2, and A3 is in table II.
Fig. 7. Energy power consumption and power parameters of different areas
(a) A1, (b) A2, (c) A3
TABLE II. EXAMPLE OF IDENTIFICATION FEATURE SELECTION
C. Home Energy Management of Mobile Device Users can use mobile device to remote monitor the state and
power energy consumption of appliances in house when smart plug connected to Wi-Fi successfully. In Fig.8, shows the power state and parameters of the personal computer on user mobile screen including in use (green), standby (yellow) and cut off (red).
Fig. 8. Displaying the different power state and parameters on user mobile device (a) in use (green light), (b) standby (yellow light), (c) cut off (red light)
IV. CONCLUSIONS Due to the growing number of household electrical
appliances, and its power consumption. The energy power management, such as standby power control, must be effectively implemented. In this paper, the proposed structure is built using multiple low-cost microprocessors to increase its feasibility that contains smart meter, smart plug, and smart home energy-saving
Date Time
Sampling Scale State
(1:ON/ 0:OFF)
10 minute s before
an hour before
yesterday before
last week
before
last month before
4/19
19:00 18:50 18:00
4/18
18:00
4/12
19:00
3/19
19:00 1
4/12
19:00 18:50 18:00
4/11
19:00
4/01
19:00
3/12
19:00 1
Area Schedule arrangement of Standby Mode
Appliances Operation period Standby period
A1 printer, monitor, computer, sound
7 hours (17:00~24:00)
17 hours (7:00~17:00 and
24:00~7:00)
A2 air conditioner 9 hours
(7:00~9:00 and
17:00~24:00)
15 hours (9:00~17:00 and
24:00~7:00)
A3
home theater,
LCD TV,
TV box, DVD player
9 hours (7:00~9:00 and
17:00~24:00)
15 hours (9:00~17:00 and
24:00~7:00)
741
system. A single phase prototype meter system was built, and the desired functions were tested in the laboratory.
REFERENCES
[1] Ming-Tang Chen, and Che-Min Lin, “Development of A Smart Home Energy Saving System Combining Multiple Smart Devices,” in 2016 IEEE International Conference on Consumer Electronics - Taiwan (ICCE-TW), 2016, pp. 1-2.
[2] Lei Jiang, Suhuai Luo, and Jiaming Li, “An Approach of Household Power Appliance Monitoring Based on Machine Learning,” in 2012 Fifth International Conference on Intelligent Computation Technology and Automation (ICICTA), 2012, pp. 577-580.
[3] Minyoung Sung, and Younghoo Ko, “Machine-Learning-Integrated Load Scheduling for Reduced Peak Power Demand,” IEEE Transactions on Consumer Electronics, vol. 61, no. 2, pp. 167-174, May 2015.
[4] Chao-Hsien Lee, and Ying Hsun Lai, “Design and Implementation of a Universal Smart Energy Management Gateway based on the Internet of Things Platform,” in 2016 IEEE International Conference on Consumer Electronics (ICCE), 2016, pp. 67-68.
[5] Joon Heo, Choong Seon Hong, Seok Bong Kang, Sang Soo Jeon, and Younghoo Ko, “Design and Implementation of Control Mechanism for Standby Power Reduction,” IEEE Transactions on Consumer Electronics, vol.54, no. 1, pp. 179-185, Mar. 2008.
742
