The energy demand in the world is always increasing day by day. This is due to the advent of new electronic components in the daily lives of us. The demand for power is always increasing and the generation is decreasing drastically. This leads to a situation where the energy demand is not matching the energy supply. Developed countries have come up with load controlling techniques in order to balance the supply demand mismatch [1]. The load controlling techniques can be broadly classified as load based control and price based control. Load based control is a method of turning off the loads of the consumers when the generation is coming down. The latter technique involves deciding the price of the electricity consumption based on the demand. In times where the demand is less, the electricity price is less and vice versa. The consumer has an option of purchasing the power when desired and also economical. But the disadvantage in the price based control technique is the consumer’s electricity bill increases if power is consumed in peak times.

The load controlled markets will not offer comfortability to the user to use all the appliances at all the times. Hence the concept of energy management technique started in the early 90’s ensured that the energy is managed well in load controlled markets. The energy management techniques discussed so far in the literature is concentrating on the generation perspective, whereas this paper addresses the load management technique with economic utilization as an important function. Power failures are frequent and a major issue that is to be addressed in under developed and developing countries. Hence the energy controller designed should also be able to supply from alternate source of electrical supply.

2. LITERATURE REVIEW ON ENERGY MANAGEMENT SYSTEM IN A SMART HOME:

Home energy management system, HEMS in short, is a research area which is concentrated on the controlling and managing energy in small scale usually in homes. The HEMS is integrated with the battery, alternate energy sources and grid supply to intelligently manage and control power in a smart home. Battery management system is also integrated into the HEMS to yield better results in a smart home. As discussed earlier, the power system is dynamic in nature which may interrupt the supply power at any time. Hence HEMS should take care of the fluctuation of supply and provide uninterrupted supply of power to the smart home.

The literature classifies the loads into deferrable or non-critical and non-deferrable or critical loads. The classification is based on the practice of the consumers in the recent past and even regulations that govern the electricity usage. The work is significant as a model house is built and a real house situation for the consumption of electricity is simulated. The algorithm designed, classifies the loads into critical and non-critical loads based on the inputs provided to the controller. The loading profile of the smart home is also taken into consideration while deciding the criticality of the loads. The energy management system designed in the literature will charge the battery source from a solar panel located at the roof top. This is done so as to reduce the electricity bill of the consumer. In spite of the economic advantage in this literature, this may not be fully applicable for countries like India, China, Taiwan etc. as the solar energy is not promising round the year [2].

There are also many literature available that discusses the controlling the energy flowing to charge the battery banks during peak and off peak periods. The integration of the solar and wind energy to charge the battery during the peak period is also discussed in the literature. The developed control system using a microcontroller decides the source of the input to charge a battery based on the presence of peak and off peak period. The peak time of the days are assumed to the communicated a day in advance to the consumer for the controller to work [3].

The comfortability of the user is not taken into account for the previous literature. But this optimization and cost are clearly discussed in the work. The choice of deferrable and non-deferrable loads are provided to the user according to the need. The user can provide his choice of loads as one of the inputs of the algorithms. The optimization of cost of electrical energy is carried out using various algorithms such as game theory, Genetic Algorithm, Particle Swarm Optimization etc. The algorithm also discusses the concept of ON grid and OFF grid mode. OFF grid mode is a situation when the energy management system is unable to communicate with the power grid due to the failure. The energy management systems then decides to operate as a standalone unit by utilizing only alternate source of power in the house. In this literature, time of use price is used as the input to system and battery is charged only to its one tenth of the rated condition [4].

The novel method is to combine artificial intelligence, demand response and its optimization principle in energy management system for efficient use. The system developed is capable of learning the usage of loads and classify the loads into regulate able and non-regulate able loads. The system after the learning process decides the usage of sources for the no deferrable loads to be switched on or off. The algorithm also prioritizes the usage of High Voltage AC (HVAC) equipment over other loads [5].

The concept of load shifting principle is also used in literature [6]. The independence of the user to use any equipment at any given point of time is limited. Hence the load principle algorithm can also prove to be a complicated issue at least in developing countries which has no or less control on load. The answer to all the above problems lies in developing a controller which decides the source of electrical power for uninterrupted supply.

2.1 Summary of the Literature review:

· The concept of energy management in a home is relatively new and there are lot of gaps that needs to be addressed immediately

· Usage of devices is limited or the freedom of the consumer is hindered in many cases as the algorithm takes into account of type of loads and price factor

· Usage of alternate power supply in cases of power failure is to considered in many cases which is a beneficial to provide uninterrupted supply of power

· Charging of the batteries using the concept of SOC is not considered effectively

2.2 Research Gaps:

· The power consumption in a home is limited and can be predicted using any load forecasting techniques which is to be considered as one of the input to the controller block

· An intelligent controller is necessary to manage power in the smart home which can take care of both economic and continuous power supply

· Battery charging is to be considered as an input to the intelligent controller to increase the efficiency and life cycle of the battery banks

3. ECONOMIC ENERGY CONTROL FOR A SMART HOME:

Based on the literature review and research gaps found, there is a requirement of designing an intelligent controller which takes care of the above problems discussed. The intelligent controller is decided to be fuzzy logic controller which takes input as electricity price from the supply company, the forecasted load of the house, SOC level of the battery and the availability of the supply from the grid. The controller then decides the charging rate of the battery and the supply to be used as grid or from the battery.

Based on the load requirement of a normal house in India, 2.4 kW of power is expected in a day wise consumption curve. The loads in India is divided into agricultural, domestic and industrial loads. The agricultural loads is supplied at almost zero or minimal cost to the farmers. The domestic loads are the loads which are dynamic in nature and there is needs of comfortability to the user. The domestic loads could be managed by a small batteries and UPS. Hence the battery of capacity 150 Ah would be sufficient. But time taken to charge the battery to its full capacity is almost three hours. Hence the input requirement for the system would be modified as below:

· Electricity Price at (t) – The electricity price is one of the important factor to be affecting the electricity bill. Electricity price is assumed to be in dynamic nature. The price depends on the demand from the consumer. The lower demand from the consumer implies lower fees and higher demand from the consumer implies higher price. The rage of the price is maintained from rupee 1 to rupee 6 in the research.

· Load Demand at (t+3) – The load demand of the house for next three hours has to be known before in hand as it is one of the inputs to the system.

· State of charge of the Battery at (t) - The percentage of the battery available in the UPS unit.

· Supply Power available at the battery at (t) – Available or not available

The outputs of the system would be:

· Charging rate of the battery at (t) – Slow, Medium and fast

· Switch to electricity supply board power (t)

· Switch to battery source (t)

Note that, ‘t’ indicates the same instant of time,‘t+3’ indicates time instant after three hours. Hence in order for the designed algorithm to work efficiently, the load demand ahead of the three hours is to be known. This is required in order to charge the battery bank efficiently and safely. The algorithm which is designed for the energy management system in a smart home is as shown in the Figure 1.

Figure 1: Algorithm for Energy Management in Smart Home

The algorithm designed first gets the input such as electricity price, load demand of (t+3) hour, Present SOC of the battery and selling price of the electricity in a price dependent market. The algorithm then checks the availability of supply power at (t+3) hour and decides to charge the battery accordingly if the electricity price is low and battery SOC is also low. The algorithm also decides on how and at what rate the battery has to be charged. The algorithm also takes care of connecting the present supply to the grid or the battery according to the price and availability.

4. DESIGN OF FUZZY LOGIC CONTROLLER IN A SMART HOME:

Fuzzy logic is one among the artificial intelligent technique that are widely used for controlling purposes. The emergence of fuzzy logic dates back to early 1965 when Lotfi Zedah published his work in one of the conferences. The fuzzy system involves the output only in the range of 1 or 0 and can never exceeds theses limits. There were also oppositions from conventional mathematics about the validity of the fuzzy systems at its emergence. But nevertheless the concept was highly appreciated as this system could answer vague answer which conventional mathematics couldn’t have answered. The main advantage of fuzzy logic is instead of bi-valued outputs such as 0 and 1, it has the ability to answer the degree of truth.

The conventional controllers is designed only on the mathematical model from the system. The mathematical model for all the systems in the world are impossible or difficult to obtain. In fuzzy system, there is no need of developing a mathematical model of the system. The fuzzy logic controller is able to take care the uncertainties in system without the need of equations for the system.

Figure 2: Fuzzy Logic Architecture

The block diagram of the fuzzy logic system is shown in the figure 2. The inputs from the real world that are ranging from all scales are to be converted to the fuzzy inputs from zero to one. This process of conversion is called as fuzzification. The fuzzy outputs finally needs to be converted into the real world needs to be converted into the real world outputs. This process is called as defuzzification. In the process, there is a fuzzy inference engine, where the set of fuzzy rules are to be designed by the user. The rule matrix contains a set of rules that are decided upon and needs to be evaluated for all input conditions.

4.1 Membership Functions:

The values which gives the degree of true and false in the universe of discourse is called as membership function. The membership functions are present in both fuzzification block and defuzzification block. The real world valued input range is normally represented in the x axis and the degree of truth is represented along y axis. The shapes of membership function can be ranging from singleton, Gaussian, Gbell, Gaussian2f etc. A typical example of membership function is shown in figure 3.

Figure 3: Fuzzy Logic Membership shapes

4.2 Fuzzification:

The fuzzification is a process of converting the real valued input variables into the crisp values using the defined membership functions.

Fuzzy set A of universe X is defined by a function μA(x), called membership function, for set A

μA(x): X → {0, 1}, where μA(x) = 1 if x is totally in A;

μA(x) = 0 if x is not in A; 0 < μA(x) < 1 if x is partly in A;

4.3 Evaluation of Rules:

The rules of the fuzzy systems are applied to the fuzzified inputs from the fuzzy inference engine. The common rules that are implied in the fuzzy systems include AND, OR and NOT functions. In other names for these words include min, max and compliment functions. The following equations represent the relationship between the membership function.

μA∪B (x) = max [μA (x), μB (x)]

μA∩B (x) = min [μA (x), μB (x)]

μĀ(x) = 1-μA(x)

The rule base contains number of IF-THEN fuzzy rules, which are the relations between the fuzzy inputs and the outputs. Based on the IF-THEN fuzzy rules, there exists a nonlinear relationship between the number of inputs and which decides what degree of the truth is present for a particular membership function.

Mamdani and Sugeno type inference methods are used commonly to evaluate the fuzzy inputs and decide among the outputs. Minimum operator is used as a reference in Mamdani type whereas maximum operator is used in the Sugeno type inference method.

4.4 Defuzzification:

The defuzzification is a process of converting the fuzzy output from the fuzzy inference engine to the normal crisp set which is real value readable to the common person. The defuzzifier block combines all the fuzzy set values and gives a single value number which is the equivalent of all the fuzzy outputs.

The defuzzification methods have different types such as Centroid method / Center of area, mean of maximum, Sum of maximum etc. In this research, centroid method which finds an optimum solution by taking the weighted mean of all the fuzzified outputs.

4.4 Designing the system using MATLAB fuzzy tool box:

The system discussed above is designed in the MATLAB fuzzy toolbox. The system is as shown in the figure 4. The inputs such as electricity price, load demand and the State of charge (SOC) is divided into three membership functions as low, medium and high. This is shown in the figure 5,6,7 and 8 respectively.

Figure 4: Energy Control Module using Fuzzy Logic

Figure 5: Membership Function of Electricity Price

The input electricity price in India is between Rupee 1 to Rupee 6 and is been divided into three functions as shown in figure 5.

Figure 6: Membership Function of Load Demand

The load demand for the particular home can be obtained from the previous load profile and the load demand in the research is referred from [7].

Figure 7: Membership Function of SOC

Figure 8: Membership Function of Supply Power Availability

5. TESTING AND VALIDATION OF THE SYSTEM:

The developed system in fuzzy logic is tested using Simulink software. The fuzzy logic block is used as a controller block in the Simulink to test for the validity of the developed model as in figure 9.

Figure 9: Simulink testing of Energy Control Module

Here the values of the system inputs are arbitrarily obtained [8] and fed into the system and checked for the outputs from the system. The day ahead electricity pricing model is obtained from the supply company and is shown in the figure 10.

Figure 10: Day ahead Electricity Price from Supply Company

The rate of electricity consumption pattern as referred from the literature is shown in the figure 11. As we can see from the figure 11, the electricity rate is high during the periods in the evening and lower in the morning. This is because, the demand for the electrical energy is high only during the evening as the usage of domestic loads would be high. The same is lesser in the morning as the people will be working in offices and industries and the corresponding loads in the offices and the industries will be high.

It can also be seen from the figure 11 that the electricity price is high in the morning times as people tend use the electrical devices for their daily chores. The usage includes geyser, toaster, chimneys and heaters if any. The example of the load profile is just an example from a smart home and the similar pattern can be found in many residential loads across the country. The considered load profile in the night is high as people will use air conditioners and heaters during summer and winters respectively.

Figure 11: Load Consumption of the Smart Home

As the grid power is varying widely in nature, the availability of grid power for the particular area is taken as a data set from the supply company for a particular day. The data set is shown in the figure 12. From the figure it can be seen that there is no power at 7 and 8 in the morning for examples. The same is also reflecting the time 16 and 17 hours.

Figure 12: Grid Power Connection to the Smart Home

The simulations are run for all the inputs for the duration of 24 hours in a day. The response of the outputs from the controller is recorded and analyzed. The system is able reduce the electricity consumption in a smart home in comparison to the electricity consumption without the controller. This is as shown in the figure 13.

Figure 13: Comparative cost of electricity with and without the controller

This energy saving is obtained from the designed system as it charges the battery according to the need and price. It also connects the smart home to the battery or to the grid by recognizing the demand and it mismatch. From the figure it can be seen that the energy saving is a huge advantage and the system can be applied to the smart home. The smart home is connected to the grid only at the times when the price of the electricity is less as shown in figure 14. The price of the grid is also an input data set obtained from the supply companies. The electricity pricing is obtained from the supply companies a day prior.

Figure 14: Grid Connection of a Smart Home with Economic Controller

The charging rate of the battery is also shown in the figure 15 which is based on the dynamic pricing of the electricity with the load demand for the next three hours in the memory [9]. The decision of rate of charging the batteries is also one of the economic consideration in the research carried out.

The charging rate of the battery is divided into slow charging, fast charging and medium charging. The type of charging from the controller is decided from all the factors influencing the economy of the electrical consumption.

6. CONCLUSION AND FUTURE WORK:

The energy controller is required in the Indian energy market which is very dynamic in nature [10]. The consumption bill of the homes is rising day by day as more number of equipment is being added in exponential fashion for our life cycle. The energy controller designed takes care of the imminent non linearites in the smart home by considering the dynamic ranges of inputs, outputs and the supply availability.

The fuzzy logic controller implemented in the research utilizes no mathematical model in the system. Hence this type of controllers can be used in many applications in which finding the equations of relations is highly difficult. The accuracy of the designed controller is as comparable to the normal controllers used elsewhere. The fuzzy logic controller technique is easy to design and not time consuming. But the disadvantage exists, where the designer needs to know the fuzzy logic inputs and the expected outputs for designing the rule base.

With the implementation of the economic energy controller, the electricity bill of the smart home is drastically reduced up to 42%. This could be seen from the comparative analysis as in figure 13. The future of this research work can be considering more inputs from the renewables. The renewable energy is gaining interest more day by day and can be implemented to the designed system with some slight modification to the existing algorithm. The fuzzy logic can also have more number of membership functions and it can be evaluated with different membership shapes.

7. CONFLICTS OF INTEREST:

The author declare no conflict of interest.

8. REFERENCES

1. Palensky P, Dietrich D. Demand side management: Demand response, intelligent energy systems, and smart loads. IEEE transactions on industrial informatics. 2011 Aug;7(3):381-8.

2. Boynuegri AR, Yagcitekin B, Baysal M, Karakas A, Uzunoglu M. Energy management algorithm for smart home with renewable energy sources. InPower Engineering, Energy and Electrical Drives (POWERENG), 2013 Fourth International Conference on 2013 May 13 (pp. 1753-1758). IEEE.

3. Al-Ali, A.R., El-Hag, A., Bahadiri, M., Harbaji, M. and El Haj, Y.A., 2011. Smart home renewable energy management system. Energy Procedia, 12, pp.120-126.

4. Chaimaa ESSAYEH, M. R.-F. (2016). Towards an Intelligent Home Energy Management System for Smart MicroGrid Applications. 2016 International Wireless Communications and Mobile Computing Conference (IWCMC) (pp. 1051-1057). Paphos, Cyprus: IEEE.

5. Dong Zhang, S. L. (2015). An Optimal and Learning-Based Demand Response and Home Energy Management System. IEEE Transactions on Smart Grid.

6. Mohammad Rastegar, M. F.-F. (2012). Load commitment in a smart home. Elseviaer, 45-55.

7. Devana, M. V. (2015). Design and Implementation of Intelligent Energy Management System for Residential Applications. Bangalore: M.S. Ramaiah School Of Advanced Studies.

8. Hong-Tzer Yang, J.-T. L.-I. (n.d.). A Load Forecasting Method for HEMS Applications. Taiwan: National Cheng Kung University.

9. Prajwal, K.T. and Gupta, V.S., 2018, January. Smart home energy management system using fuzzy logic for continuous power supply with economic utilisation of electrical energy. In 2018 2nd International Conference on Inventive Systems and Control (ICISC) (pp. 274-279). IEEE.

10. Timothy J. Ross, J. M. (2007). fuzzy logic with engineering applications. Wiley India Pvt Ltd.

Received on 30.04.2019 Accepted on 10.05.2019

Research J. Engineering and Tech. 2019;10(2):83-93.

DOI: 10.5958/2321-581X.2019.00015.1