3. Evolutionaryalgorithms 3.1. Cuckoo search optimization
Cuckoo Search is an evolutionary population-based optimiza- tion method [25–29] . It is an evolutionary search which relies on natural process of birds ﬂocking for food randomly. It is based on the obligate brood parasitic behavior of some cuckoo species in combination with levy ﬂights of some birds and fruit ﬂies. It solely enhances the behavior of laying eggs and breed- ing of cuckoos. They exist naturally in two forms: matured cuckoos and eggs. Every cuckoo tries to place its egg in other nests in order of not being detected by the parent cuckoo, where it all depends on the resemblance of the alien egg and the host egg. In this step, the alien eggs are detected and being thrown out of the nest. Naturally the cuckoos often make mis- takes as the eggs resemble quite high. So, there is a probability involved in detecting the alien eggs which is used as a param- eter for the optimization algorithm. After this process, the eggs hatch and the cuckoos mature. They tend to ﬁnd a globally optimal solution or habitat. Breeding for food has always been a quasi-random process since, they are not aware of the geo- graphical location of the best habitat. The birds tend to con- verge toward the best habitat acquired by a bird in the best position. In this way, the whole population reaches the habitat. This best environment becomes their new place for breeding and reproduction. Further, breeding for food is one of the fea- tures included in Levy´ ﬂights  .
DEDP is always one of the most important research problems in the current growing scenario of power system engineering domain and as well as the case of rural electrification in renewable energy system domain. This research focused on developing certain nature inspired evolutionaryalgorithms and neural network architecture models for solving dynamic economic dispatch problems in power system environment. In power system modules, it is very important to have intelligent mechanisms in order to obtain the optimalloaddispatch solution for the test beds considered. As a result, in this research work steps are taken to obtain an optimalload solution with a satisfactory output on cost factor. Evolutionary optimization algorithms are proposed in order to obtain near optimal solutions. Also, the contribution on a hybrid renewable energy system constituted of photovoltaic (PV)-wind- diesel-battery-converter systems was found to meet the power requirement of rural area educational institution.
algorithm  have been extensively articulated to obtain the global optimal solution. The problem has been reduced to a single objective problem by treating the emission as a constraint . This formulation has some difficulty in getting the trade off relations between cost and emission. Then minimizing the emission has been handled as another objective in addition to the cost. Recently, the studies on evolutionaryalgorithms have shows that these methods can be efficiently used to eliminate the most of the difficulty of classical methods [14, 15]. Various solutions of ELD and CEELD have been reported recently in the literature [19-20]. GA has been applied on a three generator test system considering CEELD . Further in this paper GA has been extended on two different standard test systems and extensive analysis is done by presenting a short term thermal generation scheduling for the test system-1.
In this paper, we have studied the Economic LoadDispatch problem (ELD) and a recent approach, Genetic Algorithm has been proposed to solve it. From the above result analysis we see proposed algorithm has better ability to save total cost and computational time as compared to other conventional methods. GA method can overcome the disadvantage of premature convergen of conventional methods and can obtain better solution with better efficiency and convergence. Hence it can be concluded that proposed method is easier and computational efficiency is better as compared to conventional methods.
In this paper, PSO technique was applied to the Nigerian thermal power plant to solve ELD solution to obtain better converging criteria. During the iteration, acceleration constant was varied to improve the searching capability of PSO. Thus, optimal solution of the problem is obtained assuring constraint satisfaction. Thus the results obtain, using particle swarm optimization were converging in nature.
A controller is a device which monitors and affects the operational conditions of a given dynamical system. The operational conditions are typically referred to as output variables of the system which can be affected by adjusting certain input variables. Controllers are required for controlling the given plant so that the plant can perform according to our requirements. The PID controllers are widely used in industrial applications to provide optimal and robust performance for stable, unstable and nonlinear processes . Controllers must be simple and of low order so that easy analysis and required change in the system can be done. The control system performs poor in characteristics and even it becomes unstable, if improper values of the controller tuning constants are used . So it becomes necessary to tune the controller parameters to achieve good control performance with the proper choice of tuning constants. The basic function of controller is to use a proposed tuning algorithm and to maintain the output in the desired or required range of output. Setting of the proportional, integral and derivative values of a controller to get the best possible control for a process using a tuning algorithm is called tuning of a PID controller .
V. GENETIC ALGORITHM
Genetic algorithms  are search algorithms based on the process of biological evolution. In genetic algorithms, the mechanics of natural selection and genetics are emulated artificially. The search for a global optimum to an optimization problem is conducted by moving from an old population of individuals to a new population using genetics-like operators. Each individual represents a candidate to the optimization solution. An individual is modeled as a fixed length string of symbols, usually taken from the binary alphabet. An evaluation function, call fitness function, assigns a fitness value to each individual within the population. This fitness value is measure for the quality of an individual. The basic optimization procedure involves nothing more than processing highly fit individuals in order to produce better individuals as the search progresses. A typical genetic algorithm cycle involves four major processes of fitness evaluation, selection, recombination and creation of a new population. Although the binary representation is usually applied to power optimization problems, in this paper, we use the real valued representation scheme for solution. The use of real valued representation in the GA is claimed by Wright to offer a number of advantages in numerical function optimization over binary encoding. Efficiency of the GA is increased as there is no need to convert chromosomes to the binary type; less memory is required as efficient floating-point internal computer representations can be used directly; there is no loss in precision by discretisation to binary or other values; and there is greater freedom to use different genetic operators. For the real valued representation, the k-th chromosome
The economic loaddispatch plays an important role in the operation of power system. The main objective of this paper is to determine the optimal combination of power outputs of all generating units so as to meet the required demand at minimum cost while satisfying all types of constraints. In this paper the lambda iteration method and the two main types evolutionary optimization technique genetic algorithm and particle swarm optimization which are generic population based probabilistic search optimization algorithms and can be applied to real world problem are respectively applied to solve an ELD problem and at last the comparison between all three method has been presented. The PSO provides the generation level such that the generation level is coming out to be lower than the cost resulted with genetic algorithm method.
search. The algorithm reduced total execution time, and was tested with 3, 13 and 40 units.
The EA-based (and by extension, the AI-based) techniques for solving power system optimisation as reviewed so far have covered problems involving smooth and non- smooth functions, encoding type, genetic operators, line losses consideration, constraints (including generators limits, power balance, ramp-rates and prohibited operating zones), computational resources, valve-point loading effects, multiple fuels, number of generating units, environmental consideration, etc. We described and evaluated a new EA approach for solving the optimal flow problem for ELD in the electricity generation industry in . The method combined a standard EA with smart mutation and hill-climbing techniques, and considered benchmark instances of the ELD problem involving minimum/maximum generation limits, power balance, ramp-rates and prohibited operating zones. Violation of either of these constraints introduces the concept of penalties, and these in turn provide the basis for the smart mutation operator. Our smart EA was compared with a basic EA and reported results for other recent algorithms, on three benchmark cases involving 6, 15 and 20 generating units. On the larger two of these problems we find better solutions than have so far been reported in the literature. In later chapters, the review will extend to dynamic problems and bidding context in a deregulated power market. However in the next section we focus on literature in the EA area that pertains to the ‘smart mutation’ approach explored in this thesis.
Based on the behavior of the bats, developed by Yang  is an interesting optimization technique called Bat Algorithm, which is an optimization algorithm, inspired from the echolocation of microbats. The capability of echolocation of bats can find their prey and perceive different types of insects in complete darkness. In most of the cases they use short frequency signals to analyze the object. Echolocation of bats works as sonar in bats, emits a loud and short pulse sound, and it hits an object after a fraction of time, the echo returns back to their bats ears. In addition, this makes bats being able to distinguish the difference between an obstruction and a prey, allow them to hunt even in complete darkness. Bat algorithm is a considered as a new metaheuristic algorithm based on Bat behaviour. The optimal solution of economic loaddispatch (ELD) is obtained and it is used to optimize the total generating cost of power plant by using the proposed bat algorithm. This bat algorithm has good convergence and better quality of solution than PSO and IWD reported in . The main advantage of this technique is easy is implement and capable of finding feasible global optimal solution. BAT Algorithm has priory of accuracy and better efficiency compared to other algorithms.
Intelligent Water drop (IWD) technique for load balancing is swarm based optimization technique. It is inspired by natural rivers and how they find almost optimal paths to their destination. These near optimal or optimal paths follow from actions and reactions occuring among the water drops and the water drops with their riverbed . In the IWD technique , several artificial water drops cooperate to change their environment in such a way that the optimal path is revealed as the one with the lowest soil on its links.IWD has two important properties. Firstly, soil i.e the amount of soil it carries. Secondly, velocity i.e the velocity with which it is moving . The IWD soil is increased by removing some soil of the path joining the two locations. The amount of soil added to the IWD is inversely proportional to the time needed for the IWD to pass from its current location to the next location. This duration of time is calculated by the simple laws of physics for linear motion. Thus, the time taken is proportional to the velocity of the IWD and inversely proportional to the distance between the two locations. Another mechanism that exists in the behavior of an IWD is that it prefers the paths with low soils on its beds to the paths with higher soils on its beds. Uniform random distribution is used to implement the behavior of path choosing among the available paths of soil. The lower the soil of the path, the more chance it has for being selected by the IWD .
In the above subsections, some hybrid methods have been discussed that are used to obtain the optimized solution for ELD problems. The researches confirm that the PSO method itself can be used as a powerful and useful technique for obtaining the optimal solution. If the global best and local best positions are identical, the algorithm might get stuck into local optima. It is a major drawback of PSO. To eliminate this drawback, many hybrid methods combining PSO with other global optimization algorithms like GA, IF, EP, FA, ABC, GSA has been formulated. In addition to above formulated hybrid approaches, a new hybrid PSO is suggested by the author (Santra, et al., 2016) . This approach combines the PSO and ACO algorithm together. Here new solutions or members can be generated at each iteration using PSO and then ACO can be applied for the fine-tuning of the members. 3. DISCUSSION
However the papers approach solely doesn‟t rely on evolutionary techniques, as for the purpose of transmission loss computations at various steps of algorithm Load flow analysis has to be done. Load flow computations have been done using MATPOWER 6.0 , which uses Newton Raphson method itself.This toolbox also helped in deciding the MVAr output of the STATCOM device being used .
Reactive power management means to procure economic reactive power and dispatch it optimally. Many researcher has calculated procurement cost for reactive power from different reactive power sources like synchronous generator, Synchronous condensers, Capacitors and FACTS devices. And total cost is minimized by considering reactive power transmission charges (RPTC). Reallocation of reactive power generation enhances voltage stability which can be achieved by changing generator voltages, transformer tap settings, and switchable reactive power sources in system.By providing reactive power need locally and by rearranging of reactive power supply in the system, losses can be minimized. ORPD problem can be solved by optimizing various functions like system cost minimization, improving the voltage stability, losses minimization. The reactive power dispatch problem can be resolved optimally by Non-Linear Programming technique which has many limitations as long execution time, insecure convergence properties, and algorithmic complexness. While in, the gradient-based methods gives local minima and the solution obtained will not be the optimal one. And in sensitivity analysis method, the linear objective function is used and constraints are around an operating point[2-5]. In the third type of method -Heuristic methods have been implemented to get impressive optimal solution in the problem space .
Economic loaddispatch is a challenging task in power system. ELD model characteristic should be nonlinear due to presence of valve point loading effect and presence of various constraints. In the proposed work two case study are considered, the ELD problem with valve point loading effects is solve by using various variants of PSO.The test results obtained by MRPSO shown table 2 and table 4, the rsults demonstrated that the proposed MRPSO algorithm is capable of achieving global solution, it is computationally efficient and give better optimal results than other PSO methods. Overall, the MRPSO algorithms have been shown to be very helpful in studying optimization problems in economic loaddispatch problem.
2. Related work
Traditionally the problem of ELD was solved by using Gauss-Siedel or Newton-Raphson methods in combination with Lagrangian multiplier method [1, 2]. The problem with these methods is that the convergence depends on initial guess, size of the system and the possibility of local minima. The fuzzy optimization was implemented to solve the same challenging problem [5-7]. However, all these aforementioned papers solve this problem using single objective optimization techniques. These methods required the algorithm to run many times to get all the Pareto optimal points. To solve this challenging and interesting problem researchers have proposed multi- objective optimization algorithm such as particle swarm optimization, non-dominated sorting genetic algorithm etc. and solved the problem at one run [9-15]. In this paper, we have applied three techniques of weighted sum approach to solve this problem. B.Y. Qu et al. have applied multi-objective evolutionary programming to solve environmental economic dispatch problem . A.Bhattacharya et al. have applied gravitational search algorithm for multi-objective optimal power flow in 
The goal of MO optimization is two-fold: finding an ap- proximation set of non-dominated solutions that is close to the Pareto-optimal front (i.e. proximity) and as diverse as possible (i.e. diversity, especially in the objective space) . Standard MOEAs steer the population toward the optimal front while trying to preserve the diversity by different mech- anisms, such as the selection based on crowding distance in NSGA-II  or the environmental selection in SPEA2 . However, it has been showed that these mechanisms are in- sufficient for achieving a good scalability and that differ- ent parts of the optimal front should be processed sepa- rately . State-of-the-art MOEDAs therefore often im- plement mixture probability distributions by clustering the selected solutions in the objective space and building a link- age model for each cluster separately (e.g. in mohBOA  and in MAMaLGaM ). Studies [10, 11] have noted the difficulty for finding the entire optimal front of some de- composable problems, especially the extreme regions of the optimal front, as for the studied problems the niches on the extremes become exponentially smaller than the niches in the middle. Furthermore, for MO optimization in gen- eral, selection tries to exploit all objectives simultaneously, thus reducing the pressure towards approaching the optimal
Abstract - Multicore hardware systems are proving to be more efficient each passing day and so are the scheduling algorithms for these systems. The potential speedup of applications has motivated the widespread use of multiprocessors in recent years. Optimal multiprocessor scheduling algorithms remain a challenge to the researchers. Out of the number of algorithms proposed and analyzed we here compare and examine three of them: the classic global EDF, the optimal P-fair algorithm and a newer LLREF which has worked upon the strengths of P-fair. They are compared in terms of task migrations and required number of scheduler invocations and schedulability of a variety of tasks. Results are verified on the basis of a set of randomly generated tasks.
University of Zaragoza, Spain. Although these algorithms are suitable for many aspects related to Computer Science studies, and can be very powerful to solve relatively complex problems, they are usually presented in a series of lectures in which theories and concepts are simply communicated to the students. Therefore, it is very common to observe that students have difﬁculties to correctly understand the operation of EAs and Evolution Strategies for solving problems, and how to apply them to solve problems based in real cases. Additionally, the scope and limitations of most of these mechanisms are usually presented by professors only in a theoretical way, which does not help too much students to understand them adequately. To address this issue, many authors have proposed to apply useful techniques and applications (e.g., interactive games, tutorials, AI-based applications, etc.) -. These approaches allow professors to present theoretical concepts to the students in a more interactive and attractive way.
The approach presented in this paper was applied to economic emission loaddispatch optimization problem formulated as multiobjective optimization problem with competing fuel cost, and emission. The algorithm main- tains a finite-sized archive of non-dominated solutions which gets iteratively updated in the presence of new solutions based on the concept of ε -dominance. More- over, local search is employed to explore the less- crowded area in the current archive to possibly obtain more nondominated solutions. Also to identify the best compromise solution Topsis technique was applied by incorporating relative weights of criterion importance. The following are the significant contributions of this paper: