LOAD BALANCING OF FEEDER USING FUZZY AND OPTIMIZATION TECHNIQUE

Transcription

1 International Journal of Electrical Engineering & Technology (IJEET) Volume 9, Issue 4, July- August 2018, pp , Article ID: IJEET_09_04_008 Available online at ISSN Print: and ISSN Online: Journal Impact Factor (2016): (Calculated by GISI) IAEME Publication LOAD BALANCING OF FEEDER USING FUZZY AND OPTIMIZATION TECHNIQUE B. Phani Ranga Raja and K. Jyothi Sree Assistant Professor in Usha Rama College of Engineering and Technology, Department of Electrical and Electronics Engineering ABSTRACT Phase balancing generally occurs during minimization of loss, planning and restoration of energy in distribution systems. Some areas may be over loaded due to fault in distribution. In order to overcome these problems, power controlling and there by controlling of load is required for those areas. It leads to load balancing technique. In these papers, load balancing is achieved by fuzzy logic control. Input to the fuzzy is the total load of the feeders. The output of the fuzzy step is the input to the load balancing system. Load balancing system utilizes optimization technique to convert kilowatts values into load points and specify the load points. Keywords: Load balancing, fuzzy logic, feeder, optimization and MATLAB Cite this Article: B. Phani Ranga Raja and K. Jyothi Sree, Load Balancing of Feeder Using Fuzzy and Optimization Technique: International Journal of Electrical Engineering & Technology, 9(4), 2018, pp INTRODUCTION Power system comprises of generation, transmission and distribution of power. After transmitting the power to substations it distributes to the consumers. Fault on distribution may lead to either overload or under load. So we avoid these problems by controlling the power and hence controlling the load. The overload problem is overcome by using load balancing technique. In this we explain the details of load balancing and steps for how to design and implement a load balancing in power distribution. Consumption of power on consumer side is highly volatile. Consumption of power varies time to time, day to day and year to year. It is difficult to save the large amount of power by using buffers; to overcome this problem, power controlling is required. Power controlling can be achieved by automatic generation control (AGC). It controls power generation as requires or when load changes and is implemented on power station side. It is highly economical for low power consumption. In that cases load balancing is preferable. Power controlling is implemented on distribution side. The basic function of the system during overloaded situations balances load from over loaded area to low loaded area and these achieved by the open/closed switches. In most of the cases, the phase voltage and current unbalances can be greatly improved by suitably arranging the connection phases between the distribution transformers and a primary feeder. It is also 74 editor@iaeme.com

2 Load Balancing of Feeder Using Fuzzy and Optimization Technique possible to advance the phase current unbalances in every feeder segment by means of changing the connection phases. In these paper loads balanced technique is designed and explained by fuzzy logic toolbox. It has many advantages and simple to understand and implantation is easier. 2. REPRESENTATION OF THE FEEDERS 2.1. Representation of the Feeder The typical distributor feeder is a three-phase, four-wire system with a radial or open loop Structure. To optimize the system development effort and the current unbalance, communication between the feeder and distribution switches must be arranged appropriately. The loads are connected in most of cases are single phase. In this paper we are assume that each feeder contains 50 loads or connections to it, so the total load to the three feeders is 150 connections as shown in figure 1. In figure each load can be connected through switches and only to the one of three phases. Figure 1 Distribution feeder 2.2. Current Phase Balancing Technique To balance the network, the engineers and technicians must change the stages manually after some field measurement. The changes made in different area affect the size of the conductor, but in most of the cases, the size of the phase conductor for the entire line of the feeder is the same. The current technique could be improved by changing the open/close of the feeder switches so that load currents can be transferred from feeder-to-feeder, i.e. from heavily loaded to low loaded feeders.. PROPOSED TECHNIQUE In this paper, a fuzzy logic-based load balancing technique along with combinatorial optimization oriented system for implementing the load changing. The flowchart of the proposed system is shown in Fig editor@iaeme.com

3 B. Phani Ranga Raja and K. Jyothi Sree Figure 2 Flow chart for proposed system Average unbalance is calculated by the equation Average Unbalance/Phase = ( LoadPh1 - LoadPh2 + LoadPh2 - LoadPh3 + LoadPh3 LoadPh1 )/3 Here the average unbalance is checked at defined time. If the unbalance is not over the defined limits, means the load is balanced within limits. So there is no need to balance the load. However, if the load is greater than the defined limits, load balancing using fuzzy logic is required. The negative value indicates that phase is overloaded and must release some load. And positive value indicates that phase is lightly loaded than limits and it can receive some load to get balanced. Once average unbalance is known, the only thing remains is to adjust that unbalance with another phase of the system to make system perfectly loaded. Now if the system is not balanced then we will need to calculate an unbalance change. So our system will have change as output that will give an exact amount of unbalance in the system. Once change is known, the exact amount of load transfer can be done using switching. This is because the actual load points for any phase might not result in an optimum combination which sums up to the exact change value indicated by the fuzzy step. So, we implement the best possible change from the implementation system and iteratively check the system unbalance until we achieve the average value below 10kw. 4. FUZZY LOGIC CONTROLLER: INPUT AND OUTPUT To design the fuzzy logic controller, first we design the input and the output variables. We choose the input as Load, i.e., the total phase load (kw) for each of the three phases, and the output as Change, i.e., the change of load (kw, positive or negative) to be made for each phase. Fig 3 shows the fuzzy logic controller in MATLAB. Figure 3 Fuzzy logic controller 76 editor@iaeme.com

4 Load Balancing of Feeder Using Fuzzy and Optimization Technique Now the next step is to define the range of inputs. So to provide fuzzy membership functions, and represent input logic in a simple way, I have divided ranges in different regions shown in table 4.1 below. Table 4.1 Ranges for Input Load Variable S. No. Input Load Fuzzy logic Description Linguistic term kw range 1 Very Less Loaded VLL 0 to Less Loaded LL 70 to Medium Loaded ML 130 to Perfect Loaded PL 200 to Slightly Overloaded SOL 250 to Medium Overloaded MOL 330 to Overloaded OL 400 to Heavily Overloaded HOL 470 to 600 In paper, the total load is 300. So as shown in above table 4.1, the system below 200 loads is considered as less loaded. System between 200 to 300 loads is considered as balanced system where no load balancing is required. Now when any system goes above its defined limits, it is considered overloaded. Here also above 300 loads, the system is considered overloaded. And no system can handle double loads than defined. So once system goes over 600 loads, the entire system should be cutoff to prevent system damage. The output changes of the fuzzy logic are related to the input. For example, if the system is medium loaded, changes should be such that the system again becomes perfectly loaded. So the output also should be divided in ranges same as input. It is shown in table 4.2 below S.No 1 Table 4.2 Ranges for Output Change Variable Output Change Description High Subtraction Fuzzy logic Linguistic term kw range HS -300 to Subtraction S -200 to Medium Subtraction MS -130 to Slight Subtraction SS -100 to 50 5 Perfect Addition PA 0 to Medium Addition MA 70 to Large Addition LA 130 to Very Large Addition VLA 200 to editor@iaeme.com

5 B. Phani Ranga Raja and K. Jyothi Sree Consider table 4.3. Here if the system is very heavily loaded, means very high load must be subtracted. So, values above perfect addition give the amount of subtraction to make system within limits from overloading. Perfect addition determines the balanced system. And if the system is less loaded, more loads should be added. So variable below perfect addition give the amount of addition to make system within limits from less loaded. To provide these ranges in input and output, double clicking on respective input or output signal box will open membership function editor. Now we have total of 8 membership functions for input Load and 8 membership functions for output Change. So we have to select 8 membership functions and give each membership function a name and a range as given in above table 4.1 and table 4.2. Below figure4 shows 8 membership functions defined for input variable Load. Figure 4 Membership functions for input Now, next step is to provide membership functions for output variable Change as defined in table 4.2. Here 8 membership functions are determined with the range of -300 kw to 300 kw. It is shown in figure 5 below, Figure 5 membership functions for output 5. FUZZY RULES AND SURFACES Now next step is fuzzy inference. Fuzzy inference provides set of rules that describe the control action of entire fuzzy system. The output changes depending upon the input and the rules defined in fuzzy inference system. So, to provide proper balancing, IF-THEN rule set is used. It means, IF (input load is this) THEN (output changes to this). Fuzzy set rules are the very important criteria to provide the effective load balancing. The rules I have used in this are described as below, table editor@iaeme.com

6 Load Balancing of Feeder Using Fuzzy and Optimization Technique Table 4.3 Rules for Input and Output variables Rule No. Rule Description 1 If Load is VLL then Change is VLA 2 If Load is LL then Change is LA 3 If Load is ML then Change is MA 4 If Load is PL then Change is PA 5 If Load is SOL then Change is SS 6 If Load is MOL then Change is MS 7 If Load is OL then Change is S 8 If Load is HOL then Change is HS Now as per the table above, rules are included in fuzzy design as shown in figure6 below. Figure 6 Ruler Editor in MATLAB Corresponding to the fuzzy input, output variables and the associated rule set, the fuzzy surface is shown in Figure 7 depicting the nonlinear relationship between the input and the output variable. Figure 7 Fuzzy surface for input and output variable 6. RESULTS The results for the load balancing are shown below. We can adjust the input load value by moving the red line shown in figure8 and we can also apply input value in the box named input editor@iaeme.com

7 B. Phani Ranga Raja and K. Jyothi Sree Figure 8 Result set for input and output variables Here in figure 5.1, the total load is 316 kw. So it means the system is slightly over loaded (SOL). So it can release some load to other feeders. Now as the total load is 300, the change should be such that total of the received load and current load is around 300 kw. It means the change must be positive and it has to be added to the actual load value. So, here change indicated from fuzzy logic is -25 kw. So total system load goes to 316 kw + (-25) kw = 291 kw. Hence, the system is balanced in this case. Now as per the control flow diagram in figure 4.1, the control logic checks the system again and finds that it is balanced now. So no further action will be needed until system goes off balance next time. It will continue checking the system at defined time intervals 7. ERROR CORRECTION In all cases, the final load after adding or subtracting a change load is not equal to 300 kw exactly. It means the system load is not fixed and changes every time. That is not possible in our system as we have assumed that the final load remains constant every time. So this method will give erroneous results. So error correction is implemented to make a system entirely balanced. The average error AE is, AE = round Now, the error matrix ΔP error is calculated by, ( The final load change after removing all the error components is given by ΔP, as shown in equation. ΔP = ΔP fuzzy - ΔP error 8. LOAD CHANGE IMPLEMENTATION SYSTEM In this section, we discuss the implementation system for implementing the load changes amongst the three unbalanced phases to achieve the balance condition. Input to the implementation system, the load change values (positive and negative), comes from the fuzzy step as described in the previous section. First, the system decides if the given load changes are at all possible considering the actual load values of the three phases. If it is possible, then the operation of the implementation system is divided into two steps, determine and distribute. ) 80 editor@iaeme.com

8 Load Balancing of Feeder Using Fuzzy and Optimization Technique In the determine step, the system selects the phases with negative load changes, i.e., load points to be released. Considering the change values, it selects optimally the sets of load points to be shifted from the releasing phases. First, for all the three phases, it is checked whether the load change value (absolute) is greater than the minimum load point of that phase. If this condition fails for at least one phase, the entire load change operation is not possible. Because, in that case, the sum of the total change will not be zero, i.e., there will be some change in the total load value which is not possible. Next, we determine the load points to be shifted for the releasing phases. We iterate over the three phases to check the negative (releasing) ones. For each of the releasing phases, first we calculate the average load value of that phase. Then, we determine the number of load points to be shifted from that phase by dividing the load change value by the average value and rounding the result. In the distribute step, the system considers the phases with positive (receiving) load changes and distributes the load points, selected in the determine step, optimally among those receiving phases. First, we concatenate the changing load points, from the determine step, into a single change vector. Then, if there is a single receiving phase, we allocate all elements of the change vector to that phase. Otherwise, if we have two receiving phases, we have to go for optimal distribution of the load points amongst the two phases for the two receiving phase case; first we get the average of the change vector. Then we divide the fuzzy change value of one of the receiving phases with this average value to determine how many load points should be shifted to that receiving phase 9. CONCLUSION Load balancing is a critical requirement of the power system to ensure that entire system works without overloading. This paper gives us a good understanding of load balancing. In this paper, we have presented a fuzzy logic-based load balancing system along with a combinatorial optimization-based implementation system for implementing the load changes. The input to the fuzzy step is the total load (kw) per phase of the feeders. Output of the fuzzy step is the load change values, negative value for load releasing and positive value for load receiving. Sum of the positive and negative values is zero, i.e., the total load remains unchanged for the entire phase balancing. The output of the fuzzy step is the input to the load changing system. It also performs the optimal inter-changing of the load points between the releasing and the receiving phases. The load balancing system is tested at the three-phase, four-wire unbalanced feeders. Application of the proposed system is substantiated by detailed application example using Matlab for the simulations. The proposed phase balancing system using the fuzzy logic and the implementation system is practically effective for reducing the feeder unbalance. The phase balancing techniques and the systems presented in this paper could be generically extended further for other distribution systems and feeder load configurations editor@iaeme.com

9 B. Phani Ranga Raja and K. Jyothi Sree REFERENCES [1] D.C. Walters, G.B. Schele, Generic algorithm solution of economic dispatch with valve point loading, IEEE Transactions on Power Systems 8 (3) (1991) [2] S. Cinvalar, J.J. Grainger, H. Yin, S.S. H. Lee, Distribution feeder reconfiguration for loss reduction, IEEE Transactions on Power Delivery 3 (3) (1988) [3] M.E. Baran, W. Kelly, State estimation for real time monitoring of distribution systems, IEEE Transactions on Power Systems 9 (3) (1994) [4] A. Ukil, M. Siti, J. Jordaan, Matlab based fast balancing of distribution system using heuristic method, SAIEE Transactions (under review). [5] A. Ukil, W. Siti, J. Jordaan, Feeder Load Balancing Using Neural Network, Lecture Notes in Computer Science 3972 (2006) [6] M. Siti, D. Nicolae, A. Jimoh, A. Ukil, Reconfiguration and Load Balancing in the LV and MV Distribution Networks for Optimal Performance, IEEE Transactions on Power Delivery 22 (2007) [7] C. C. Liu, S. J. Lee, K. Vu, Loss minimization of distribution feeders: Optimality and algorithms, IEEE Transactions on Power Delivery 4 (1989) [8] C. S. Chen, M. Y. Cho, Energy loss reduction by critical switches of distribution feeders for loss minimization, IEEE Transactions on Power Delivery 4 (1992) [9] T. P. Wagner, A. Y. Chikani, R. Hackman, Feeder reconfiguration for loss reduction: An application of distribution automation, IEEE Transactions on Power Delivery 6 (1991) [10] T. H. Chen, J. T. Cherng, Optimal phase arrangement of distribution transformers connected to a primary feeder for system unbalance improvement and loss reduction using genetic algorithm, IEEE Transactions on Power Systems 15 (2000) [11] El-Hawary ME Electric power applications of fuzzy systems. IEEE Press, Piscataway, [12] R.C. Bansal, Bibliography on the fuzzy set theory applications in power systems ( ), IEEE Transactions on Power Systems 18 (2003) [13] M.A. Kashem, G.B. Jasmon, V. Ganapathy, A new approach of distribution system editor@iaeme.com