Real power loss diminution by rain drop optimization algorithm

Received Apr 29, 2020 Revised Mar 15, 2021 Accepted Apr 16, 2021 In this work Rain Drop Optimization (RDO) Algorithm is projected to reduce power loss. Proceedings of Rain drop have been imitated to model the RDO algorithm. Natural action of rain drop is flowing downwards form the peak and it may form small streams during the headway from the mountain or hill. As by gravitation principle raindrop flow as stream then as river form the peak of mountains or hill then it reach the sea as global optimum. Proposed Rain Drop Optimization (RDO) Algorithm evaluated in IEEE 30, bus test system. Power loss reduction, voltage deviation minimization, and voltage stability improvement has been achieved.


INTRODUCTION
Real power loss reduction is main objective of this work. Various methods conventional [1]- [6] and Evolutionary techniques [7]- [19] are applied to solve the problem. Rain Drop Optimization (RDO) Algorithm applied to reduce the power loss. Rain drop actions have been imitated to model the algorithm. Natural behavior of rain drop is flowing downwards form the peak and it will form small streams during the progression from the hill. With reference to the gravitation the raindrop flow as stream then as river from the peak of mountains or hill then finally it reaches the sea. Reaching the sea by passing many areas including valleys is global optimum. Raindrops are engendered in arbitrary mode in the initial iteration itself, then each rain drop will assign a neighborhood by itself then in arbitrary mode neighbor points are produced. Projected algorithm begins with arbitrarily engendered solution afterwards exploration has been done sequentially around the present point until end of the end criterion and also revision of the present value will be there throughout the procedure. Rain Drop Optimization (RDO) Algorithm evaluated in standard IEEE 30, bus test system. Voltage deviation and power minimization achieved along with voltage enhancement.

PROBLEM FORMULATION
Solving the optimal reactive power dispatch (ORPD) problem plays a significant role in the efficient operation and planning of the power system. The aim of solving the ORPD is to determine the best operating point of system for maximizing the voltage stability, minimizing the system loss and the voltage deviations. The best operating point includes the terminal voltages of the generators, taps of transformers and the injected reactive powers of the shunt compensators. The solution of the ORPD problem is formulated as an Fitness function (F 1 ) (6) is defined to diminish the power loss (MW), Fitness function (F 2 ) minimization of Voltage deviation is (7), Fitness function (F 3 ) (8) voltage stability index (L-index) is (9), (10), (11).

RAIN DROP OPTIMIZATION ALGORITHM
Natural behavior of rain drop is flowing downwards form the peak and it will form small streams during the progression from the hill [20]. In the projected algorithm rain drops will be a particle in the population, and it has been described as (24), DP i symbolize rain drop, number of variables indicated by "n" and Y i,k indicate the k th variable in the problem, size of the population is defined by "m". DP i has been assumed as a vector or point in the "N" dimensional axis and through uniform distribution it has been engendered as function with constraints as (25).
No main point for a single rain drop RD i then predominantly in the position of stationary. Then the rain drop has been take out of this situation through a procedure of explosion is defined by (30),

NHP exploration = neighbor point × explosion base × explosion counter
The ranking of the rain drops in iteration's is given by (31) and (32), order(D1 t i ), order(D2 t i ); are orders at iteration t (33) ω 1 , ω 2 ; are weight cofficients and ranking t i is rank of rain drop RD i Raindrops are engendered in arbitrary mode in the initial iteration itself, then each rain drop will assign a neighborhood by itself then in arbitrary mode neighbor points are produced. Throughout the procedure of the optimization the neighbor point will be within in the exploration space limits and if any engendered beyond the exploration space limit then it has been modernized by (34) and (35). Subsequently for each rain drop and its neighbor points cost function will be computed then the comparison of values between rain drop and neighbor point will be done when there is dominant or main neighbor point found then the rain drop move towards to that point by altering its position. then go to step "n" − Otherwise exploration process will be applied − Fix the rain drop as live − Swap the rain drop with dominant neighbor point − Drop number +1 − Is drop number <= number of population? If yes go to step "g" − Otherwise iteration +1 − Even after applying the exploration process no neighbor point found means then that particular rain drop is marked as motionless − After creation of the worth order list then lower order value rain drops are marked as motionless − Any live or active rain drop found ; Itertaion <= maximum itertaion − If yes go step "f" − Otherwise go to next step "w" − For all rain drops the cost function value will be computed − Discover the rain drop which possess the minimum cost functional value − Then optimal solution is the cost and raindrop position − End

SIMULATION RESULTS
Projected Rain Drop Optimization (RDO) Algorithm evaluated in standard IEEE 30 bus system [21]. Active and reactive power consumption is 2.8340 and 1.2620 per unit in 100 MVA base. Table 1 and  Table 2 shows the parameters [21]; then Table 3 to Table 4 shows the comparison results. Figures 1-4 gives the graphical comparison.  revision of the present value will be there throughout the procedure. Throughout the procedure of the optimization the neighbor point will be within in the exploration space limits and if any engendered beyond the exploration space limit then it has been modernized. Rain Drop Optimization (RDO) Algorithm evaluated in IEEE 30, bus test system. Voltage stability enhanced, power loss reduced with voltage deviation minimization.