Real Time Electricity Monitoring using Smart Energy Meter in a Smart LAN based Network

Transcription

1 Real Time Electricity Monitoring using Smart Energy Meter in a Smart LAN based Network Aakash Bansal 1, Varun Goyal 2 Department of Electronics and Communication Engineering, Maharaja Surajmal Institute of Technology, New Delhi, India Abstract With the advancing technology, every service is shifting towards making systems automated, portable and easy to use. Supporting the change in time, the paper proposes a technique for real time monitoring of electricity consumption using smart energy meters, connected with each other in a smart network for theft analysis. Wirelessly enabled smart energy meters are connected in a LAN network, which is further connected to a bigger network, similar to a GSM Mobile Network. The main aim of the smart energy meter is the detection of energy thefts in local distribution of electricity throughout the country. To serve the purpose, server runs real time analysis and report consumption of electricity regularly. The abrupt changes in the received data at the server end is predicted as energy theft cases. Index Terms Electricity Theft, Smart Grid, Energy Meter, Real Time Monitoring, LAN Application I. INTRODUCTION Food-Cloth-Shelter being the basic necessity for a human being, the advancing time has added the need for electricity to the list. Almost all the appliances, either medical, industrial or domestic makes use of electricity to make lives easier. To fulfil the need for electricity, most of the governments have started privatisation of electricity to organisations like Tata Power. To maintain a balance between demand and supply of electricity, these organisations charge users according to their usage under government bodies. To monitor the consumption of each household, organisations installed energy meters, which are used to read the regular consumption in terms of units. These units specify the electricity consumption, and hence are used for billing. To save money, people started looking out for a way out, and hence started stealing electricity. Many people started taking electricity directly from poles without an energy meter in between. Many started using magnetic effects to fail energy meters observations. Each time organisations introduce a new solution to the electricity theft, another way for the same is discovered by the public.to solve this issue of regular electricity theft, we propose a wireless local area network of smart energy meters connected on ISM band of 2.4 GHz. The network is similar to a GSM implementation, and does not need any internet connectivity, and hence making it implementable for developing and underdeveloped countries as well that majorly faces the problems of energy theft. Smart Energy Meter transfers data regarding electricity consumption to the central node of the local network, that does the necessary computations for real time analysis and predicts energy thefts in the local network. Using this data, the organisations and local governments can detect the source of energy theft, and necessary actions can be taken on time. The following paper describes the above specified technology in detail. The upcoming section describes the operation of traditional energy meters, and how these traditional meters are converted into smart energy meters. This section is followed by the description of the wireless network proposed in the paper, and techniques for data sharing. The next section describes the algorithms used for computation for real time analysis of theft detection and alarming to the concerned authorities. The paper ends with a test case and the results built for the proposed technique. II. ENERGY METERS Traditional Energy Meters are either electronic or electro-mechanical in nature. They measure the electricity consumption in terms of kw-hour, and displays in terms of units. Each pulse represents one kw-hour of energy consumption in the household. 410 Aakash Bansal, VarunGoyal

2 The most commonly used energy meter is the electromechanical induction watt-hour meter. Such meters count the number of revolutions of a non-magnetic, but electrically conductive disc. This disc rotates at a speed proportional to the power passing through the meter, and hence it measures the power consumed by the household. Another popular energy meter available in the market, and is presently used is a solid state electronic energy meter. Such energy meters use solid state devices for the measurement of current and voltage supplied to the load, and using the same, calculates the power consumed by the household and is displayed on the screen along with the present date and time. It also have some indicating LEDs that indicates conditions such as phase, fault, calibration, power consumed, etc. The same can be seen in figure 1. Figure 1 Interface of a Generic Energy Meter Figure 2 Block Diagram for the Smart Energy Meter III. IMPLEMENTATION OF A SMART ENERGY METER One of the biggest challenge in the regularly upgrading technology is to make things compatible with the past technology as well so that no physical damage in the pre-installed devices. 411 Aakash Bansal, VarunGoyal

3 In order to make the pre-existing energy meters smarter, an additional device is attached inside the system. In the presently used meters, one of the indicators indicate the consumption of power. The LED blinks with every change in the meter readings. Hence, using the signal pulse generated at this indicating LED, the additional device reads the regular consumption in real time. Using a microcontroller, the data recorded is stored temporarily and is transmitted to the central server of the network for further analysis. The circuit diagram for the same can be seen in figure 3. It consists of a controller, ATmega16 for the reception of data from old conventional energy meters. Using a transistor as a switch, the signal bursts from the indication section of the generic energy meters is received at the microcontroller in form of pulses. The signal remains high throughout the time. Whenever there is an addition of 0.1 kwatt-hour, a pulse is generated at the base, enabling the transistor, and a pulse of digital logic low is received. Using this technique, the microcontroller is programmed to keep appending the value with minor computations for each pulse received, and store the same within itself. The next important part of the circuit for Real Time Monitoring is a Wireless Node. The proposed Smart Energy Meter uses nrf24l01+ to participate in a local LAN. nrf24l01 is a single chip transceiver module that works on a 2.4 GHz ISM Band. It is an open source transceiver module and comes with an embedded baseband protocol engine. It has an inbuilt power amplifier and a low noise amplifier for improved efficiency. The range of the wireless node can be increased easily with the help of antennas. One can use any type of antenna as with the module as long as it has an SMA connector. The transceiver module connects the device to a local network of a community, and transmits the collected results at a central location in the local network. This central hub of the local network does the necessary computations to analyze the consumption patterns and predict electricity theft, and transmit details to the host accordingly. IV. NETWORK ARCHITECTURE The Smart LAN or Local Area Network can be considered as an analogy to the conventional GSM Architecture used in the same fashion for this application with some changes to the same. Each specified area will be covered by a Base Receiver Station (BRS). The BRS will communicate with all the energy meters installed in houses throughout the coverage area, collects their data, record and analyze consumption pattern per household and predict the theft accordingly. This communication between households and a Base Receiver Station is depicted in figure 4. Figure 3 Wireless Network of Household Meters Connecting to a Central Hub. The Dotted Lines represent the Wireless Link between Household and Base Receiver Station (BRS). The similar networks will be used to analyze multiple households at the same time. And for each result built, data will be transmitted from these base receiver stations to a central node of the complete network. This central node can be kept in direct control of the organization and concerned authorities to take immediate action against the theft. The overall proposed network can be seen in figure 5. The network is divided into three different sections, namely, Data Collection Section, Base Control Section, and Central or Main Control Section. 412 Aakash Bansal, VarunGoyal

4 A. Data Collection Section The data collection section covers the base receiver stations and the household energy meters. Smart Energy Meters installed in the households uses a wireless link on a band of 2.4 GHz to transmit the collected data from conventional meters to the Base Receiver Station using nrf24l01 transceiver modules. Base Receiver Stations use the collected data to analyze the electricity consumption and detect cases of energy theft in the coverage area. The collected and analyzed data is sent to the Base Control Section for further processing and providing necessary back bone to the data collection Section. B. Base Control Section The Base Control Section acts as a back bone for the base receiver station architecture and data collection section. This section collects data from different base receiver units and routes the data for further processing and other necessary functions such as billing to the main or central control section. The complete network is unidirectional and hence data travels only from data collection section to the main control section. C. Main Control Section The Main Control Section or the Central Control Section is the central hub for the complete network. The final data regarding consumption of energy for each household is collected at the central hub. The hub has a user database for storing the consumption data, and verifying active and inactive connections. It also has a billing unit which generates bill for each smart meter, hence making the system automated. V. PROPOSED ALGORITHM The algorithm described below uses a microcontroller to control the functionality of the system, a wireless module to build the network of energy meters in the locality, and an additional circuit to make present energy meters smart. Figure 4 Wireless Network Architecture for Real Time Consumption Monitoring with Smart Energy Meters Initially, as described earlier, the system will first receive data from the traditional energy meter using the switching characteristics of transistor and will generate a pulse at the analog input of the microcontroller. For each change in pulse, the microcontroller will keep increasing the data specifying the energy consumption. Now, this microcontroller also has a counter, and after every one minute, the fresh data is transmitted to the 413 Aakash Bansal, VarunGoyal

5 Power in kw International Journal of Electronics, Electrical and Computational System local network using the wireless node nrf24l01 transceivers. The transceiver module transmits a particular sequence of data in form of a code word to maintain accuracy and security in the communication. The code word includes the unique address of the transmitting smart energy meter, data regarding consumption in bits, working status of the energy meter, parities and checksum for minor corrections and encapsulation to keep the data secure. The following data in form of codeword is generated by the attached microcontroller and using the same, the transceiver module transmits information to base receiver station. After receiving the data, the base receiver station appends the data-set of the corresponding energy meter, and using basic mathematical functions, it does the analysis for detection of energy theft. At the base receiver station end, for each data received, it compares the present data with past three data-set and checks for the continuity of consumption. If there is a regular consumption going on, there will be a linear increase in the curve. In case, the energy consumption stops, the data received must be approximately equivalent to the previous data. But, if there is any kind of hampering done, leading to the drop of signal, or in case the LED signaling of the energy meter is stopped, the same can easily be identified by analyzing the data received from the microcontroller and the smart energy meter, and using the same, one can easily detect an intentional tempering of the energy meter. On the base receiver station end, as per the proposed algorithm, for any detection of such malicious activities, concerned authorities are informed in real time to take necessary action. Due to limit of memory at base receiver station, the BRS is proposed for an ability to store a data of a about 24 hours and then overwriting the same, to keep the analysis unaffected. With the help of this proposed algorithm, the real time analysis of electricity consumption can be done throughout the day, and cases of electricity theft can be reduced on a significant scale. The next part shows the observations of the presented work, and the analysis done at the Base Receiver Station end. VI. OBSERVATIONS AND RESULT The above proposed network and smart energy meter has been developed in form of a prototype as a proof of concept. To demonstrate the same, five of the energy meters have been modified to smart energy meters, and are communicating with a central hub which acts as a base receiver station. This base receiver station collects the data from different energy meters, and using the microcontroller, it feeds data to the computer via serial communication. Using the computation tool, Microsoft Excel, the necessary analysis of the received data is done and the same has been presented in the section. 16 Base Receiver Station Data Reading Count Main Meter SM1 SM2 SM3 SM4 SM5 Figure 5 Result Output for five Energy Meters at Base Receiver Station 414 Aakash Bansal, VarunGoyal

6 Using the circuit shown in figure 3, the energy meters are upgraded to smart energy meters. Instead of tempering with the internal connections of the meter, a light sensor has been used which is placed right on the indicating LED. This ensures the security of the pre-installed traditional energy meter without any internal damage and tempering. The sensor detects each blink of the indicator LED, and sends the respective signals generated to the microcontroller for further processing of data at smart interface end..the microcontroller then uses the received signal to generate data specifying electricity consumption out of it. Using the data, after necessary computations, the microcontroller uses the nrf24l01 transceiver module to transmit data to the central base receiver station. Now since data is being received from multiple ends, the network is using polling technique in this case to collect data from all fivetransmitting energy meters. The collected data is later processed by Microsoft excel. A data set for about thirty minutes is tested and presented in this paper. The curve presented in Figure 6 shows the data received from five different energy meters and the actual power being provided by the electricity board to that specific area. Observing the data, one can see there is an abrupt change in the data set of smart energy meter 1. Also, if observed for each time slot, the summed-up result of all five smart meters isn t equating from reading count 15 to 23, which itself represents some malicious activities. In this time slot, only smart energy meter 1 seems to produce the most unexpected results. Hence, it can be deduced that there is some tempering being done with energy meter 1 itself. At the end, this data is transmitted to the concerned authorities and organisations to take necessary action within time. The demonstration and final results for the proposed system is hence presented, and is observed to be operating perfectly. REFERENCES [1] Stephen McLaughlin, Brett Holbert, Ahmed Fawaz, SamanZonouz, A Multi-Sensor Energy Theft Detection Framework for Advanced Metering Infrastructure, IEEE Journal on Selected Area in Communications, Vol. 31, No. 7, July [2] Rong Jiang, Rongxing Lu, Jun Luo, Energy Theft Detection Issues for Advanced Metering Infrastructures in Smart Grid, Tsinghua Science and Technology, Vol. 19, No. 2, April [3] G. L. Prashanthi and K. V. Prasad, Wireless Power meter and Intimation System using GSM and Zigbee Networks, Journal of Electronics and Communication Engineering, Vol. 9, Issue 6, Ver. 1, Nov 2014 [4] Alvaro A C ardenas, Saurabh Amin Galin Schwartz, Roy Dong, A Game Theory Model for Electricity Theft Detection and Privacy-Aware Control in AMI Systems, Fiftieth Annual Allerton Conference, October [5] W. A. Doorduin, T. Mouton, Feasibility Study of Electricity Theft Detection using Mobile Remote Check Meters, IEEE AFRICON 2004, October [6] Sergio Salinas, Ming Li, Pan Li, Privacy-Preserving Energy Theft Detection in Smart Grid: A P2P Computing Approach, IEEE Journal on Selected Areas in Communications/Supplements, Vol. 31, No. 9, September [7] Saurabh Amin, Galina A. Schwartz, Game-Theoretic Models of Electricity Theft Detection in Smart Utility Networks, IEEE Control Systems Magazine, February [8] S. McLaughlin, P. McDaniel, Energy Theft in the Advanced Metering Infrastructure, Proceedings of International Conference on Critical Information Infrastructure Security, Springer-Verlag, pp , [9] Pandurang G. Kate, Jitendra R. Rana, ZigBee Based Monitoring Theft Detection and Automatic Electricity Meter Reading, 2015 International Conference on Energy Systems and Applications, November [10] D. N. Nikovski, Z. Wang, T. Muso, Smart Meter Data Analysis and Power Theft Detection in Machine Learning and Data Mining and Pattern Recognition, Springer Journal for Computations [11] N. Mahalakshmi, E. Elavarasi, Design of Intelligent SMS Based Remote Metering System for AC Power Distribution to HT and EHT Consumers, International Journal of Computational Engineering Research, Vol. 2, Issue 3, pp , June [12] MdMejbaulHaque, Md. Kamal Hossain, Md. Rafiqul Islam Shiekh, Microcontroller Based Single Phase Digital Prepaid Energy Meter for Improved Metering and Billing System, International Journal of ower Electronics and Drive Systems, Vol. 1, No. 2, December [13] Z. Xiao, Y. Xiao, D.H. Du, Non-repudiation in Neighborhood Area Networks for Smart Grid, IEEE Communication Magazine, Vol. 51, No. 1, pp/ 18-26, [14] B. Loeff, Deputizing Data: Using AMI for Revenue Protection, Utility Automation and Engineering [15] J. Nagi, K. Yap, S. Tiong, S. Ahmed, Detection of Abnormalities and Electricity Theft using Genetic Support Vector Machines, IEEE TENCON Region 10 Conference, pp. 1-6, [16] A. Molina-Markham, P. Shenoy, D. Irwin, Private Memoirs of a Smart Meter, Proceedings of the Second ACM Workshop on Embedded Sensing Systems for Energy Efficiency in Building, pp , Aakash Bansal, VarunGoyal