Reconfiguration with Simultaneous DG Installation to Improve the Voltage Profile in Distribution Network Using Harmony Search Algorithm

Due to dynamic nature of loads, total system load is more than its generation capacity that makes relieving of load on the feeders not possible and hence voltage profile of the system will not be improved to the required level. In order to meet required level of load demand, Reconfiguration & DG units are integrated in distribution network to improve voltage profile, to provide reliable and uninterrupted power supply and also to achieve economic benefits such as minimum power loss, energy efficiency and load leveling. This work proposes minimization of real power losses and improvement of voltage profile using network reconfiguration in the presence of distributed generation. Generally distributed generations (DG) are preferred with objective of minimizing real power loss and improving voltage profile in distribution system. In this work A meta heuristic Harmony Search Algorithm (HSA) is used to simultaneously reconfigure and identify the optimal locations for installation of DG units in a distribution network. Sensitivity analysis is used to identify optimal locations for installation of DG units. The proposed method has tested in MATLAB for 33-bus and 69- Bus radial distribution systems at three different load levels and the analysis is presented for loss minimization.


Objective Function
My objective function is to maximize the power loss Reduction in distribution system using simultaneous Network Reconfiguration and DG Installation problems using Harmony Search Algorithm (HSA). The objective function to maximize the power loss Reduction is described as (1)  (2)     system radial 1 -

Bulletin of EEI
The Voltage, Real power and Reactive power at respective buses are calculated by using an efficient method for Load-flow solution of radial distribution networks proposed by S. Ghosh and K. S. Sherpa [17]. Those equations are derived from the single-line diagram as shown in Figure 1.
The following equation is used to calculate the power loss between the buses K and K+1.
The total power loss of the feeder can be calculated by summing up the all individual line sectional losses of the feeder, which is given Power losses in a distribution network are reduced by using so many methods like placing capacitors at load ends, by injecting reactive power at load side using synchronous condensers, Network reconfiguration and DG installation etc. In that Network Reconfiguration and Distributed Generation (DG) installation are the best methods to minimize the power losses in distribution systems.

Power Loss Reduction using Network Reconfiguration
After network Reconfiguration of the distribution system, the power loss between the line sections is calculated as follows.
The total power loss of the feeder after network reconfiguration can be calculated by summing up the all individual line sectional losses of the feeder, which is given by The Net Power loss reduction using network Reconfiguration is the difference of power losses between before and after Reconfiguration of a distribution network, which is given by Loss  T  P  n   k  k  k  Loss  T  P  R  Loss  P  1  1  ,  ,  1 1 , ,

Power Loss Reduction using DG Installation
To improve voltage profile & minimize the power losses in distribution system, DG units are integrated at arbitrary locations. Basically DG's are integrated at load end. Integration of DG units at respective buses to improve the voltage profile is shown in Figure 2. The following equation is used to calculate the power loss by the integration of Dg units at an arbitrary location, which is given by The net power loss reduction using DG installation is the difference of power loss before and after DG installation is given by Here the question is where the DG's are installed and how to select the sizes of DG units are discussed in sensitivity analysis for DG installation topic.

Sensitivity Analysis for DG installation
A new methodology is used to determine the candidate locations for the placement of DG units using Loss Sensitivity Factors [18]. The estimation of these candidate locations basically helps in reduction of the search space for the optimization procedure Consider a line section consisting an impedance of connected between K-1and K buses as given below.
In order to calculate the active power loss in the K th line between K-1 and K buses, the values of voltage, real power and reactive power at the respective buses calculated by using equations 1, 2 and 3. The equation is as follows Now, both the Loss Sensitivity Factors can be obtained as shown below: The Loss Sensitivity Factors are calculated from the base case load flows and the values are arranged in descending order for all the lines of the given system. The line section which is having the highest LSF at which the DG unit is installed fist. Like that the locations of DG units are found out. The selection of sizes of DG units is calculated using HSA, which is discussed in application of HSA for power loss minimization topic.

Harmony Search Algorithm
Harmony Search (HS) algorithm was recently developed in an analogy with music improvisation process where music players improvise the pitches of their instruments to obtain better harmony proposed by Z.W. Geem et al. [19]. HS algorithm is simple in concept, less in parameters, and easy in implementation. So the HS algorithm has been successfully applied to various benchmarking and real world problems including traveling salesperson problem [19], parameter optimization of river flood model [20], design of pipeline network [21], and design of truss structures.
However, the major difference between GA and HS is that HS makes a new vector from all the existing vectors (all harmonies in the harmony memory), while GA makes the new vector only from two of the existing vectors (the parents). In addition, HS can independently consider each component variable in a vector while it generates a new vector, whereas GA cannot since it has to maintain the structure of a gene. The steps of HS for the generalized orienteering problem are as follows: Step 1) Initialize the problem and algorithm parameters.
Step 2) Initialize the harmony memory.
Step 3) Improvise a new harmony.
Step 4) Update the harmony memory.
Step 5) Check the termination criterion. These steps are described in the next five subsections.

Initialization of Problem and Algorithm Parameters
In Step 1, the optimization problem is specified as follows:  

Initialize the Harmony Memory
In this step, the ) ( HM matrix is filled with as many randomly generated solution vectors as the

Improvise a New Harmony
A New Harmony vector is generated by following three rules: 1) HM consideration; 2) Pitch adjustment; and 3) Totally random generation. For instance, the value of the first decision variable ) 1 ( x for the new vector can be chosen from values stored in . Value of other variables ) ( i x can be chosen in the same style. There is also a possibility that totally random value can be chosen. HMCR Parameter, which varies between 0 and 1, sets the rate whether a value stored in HM is chosen or a random value is chosen, as follows: The HMCR is the rate of choosing one value from historical values stored in HM while ) -1 ( HMCR is the rate of randomly choosing one value from the possible value range. After choosing the New Harmony vector , pitch-adjusting decision is examined for each component of the new vector. This procedure uses the PAR parameter to set the rate of pitch adjustment as follows: In the pitch adjusting process, a value moves to its neighboring value with probability of PAR , or just stays in its original value with probability a neighboring index is used for discrete-type decision variables. The HMCR and PAR parameters in Harmony Search help the algorithm find globally and locally improved solution, respectively.

Update Harmony Memory
If the New Harmony vector is better than the worst harmony in the HM , judged in terms of the objective function value, the New Harmony is included in the HM and the existing worst harmony is excluded from the HM .

Check Termination Criterion
If the stopping criterion (maximum number of improvisations) is satisfied, computation is terminated. Otherwise, Steps 3 and 4 are repeated.

Application of HSA for Poiwer Loss Minimization
Both the network Reconfiguration and DG installation problems are complex combinational optimization problems. In past many authors proposed many methods to solve the Network reconfiguration and DG installation problems separately to minimize the power losses in distribution systems. So in existing methods we will discuss about the network reconfiguration and DG installation problems independently by using HSA to minimize the

Existing Methods Only Reconfiguration Using HSA
In this, we are considering only reconfiguration to minimize the power losses in distribution system. The optimum distribution system is obtained by first generating all possible radial structures of the given network without violating constraints and subsequently evaluating the objective function. However, real distribution system contains many nodes, branches, and trees. So the conventional methods are ineffective and impractical, because of dimensionality.
Here the HSA is proven to be effective and useful approach for the network reconfiguration problem.
Let us consider a 33-Bus radial distribution system with 33 sectionalizing switches (from 1 to 32) and 5 Tie line switches (33, 34, 35, 36 and 37) which is shown in Figure 3  In this, we are considering only DG installation to minimize the power losses in distribution system as shown in Figure 5. Here we are mainly focused on optimal location and suitable sizes of DG installation. Sensitivity analysis [18] is used to calculate the optimal locations of buses to install the DG units. The Loss sensitivity factors (LSF) are calculated using load flows from base case. The LSFs are arranged in descending order and at which the bus is having highest LSF, at that place the DGs are placed first. Here we are considered only top three locations to install the DG units. In that way we can choose the number of DG's is required. The sizes of DG units are selected after the location buses by using HSA. The size of solution vector is depends on number of DG units are need to be installed.

Proposed Method (Simultaneous Reconfiguration & DG installation)
As in the above discussion we are considered Network reconfiguration and DG installation problems separately. But here we are considering the Simultaneous network reconfiguration and DG installation for the effective minimization of power losses. To reconfigure the network, initially all possible structures of given network are generated fist and by using sensitivity analysis we are finding the optimal locations for DG installation which is done simultaneously by using HSA. During optimization process the rating of DG units will vary in discrete steps at specified location by using HSA.

S os
The second solution vector HV2 is generated with the same Tie line/open switches (19,13,21,30 and 24) and same DG installation locations with different ratings is formed as shown in the Figure 7.

Test Results
In order to effective minimization of power losses in distribution systems, the proposed method (Simultaneous Network Reconfiguration and Dg installation) using HSA is applied to two test systems consisting of 33 and 69 buses. In the simulation of network, five scenarios are considered to analyze the superiority of the proposed method.
Scenario I: The system is without reconfiguration and distributed generation (Base case).

Test System 1
In this a 33-bus radial distribution system [22] with five tie-switches and 32 sectionalizing switches are considered. In the 33-bus system, the sectionalizing switches are numbered from 1to 32 and the tie line switches are numbered from 33 to37. The line data and load date of network are taken from the [9]. The total real and reactive power loads on the system are 3715 kW and 2300 KVAR. The parameters of HSA algorithm used in the simulation of network are HMS = 20, HMCR =0.85, PAR=0.3 and NI =20 and number of runs, N=9. The network is reconfigured based on the number of open switches. Using sensitivity analysis the DG units are installed at optimal locations in scenarios III, IV and V. The Network is simulated at three load levels: 0.5 (Light), 1.0 (Nominal), and 1.6 (Heavy) and simulation results are presented in Table 1.
It is clearly observed from Table 1, at light load base case power loss (in KW) is 47.06 and which is reduced to 33.27 and 23.29 using only network reconfiguration and only DG installation. The power losses are effectively reduced to 17.78 in the Simultaneous Reconfiguration and DG installation at the scenario V. Similarly the Minimum voltage is also improved to 0.9859 in scenario V, which is almost nearer to unity. Like that the power losses are also reduced at medium and heavy load also by using proposed method. The voltage profile curves drawn between Voltages with respect to the Bus nodes for all scenarios at three different load conditions is shown in Figure 9. After network reconfiguration with simultaneous DG installation, the optimal structure for scenario V is shown in Figure 10. By increase the number of DG units need to be installed we can reduce the power losses effectively.  Figure 9. Voltage profiles of 33-bus system at light, nominal, and heavy load conditions  The results are also compared with all other method like GA and RGA, but HSA gives the better results compared to other methods. This shows that for all three load levels, the improvement of voltage using proposed method (scenario V) is highest, which elicits the superiority of the proposed method over all others.

Test System 2
In this a 69-bus large-scale radial distribution system with 68 sectionalizing and five tie switches are considered. The sectionalizing switches are numbered from 1 to 68 and tile line switches are numbered from 69 to 73. Configuration, line, load and tie line data are taken from the [23].
Similar to the test system the algorithm parameters are used. Total system loads for base configuration are 3802.19 kW and 2694. 06 KVAR. Similar to test systems 1, this test system is also simulated for three scenarios at three load levels and results are presented in the Table 3. It is clearly observed from Table 3, at light load base case power loss (in KW) is 51.06 and which is reduced to 23.72 and 21.92 using only network reconfiguration and only DG installation. The power losses are effectively reduced to 11.07 in the Simultaneous Reconfiguration and DG installation at the scenario V. Similarly the Minimum voltage is also improved to 0.9860 in scenario V, which is almost nearer to unity. Like that the power losses are also reduced at medium and heavy load also by using proposed method. By increase the number of DG units need to be installed we can reduce the power losses effectively.
It is observed from Table 4, Similar to the 33 bus system, in 69 bus system also having the minimum voltage (0.9563) and maximum power loss at bus number 65 when compared to the other buses in 69bus system. By using HSA at Light load, at bus number 65, the minimum voltage is improved from the different scenarios. In that by using scenario II, scenario III, and scenario IV we are improving the minimum voltages up to 0.9841, 0.9850 and 0.9841. But by using proposed method we are improving the voltage value up to 0.9866, which is nearer to the unity. Similarly at Nominal load and Over load conditions also the bus voltages are improved effectively using proposed method.

Conclusion
In this paper harmony search algorithm is proposed for Simultaneous Network Reconfiguration & DG Installation in distribution system to minimize the Real power losses. The results show that simultaneous network reconfiguration and DG installation method is more effective in reducing power loss and improving the voltage profile compared to the different scenarios in the 33 and 69 Bus systems at three different load conditions.