The electricity demand is increasing,

and the need for electrical energy is growing in India as well as in other

countries. Energy demand is increasing 4.9 % per year on an average from 900

Terawatt-Hours (TWh) in 2013 to almost 3300 TWh by 2040 1. India holds for almost 17 % of the increase in global electricity

demand from 2013 to 2040 2. Renewable sources

like wind power, solar power, etc. are the reliable and the cost-effective

solution for the sustaining energy shortages 3.

Wind power is emerging as the fastest developing technology and provides the

largest share of the distributed generation in electrical power generation. The

installed capacity of the wind power generation has been increasing in the

world at an annual growth rate of about 20 % 4-5.

The major disadvantage of the power

generation from wind power is that its output is intermittent in nature. The

dynamic nature of wind power causes mal-operation of relays which affects the

stability and security of the system 6. The

wind generators of the wind farm functions only when the wind speed is between

the cut in and cut out speed of the wind turbine and the wind farm must be

always connected to the grid. Normally the settings of the relay are fixed by

assuming that the wind farm is connected to the power grid at its rated capacity

7. The penetration of distributed generated

power changes the conventional distribution system’s short circuit power, fault

level, amplitude of fault current and direction due to the variation in source

impedance 8. The addition of intermittent wind

power causes mal-operation in the differential relay, over reach or under reach

problem in distance relays and over current relay coordination problems 9-10.

The over current relays are the most commonly used

protection scheme for the feeders as they are economic and effective. The over

current relays have two settings; pick up current settings and Time Multiplier Settings

(TMS). To determine the above settings, two different methods are commonly

adopted, i.e., conventional approach, and optimization techniques 11. The conventional method is based on the

predetermination of all fault currents, abnormal conditions, system contingencies,

etc. The conventional methods are based on the network topology which includes

graphical selection procedure 12,

identification of Minimum Break Point Set (MBPS) using expert set 13 and linear graph theory 14.

The proper relay coordination is ensured only if the operating time of the

secondary relays is greater than the corresponding primary relays. The

conventional method is not applicable in large systems with contingencies

including generator outage, line outage etc., as it is time consuming. Relay

coordination in power network having multi generators including DG sources

becomes unfeasible through conventional optimization technique. Optimization

approaches solve the protection coordination problem by minimizing the

objective function subject to the constraints by maintaining the Coordination

Time Interval (CTI). The coordination of over current relays is formulated as

the optimization problem using linear programming approach in 15 using simplex and two phase simplex methods. In these

methods, the pickup current settings are predetermined and the operating time

of the relay is considered as the linear function of TMS. These optimization

techniques are not capable of handling the complicated problems and the

obtained results may be trapped in local minimum values 16. More over the accurate predetermination of the dynamic nature

of pick up current setting is not possible due to changes in network

topologies. When both the pickup current setting and TMS are to be determined

simultaneously, the coordination problem becomes Non Linear Programming (NLP)

problem. In 17, NLP problem is solved using

Gauss-Seidel iterative procedure. Subsequently the application of various

heuristic optimization techniques have also been proposed including genetic

algorithm approach 18, particle swarm

optimization technique 19, differential

evolution algorithm 20, ant colony

optimization 21, etc. These evolutionary

algorithms get trapped into local minima and as the result these algorithms

need to be run repeatedly by varying the parameters. In medium and large

interconnected networks, it is very difficult to satisfy all the coordination

constraints 22. Thus the relay coordination

problem is reformulated as unconstrained programming problems by using penalty

functions for avoiding constraint violations 23.