Performance improvement of self-adaptive evolutionary methods with a dynamic lower bound

Reactive power optimisation is an important technique, which is concerned with the security and economy of operation of the power system. The appropriate distribution of reactive power can elevate voltage rating, decrease network losses and maintain network running under proper conditions. An improved evolutionary programming method with dynamic mutation and metropolis selection to solve the multi-objective reactive power optimisation under the deregulation environment is presented. The multi-objective function includes the minimisation of network losses, voltage deviation and compensation cost. To solve the problems of convergence and robustness in the conventional evolutionary programming method, the mutation operators and the selection criteria that affect the convergence and robustness are considered and a dynamic mutation and metropolis selection evolutionary programming method is suggested. Introducing chaos dynamics into mutation operators of evolutionary programming, the new method adopts the certainty method of like-stochastic to obtain mutation operators which break through the conventional thought with mutation by stochastic numbers of fixed distribution. Also, it introduces the metropolis selection in evolutionary programming to construct new selection operators. Thus, the new method not only accelerates convergence but also increases precision, so is an efficient way to optimise the capacitor banks and the adjustable transformer ratio. Tested by the IEEE-30 bus system, the method is effective.

A hybrid evolutionary programming (EP) method is presented for global optimization of complex circuits. The conventional EP is integrated with a clustering algorithm to improve the robustness of the algorithm for complex multimodal circuit optimization problems. The EP generates populations around the regions of the search space which can potentially contain a minimum but may be overlooked. The clustering algorithm is used to identify these regions dynamically. In order to improve the speed of optimization, the EP is combined with a gradient-based search method in an efficient fashion. The local search is performed from the center of each identified cluster in order to find the minimum in the region very fast. The hybrid algorithm can also reduce the search space by avoiding the search in the areas that were previously investigated. This feature greatly improves the speed of optimization and prevents the premature convergence as well. The algorithm performed very well in several benchmark problems including a test function minimization and global optimization of a complex RF diplexer circuit.

Evolutionary programming method for minimizing the number of tardy jobs

Chee-Kiong SOH 1 & Yaowen YANG 2 ABSTRACT The electro-mechanical impedance (EMI) method is a relatively new nondestructive evaluation (NDE) method which uses the electro-mechanical coupling effect of piezoceramic Lead Zirconate Titanate (PZT) transducers to measure the force impedance of the structure. Due to the electro-mechanical coupling effect of the transducers, any changes in the mechanical impedance of the structure caused by the damage will affect the electrical impedance of the transducer. By comparing the impedance spectra of the damaged structure with the baseline, i.e., the impedance spectra for the pristine structure, the damage in the structure can be assessed. The quantity of the damage is assessed by a statistically calculated damage index, such as root mean square deviation or mean absolute percentage deviation or covariance change or correlation coefficient deviation. The EMI method has shown some advantages over the conventional NDE methods. However, it is not able to simultaneously identify the location and quantify the damage. This paper presents a technique integrating evolutionary programming and the EMI method to overcome this limitation. Essentially, the proposed technique first makes use of the EMI method to measure the variations of the electro-mechanical impedance of the structure. The damage is then identified by a system identification technique which is generally employed in the vibration-based method. Due to the numerous local optima in the search space, the traditional optimization strategies may not be able to find the correct solution. Evolutionary programming (EP) is then used as the system identification technique to find the global optimum. Thus, both the location and the quantity of the damage can be simultaneously identified. In order to enhance the integrated EP and EMI method, a fitness function, which can be generally applied to the other methods, is proposed to discriminate the variations caused by damages from the discrepancies caused by modelling errors. An experimental test is carried out on a cylindrical shell specimen to verify the damage detection results. The experiment demonstrated that both the location and the extent of damage can be simultaneously identified by the integrated EP-EMI method.

Scheduling algorithm based on evolutionary computing in identical parallel machine production line

In the distribution network containing distributed generators (DG), both DG and capacitors can be adjusted for reactive power, and there is a strong complementarity between the two. If the reactive power compensation ability of each grid-connected DG can be fully utilized, it will effectively reduce the voltage fluctuation and the number of equipment actions, which will help to improve the operation level of the distribution network. In this paper reactive power optimization of distribution networks containing distributed power sources has studies. An improved evolutionary programming (EP) method is proposed to solve the optimization goal, using a variable step-size evolution method and optimization strategy with small population and high convergence accuracy. The algorithm improves the random dynamic step method. In different stages of optimization, different step sizes are used to improve the optimization effect. At the same time, considering that some wind farms are equipped with AVC and some wind farms are equipped with capacitor banks, the evolutionary operator is improved. The operator evolves directly on the integer, so it is easier to find the global optimal solution. This paper has used IEEE33 bus system to verification. Results show that the DG can reduce active power loss in distribution network and ameliorate the voltage level. Meanwhile, the improved evolutionary programming method has good optimization effect, fast convergence speed and high search efficiency.

This paper presents a new approach to solving the short-term unit commitment problem in utility system using an evolutionary programming based tabu search method. The objective of this paper is to find the generation scheduling such that the total operating cost can be minimized, when subjected to a variety of constraints. This also means that it is desirable to find the optimal generating unit commitment in the power system for the next H hours. Evolutionary programming, which happens to be a global optimization technique for solving unit commitment problem, operates on a system, which is designed to encode each unit's operating schedule with regard to its minimum up/down time. The best population is selected by evolutionary strategy. A 66-bus utility power system with twelve generating units in India demonstrates the effectiveness of the proposed approach. Numerical results are shown comparing the cost solutions and computation time obtained by using the evolutionary programming method and other conventional methods like dynamic programming, Lagrangian relaxation and tabu search in reaching proper unit commitment.

In this paper, hybrid tabu search and simulated annealing (TSSA) with search space managing methods are proposed to determine the optimal number and allocation of FACTS controllers to enhance power transfer capability of power transactions between generators and loads in power systems. Particular optimal allocation includes optimal locations and parameter settings. Two types of FACTS controllers including thyristor-controlled series capacitor (TCSC) and static var compensator (SVC) are used individually. The objective function is formulated as maximizing total transfer capability (TTC) and minimizing power losses. Power transfer capability determinations are calculated based on the optimal power flow (OPF) technique. Split and non-split search space managing methods are used to improve the solution searching capability. Test results on IEEE 118-bus system and the practical Electricity Generating Authority of Thailand (EGAT) 58-bus system show that the proposed hybrid TSSA with optimal number of FACTS criteria and the split search space managing method give higher TTC and less number of FACTS controllers than those from evolutionary programming (EP) and non-split search space method.

Distributed Generation (DG) is an auspicious solution to many power system problems such as voltage regulation, power loss minimization, effective operating power factor, etc. This paper presents a modified Artificial Bee Colony (ABC) algorithm to suit the searching for the optimal DG sizing and placement in a distribution network. The problem formulation and the proposed technique are tested on a 28-bus Malaysian distribution system. In confirming the effectiveness of this method to determine the minimum power loss which respects to its optimal size and placement, the proposed population-based meta-heuristic approach is compared with the Evolutionary Programming (EP) method for validation and accuracy. The analysis is focused on the sizing impact when considering the cost factor without neglecting the importance to maintain the minimal loss. The results have shown comparatively the same with EP in terms of the computational efficiency. Significance of savings could be achieved while having a low investment of DG installation with a minimum power loss.

Genetic algorithm (GA), evolutionary programming (EP) and evolutionary strategy (ES) are called the three kinds of evolutionary computation methods. They have been widely used in many engineering fields. However, selecting individuals directly and random search lead to produce premature problem, and requirement for high precision decreases the search efficiency, these become the obstructs of application in engineering practice. This paper proposes a new algorithm of evolutionary computation, it is called bio-simulated optimization algorithm (BSO). BSO reproduces new generation through asexual propagation and sexual propagation. Here, the evolutionary operators effectively solve the problem of premature convergence. Furthermore, performance of global search and convergence are proved theoretically. Finally, Compared BSO with GA and EP in searching the optimal solution of a continuous multi-peaks function, three kinds of computation procedures are run in Matlab, the result shows that performance of BSO is superior to GA and EP.

The unified power flow controller (UPFC) is one of the most promising Flexible AC Transmission Systems (FACTS) devices for the load flow control. Simultaneous optimisation of location and parameters for UPFCs is an important issue when the given number of UPFCs is applied to the power system with the purpose of increasing system loadability. This paper presents a mathematical model about optimal location and parameters of UPFCs to maximise the system loadability subject to the transmission line capacity limits and specified voltage level. An improved computational intelligence approach: self-adaptive evolutionary programming (SAEP) is used to solve the nonlinear programming problem presented above for better accuracy. Case studies of the IEEE 30- and 118-bus systems using the proposed model and technique demonstrate that the proposed mathematical model is corrective and efficient. Furthermore, steady-state performance of power system can be effectively enhanced due to the optimal location and parameters of UPFCs.

Distribution network planning and operation require the identification of the best topological configuration that is able to fulfill the power demand with minimum power loss. This paper presents an efficient hybridization of both Particle Swarm Optimization (PSO) and Evolutionary Programming (EP) methods which is called the Evolutionary Particle Swarm Optimization (EPSO). The proposed method is used to find the optimal network reconfiguration and optimal size of Distribution Generation (DG) simultaneously. The main objective of this paper is to gain the lowest result of real power losses in the distribution network while improve the computational time as well as satisfying other operating constraints. A comprehensive performance analysis is carried out on IEEE 33 bus distribution system. The proposed method is applied and its impact on the network reconfiguration for real power loss is investigated. The results are presented and compared with the strategy of separated DG sizing and network reconfiguration.

This paper proposes a novel hybrid method for the planning of reactive power problem (RPP). The objective of this paper is to determine the optimum investment required to satisfy suitable reactive power constraints for an acceptable performance level of a power system. Due to the discrete nature of reactive compensation devices, the given objective function leads to a nonlinear problem with combined (distinct and constant) variables. It is solved by a hybrid procedure, aiming to develop the best search features of an iterative algorithm. The performance of the proposed procedure is shown by presenting the numerical results obtained from its application to the IEEE 30-bus test network. The results obtained are compared with evolutionary programming (EP) and Broyden method to determine the efficacy of the proposed method.KeywordsReactive PowerGreedy Randomize Adaptive Search ProcedureSimple Genetic AlgorithmEconomic DispatchBelief SpaceThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

To design the lifting points of large-span steel structure when the various compatibility equations are undefined in the lifting process, the programs based on improved evolutionary programming are developed by MATLAB. Lifting points design is to determine the comprehensive optimal strategy on number and distribution of lifting points, among which the minimum strain energy theory is mentioned and the secondary development technology of ANSYS-APDL is used. The performance and efficiency of the algorithms in different optimization ideas (hiberarchy optimization and synchronic optimization) and methods (the particle swarm optimization and evolutionary programming) are compared, the results indicate that the improved evolutionary programming method based on synchronic optimization idea is satisfactory and provides a new but more effective method.KeywordsLarge-span steel structureLifting points designSynchronic optimizationEvolutionary programmingSingle-point mutation

This paper discusses the solution of optimal power flow incorporating wind generation cost. A new model of wind generation cost is presented based on the probability distribution of wind farm power output. In this model, the wind farm generation cost consists of two components: the reserve capacity cost and the environmental benefit loss. Both of the two components could be attributed to the intermittence and uncertainties of wind power. A self-adaptive evolutionary programming method is employed to solve the OPF with wind power incorporated. A case study is conduced based on the IEEE New England test system (10-Geneartor-39-Bus) as benchmark. The simulation results demonstrate the effectiveness and validity of the proposed model and method.