Energy management carried in a power system by configuration process is a difficult activity. So, reconfiguration has been introduced to solve this problem. Numerous optimization topologies have been utilized to solve this problem so far. However, they exhibit some drawbacks such as convergence, etc. Hence to overcome this issue, this work formulated a new hybrid optimization topology Genetic Algorithm Enabled Particle Swarm Optimization (PSOGA) to solve the energy configuration problem with low power loss in the Distribution System (DS). The proposed topology’s effectiveness was evaluated on the IEEE 33 bus Distribution System, and the results were compared to methods reported in the literature. As a result, the suggested technique appears to be more successful than other approaches, and the power loss in buses is minimised and hence exhibits an enhanced voltage profile. Hence, it is concluded that the proposed PSOGA can be a promising topology for reconfiguration as well as energy management in DS.

Energy management plays a vital role in DS. Thus, to maximize the efficiency of energy management, proper configuration of DS is essential. The improper configuration may lead to increased loss of power, poor voltage. Henceforth, the network should be restructured to have less power loss with improved voltage levels [

To minimize the power loss over DS, numerous methodologies such as NR, DG, and optimal placement of capacitor have been adopted so far [

This turning ON/OFF condition mainly depends on the objective function formulated by the user [

Power loss minimization

Improvement in Voltage profile and enhance the efficiency of the DS [

As a result, several studies have been conducted to solve this distribution network reconfiguration (DNR) [

Traditional Multi-layer Perceptron (MLP) algorithm

Artificial intelligence (AI) based techniques and

Heuristic algorithms

Among those, Metaheuristic algorithms like Particle Swarm Optimization (PSO), Evolutionary Programming (EP), and Genetic Algorithm (GA) are becoming more popular to solve NR in DS. The NR is carried out with multiple objectives such as power loss minimization, voltage stability, and load margin. Rao et al. [

Prasad et al. 2007 formulated GA to find a solution for NR and balancing load condition in DS. The tie switches and section status are altered according to the load state and thus the NR is performed. Shirmohammadi et al. [^{nd} order cone program to minimize the operational cost of the system which is subjected to both financial and technical constraints. A multi-objective biased random key GA is utilized for meter allocation [

DG placement [

This work introduced a hybrid method (combination of PSO and GA) with the aim of attaining NR with reduced power loss and a faster convergence time. Thus, it utilizes the divergence property of PSO with the strengthening of GA, to modify the configuration of existing network. As a result, its quick convergence characteristic establishes its superiority over alternative topologies.

This section labels the hybrid PSOGA developed for solving NR problems.

The basic concept of GA has formulated based principles of Charles Darwin and finds the persistence of the fittest. PSO is a population-based optimization topology, which is stimulated by the behaviour of a flock of birds. Though these two methods are utilized for finding the optimal solution, they exhibit some drawbacks. GA has no memory capacity to store data. Similarly, PSO have poor individuals but exhibits memory capacity. So, the thinking capability of PSO and local search capability of GA together formulated a hybrid topology called PSOGA algorithm and it is shown in

The algorithm adopted to find the fitness function (minimization of power loss) is depicted below.

Step 1: Input the parameters (both electrical and topology).

Step 2: Generate both initial positions and velocity randomly.

Step 3: Execute load flow calculation.

Step 4: Evaluate the objective function, update velocity and check for all the constraints.

Step 5: If the constraint is not satisfied, execute GA.

Step 6: Finally compare the previous iteration.

Step 7: Finally, the termination criteria are verified. If maximum iterations are achieved, results are drawn; otherwise go to step 2.

The parameter configuration utilized for PSO is depicted below,

Acceleration coefficients (C1, C_{2}) = 1.6

Weight (W_{max} and W_{min}) = 0.9 & 0.38

Similarly for GA,

Crossover rate = 0.88,

Mutation rate = 0.02

The NR is being carried out; the key objective functions are power loss minimization and voltage profile improvement.

_{r}, _{x} and _{v}-Weight factor

_{L}-Power Loss (Active)

_{L}-Power Loss (Reactive)

CVD-Sum of deviation of gain value from its original value.

Normally, sum of all these should be equal to 1. (i.e., _{r} + _{x} + _{v} = 1)

Voltage limit should be,

V_{min} < V_{i} < V_{max}

V_{i}-Voltage at node ‘i’;

Reactive power limit can be given as

Q_{c Total }<_{ }Q_{d}

Q_{c Total}–Compensated KVAr in capacitor bank

Q_{d}-Load KVAr (demand side)

Apparent power limit should be represented as _{k }<_{ }S _{max}

where

S_{k}-‘k^{th}’ line power flow

S_{max}-Maximum allowable power flow

In this work, the standard IEEE 33-bus system depicted in

To verify the effectiveness of this proposed topology, three different types of load conditions are considered in this study are discussed below

Case (i): Light load = 0.5 pu,

Case (ii): Normal load = 1.0 pu,

Case (iii): Heavy load = 1.3 pu.

Thus, the analysis of this system before optimization is tabulated in

Factors | Total power (kW) | Power loss | |
---|---|---|---|

Real (kW) | Reactive (kVAr) | ||

Light load | 1857.56 | 47.20 | 25.40 |

Normal | 3715.23 | 202.40 | 111.45 |

Heavy loaded condition | 4829.52 | 358.00 | 197.41 |

Thus, under normal condition, the minimum voltage level of a bus is about 0.9130 pu When it is lightly loaded, it is around 0.9585 pu. Under the heavy loaded condition, it is equal to 0.8808 pu.

In this, tie switches mentioned above are reconfigured using the proposed topology and its results are represented in

Switches to be opened | Power loss (Real) (kW) | Power loss (Reactive) (kVAr) | Min. voltage (pu) | ||||
---|---|---|---|---|---|---|---|

Before NR | After NR | Before NR | After NR | Before NR | After NR | Before NR | After NR |

Case (i) | |||||||

33, 34, 35, 36, 37 | 7, 11, 28, 32, 34 | 47.20 | 31.99 | 25.40 | 23.102 | 0.95 | 0.97 |

Case (ii) | |||||||

33, 34, 35, 36, 37 | 7, 11, 28, 32, 34 | 202.40 | 109.48 | 111.45 | 97.14 | 0.91 | 0.94 |

Case (iii) | |||||||

33, 34, 35, 36, 37 | 7, 11, 28, 32, 34 | 358.00 | 224.63 | 197.41 | 166.94 | 0.88 | 0.92 |

Half of the usual load level is applied in this state. Thus, the system voltages after and before NR is depicted in

Similarly, the real power loss across the system gets decreases to 31.99 kW when compared to the base case value of about 47.2 kW. Thus, the power can be saved using this topology when compared to the base case condition. The tie switches 33, 35, 36 and 37 remain closed, and switches 7, 11, 28, 32, and 34 remain open under this proposed topology.

The system voltage profile after and before reconfiguration under case (ii) is represented in

Similarly, the power loss (real) across the system gets decreases to 109.48 kW when compared to the base case value of about 202.40 kW. Thus, the power can be saved using this topology when compared to the base case condition.

The tie switches 33, 35, 36, and 37 remain closed and switches 7, 11, 28, 32, and 34 remain open under this proposed topology.

Under case (iii), the load is increased by about 130% of the nominal load. The system’s voltage profile after and before NR is shown in

Similarly, the real power loss across the system gets decreases to 224.63 kW when compared to the base case value about 358 kW. Thus, the power can be saved using this topology when compared to the base case condition.

The tie switches 33, 35, 36, and 37 remain closed and switches 7, 11, 28, 32, and 34 remain open under this proposed topology.

Thus, the power loss reduction achieved by the PSOGA algorithm under different load scenarios is depicted in

From above

The effectiveness of the proposed topology with power loss reduction and in voltage enhancement is tabulated in

Load condition | Total real power loss (kW) | Voltage profile (pu) | Methodology adopted |
---|---|---|---|

Light load | 47.2 | 0.9585 | Base case |

- | - | EPSO | |

31.99 | 0.9715 | Proposed PSOGA | |

Normal load | 202.4 | 0.9130 | Base case |

120.7 | 0.990 | EPSO | |

109.48 | 0.950 | Proposed PSOGA | |

Heavy load | 358 | 0.8841 | Base case |

- | - | EPSO | |

224.63 | 0.9245 | Proposed PSOGA |

Similarly, to prove the efficacy of the proposed algorithm, a comparative study has been made with the existing topologies which utilized IEEE 33 bus as a test system. This study is carried out under normal load condition and their results are tabulated in

Methodology | Switches to be opened |
---|---|

PSOGA | 7, 11, 28, 32, 34 |

HAS | 7, 10, 14, 36, 37 |

EPSO | 33, 34, 35, 36, 37 |

SPSO | 33, 34, 35, 36, 37 |

Firefly-DNR, SBAT | 33, 34, 35, 36, 37 |

IGA | 7, 9, 17, 35, 37 |

IPSO | 7, 9, 17, 25, 35 |

These comparison findings show that the suggested algorithm (PSOGA) achieves better results than the other optimisation algorithms, reducing total power loss and it shows effective energy management.

Similarly, it improves the minimum voltage level from 0.8804 to 0.9510 pu irrespective of load variations in the system. Similarly, the real power was improved about 99.34%, 97.28% and 95.38%, respectively for three different kinds of load considered in this work. (Light, normal and heavy load).

The efficiency of every optimization topology is evaluated with the convergence time also. Better convergence time results in better results. A comparative study has been made in terms of execution timein

Topology | Execution time (Sec) |
---|---|

PSOGA | 4.04 |

EPSO | 12.2 |

PSO | 16 |

EP | 55 |

From the above

From the above analysis, the objective work for minimum power losses and improved voltage profile attain with least number of iterations has been observed. Therefore, it’s proved that the effectiveness of the proposed system better than the other existing optimization topology.

The PSOGA topology was effectively used to address the NR problem in this study. Its main objective is to reduce the power loss and voltage enhancement in DS. To examine the efficacy of the PSOGA algorithm, IEEE 33 bus is utilised as a test system. From the results, it is observed that the minimum voltage level of the system was improved to 0.95 p.u irrespective of load variations in the system. Similarly, the real power was improved by about 99.34%, 97.28%, and 95.38%, respectively for three different kinds of load considered in this work. (Light, normal, and heavy load). Thus, acquired results have proven the efficacy of the proposed topology for the NR problem in DS. It results in a better reduction in loss and voltage enhancement than other popular topologies. It helps to minimize the loss and the cost of energy tracking. Hence enhanced energy management can be obtained in DS. Therefore, the proposed topology is one of best method for finding a solution to large-scale NR in DS.

The authors would like to thank Anna University and also, we like to thank Anonymous reviewers for their so-called insights.