Search

Evolutionary algorithms show outstanding performance when they are applied to optimization problems with a few variables, i.e. problems with less than a hundred continuous variables. Large-scale global optimization with continuous variables is still a challenging task for a wide range of evolutionary algorithms. Their performance decreases when the number of variables increases because the search space grows exponentially. Classic evolutionary algorithms cannot find a good solution using the allocated resources. A cooperative coevolution approach is a good tool for increasing the performance of an optimizer in solving high-dimensional problems. The approach splits the objective vector into a predefined number of parts (subcomponents), and each part is optimized by its optimizer. This paper makes an effort to solve the problem of selecting the number of subcomponents. The paper represents a novel metaheuristic for solving optimization problems with a huge number of continuous variables. The suggested approach is based on the self-adaptive cooperation of algorithms and the cooperative coevolution approach. Each algorithm has a unique number of subcomponents. The metaheuristic automatically allocates resources between algorithms during the optimization process. Algorithms optimize the same population one by one. The proposed metaheuristic is titled COSACC, coordination of self-adaptive cooperative coevolution algorithms. We have evaluated the proposed algorithm on fifteen problems from the IEEE LSGO CEC'2013 benchmark. The study demonstrates that COSACC outperforms in average cooperative coevolution algorithms with the static number of subcomponents. Wilcoxon test has proven the results of numerical experiments. We have tested COSACC performance with other state-of-the-art metaheuristics, COSACC is a competitive approach. [ABSTRACT FROM AUTHOR]

The term optimization refers to find the best possible solution. Gradient computation and Nature evolution rule are the two main bases of almost all optimization algorithms. Single objective optimization problems deal with optimization of one objective function whereas multi-objective problems deal with two or more objective functions to be optimized. Many algorithms have been developed till now for the optimization of multi-objective problems. In this paper, we will review and compare two multi-objective evolutionary algorithms SPEA2 (Strength Pareto Evolutionary algorithm) and PAES (Pareto Archived Evolution strategy). Evolutionary algorithms use fitness function, selection, crossover, and mutation procedures to generate a new population. SPEA2 algorithm uses the population set as well as an external set known as an archive to select and generate the next population members. PAES uses a local search approach to generate new solutions and maintains an archive to refer to previously generated solutions and thereby find the dominance ranking of current and newly generated solutions. Four sets of problems (DTLZ1, DTLZ2, ZDT1, ZDT2) will be solved using these algorithms and the results of these algorithms on these test problems will be compared in this paper. [ABSTRACT FROM AUTHOR]

The paper considers a new approach to the identification of dynamic objects in the form of the first-order differential equations system. The peculiarities of the proposed approach are the absence of restrictions on the structure of the differential equations included in the system and the symbolic representation of the obtained solution. The proposed approach includes a self-configuring genetic programming algorithm for selecting the structure of differential equations. The number of algorithms corresponds to the number of equations in the system. The method of differential evolution is used to optimize numeric constants. The authors apply a self-configuring procedure for the indicated evolutionary algorithms. The proposed approach has been tested on a variety of problems. The accuracy of the obtained solution has been investigated depending on the presence of noise in the sample data. [ABSTRACT FROM AUTHOR]

The study here summarizes the work growth in "Particle Swarm Optimization (PSO)" since (1995 to 2021). Thirty-five papers have been analysed. Those research articles have been divided into various aspects in which PSO technique has been modified a lot during these years due to its simplicity. The hybridization of PSO with many evolutionary algorithms has been done during these years which helped us a lot to solve many optimizations and real-life problems. This paper discusses the progress, modification, improvement and application of PSO during these years. [ABSTRACT FROM AUTHOR]

This paper discusses a review structure in modified Maximum Power Point Tracking (MPPT) for a single phase Z-source inverter (ZSI) when connected to a photovoltaic (PV) module. Both soft computing MPPT mechanisms, Artificial Intelligence (AI) and Evolutionary Algorithm (EA) are very effective in dealing with nonlinear I-V characteristics of PV module. ZSI can be used to buck or to have boost operation at the inverter ac output voltage when the duty cycle is changed. In this paper, the selective MPPT control algorithms have been discussed in order to understand which controllers are the most effectively applied to maximize the power generation at the inverter output. At the end, this paper will tabulate a comparison between the AI and EA techniques which could reduce the complexity of implementation, tracking efficiency, control strategy and reliability towards ZSI. These tables will help in recognizing the suitable MPPT controllers that can be applied to a specific load for a new research exploration in the future. [ABSTRACT FROM AUTHOR]

Educational institutions frequently struggle to develop a workable lecture/tutorial schedule in a major universitydepartment. In this study, an evolutionary algorithm (EA)-based method is used to resolve the university scheduling problem. Throughout the process, a chromosomal representation tailored to the task is used. In order to generate workable timetables in a fair amount of computation time, heuristics and context-based reasoning have been applied. The strategy used to accelerate convergence is clever adaptive mutation. Real data from a large university has been used to validate, test, and debate the whole course scheduling system mentioned in this paper. Higher education institutions are currently concerned about students' attendance patterns. For the faculty to construct schedules automatically, a computer-based automatic timetable generator system is required. Our ideas for the "Automatic Time Table System," which consists of many applications, have been laid out. The application's timetable generation feature helps you save time. The system's automated Excel spreadsheet is utilized to maintain a record of the teachers' availability and scheduling. Our students will find this system helpful in the twenty-first century. It takes a lot of time and labor for educational staff in universities with huge student numbers to manually create a schedule. Due to this, it commonly happens that the same professor will give numerous lectures or competing classes. Automatically generating test and class schedules will be made simpler by the availability of a timetable generator. It will be produced by the system automatically, which will also help you save time. It prevents having to manually set up and manage a schedule. The purpose of this project is to develop a simple, functional, and user-friendly application that will facilitate the construction and distribution of schedules. The genetic algorithm is themain technique used to make schedules. It helps in the development of the ideal schedule by regulating all the rules. Conflicts in scheduling are not an issue for the faculty. With inputs such teacher name, data for the rooms, labs, and subject, this system will have a user-friendly, interactive, and less complicated interface. To store all the data entered as input, the system will have a carefully constructed database. The system will have an algorithm that manages all of the database's data while taking into account both hard and soft constraints. a timetable generation technique that seeks out the optimal solution using genetic algorithms. Instead of requiring time-consuming documents, the staff and students can easily view the timetable. [ABSTRACT FROM AUTHOR]

Due to the high labour costs in industrialised countries and the increasing demands on the quality of plastic moulded parts produced by the injection moulding process, companies are challenged to increase their production efficiency. Data-based approaches provide an opportunity for systematic and objective process set-up and control. Especially the process set-up can be a tedious, iterative and complex task which can consume a significant amount of time in the preparation of mass production. Artificial neural networks (ANN) can be used by evolutionary algorithms to search for suitable machine setting parameters for the desired part and process quality features. This often results in locally optimal solutions, which do not guarantee the best possible machine settings. Invertible neural networks (INN) could be a solution as with the help of these models, globally optimal solutions may be identified for definitive optimisation as the entire solution space for given quality parameters can be mapped in short time. In our approach, data for the implementation of the INN is collected in practical tests. The resulting weight and several geometric dimensions of the parts are determined as quality parameters. Based on these data, ANN are trained and utilized for the generation of synthetic data for the training of the INN. The creation of synthetic data is necessary because INN have millions of trainable parameters and therefore degrees of freedom which need to be determined using a bigger amount of training data in comparison to traditional feedforward ANN. In this paper, two different databases with a size of 10,000 samples each are being investigated. The two databases differ in the distributions of holding pressure and cooling times in order to investigate the influenceability of the prediction quality of the INN by training data. The sample settings for the injection velocity, packing pressure, mould temperature and melt temperature are uniformly distributed. Two different types of conventional and conditional invertible neural networks are implemented based on literature recommendations. Measured on validation data, the model quality (R² value) for the prediction of the part weight is 0.995 and 0.975 for the part dimensions. The machine setting parameters proposed by the INN can be analysed in regard to their impact. The holding pressure time is identified as the machine setting with the greatest influence on both part weight and part dimension. The predictions of the INN are validated in practical experiments with an R² value of 0.997 for the part weight. The weight of the parts produced with the proposed machine settings deviates on average by 0.24 % from the target. [ABSTRACT FROM AUTHOR]

The convoluted and dynamic nature of cancer growth makes it difficult to predict long-term cancer survival. Evolutionary deep learning techniques have been growing in popularity as efficient and accurate prediction techniques in recent years. Long-term cancer survival prediction has been developed to utilize evolutionary deep-learning techniques. This research combines the power of evolutionary algorithms using Particle Swarm Optimization (PSO) with some deep learning techniques to improve the performance and accuracy of the deep learning models by optimizing its parameters and minimizing the error between the predicted and actual outputs. Overall, combining evolutionary algorithms with deep learning models for cancer survival analysis prediction demonstrates the potential to significantly improve the accuracy and efficiency of cancer survival predictions, leading to better treatment decisions and improved patient outcomes. This paper uses the gbsg2 survival analysis dataset, a well-known dataset commonly used in survival analysis studies. [ABSTRACT FROM AUTHOR]

Modern artificial neural network architectures and training algorithms are able to achieve high accuracy in a wide range of problems. However, training multilayer neural networks using backpropagation or evolutionary algorithms might take a large amount of time. Extreme learning machines (ELMs) are aimed to resolve this problem by excluding training from the neural network model creation process by randomly initializing weights between input and hidden layers and computing weights between hidden and output layers. However, random weights initialization might lead to suboptimal results produced by the network. In this paper, we apply biology-inspired algorithms, including genetic algorithm with tournament selection, particle swarm optimization, and chaotic fish school search with exponential step decay, to the selection of input weights in ELM, the obtained ELM configurations are applied to solve regression problems. The results of the study show, that population-based algorithms can improve ELM accuracy. [ABSTRACT FROM AUTHOR]

Forecasting the bank profile shape of stable hydraulic channels using empirical, experimental and numerical models is of special consideration among fluid mechanic and river science engineers. In the present paper, the application of soft computing methods is evaluated in predicting the geometry of stable channel cross sections. In this way, using a combination of the Particle Swarm Optimization (PSO) algorithm with an Adaptive Neuro-Fuzzy Inference System (ANFIS) model, a novel evolutionary system called ANFIS-PSO is presented. The evolutionary model performance is assessed in comparison with a simple ANFIS model. The coordinates of points located on a channel boundary in stable state were also measured by the authors using a sensor instrument in the laboratory at 4 different flow discharge rates of 1.157, 2.18, 2.57 and 6.2 l/s. The results indicate that the evolutionary ANFIS-PSO model with Root Mean Squared Error (RMSE) and Mean Absolute Relative Error (MARE) of 0.0132 and 0.1326 performed better than the ANFIS model with 0.026 and 0.1426 error values respectively (almost 97% and 10% decrease in RMSE and MARE value for the ANFIS-PSO model, respectively). This demonstrates the high ANFIS-PSO model accuracy in predicting bank profile characteristics. The robust evolutionary ANFIS-PSO proposed can be used in designing and estimating stable channel dimensions. The second-degree polynomial equation proposed by the evolutionary ANFIS-PSO model can be utilized in predicting the coordinates of other points located on a stable boundary of a channel cross section. [ABSTRACT FROM AUTHOR]

The efficiency of the systems for the automated formation of artificial neural networks is beyond any disputes. Such systems make it possible to automatically generate ready-to-work models without participation of a domain expert while creating a model. The present paper considers a modification of the evolutionary algorithm for the automated formation of artificial neural networks of direct propagation, which helps build recurrent neural networks. The peculiarity of the algorithm lies in the network structure encoding in the form of a tree allowing the network structure to be represented more compactly, that well correlates with text analysis tasks, the dimensions of which vary from hundreds to thousands of features. To test the approach, the problems of classifying textual information were solved. The averaged test results as well as some of the obtained models are presented to demonstrate the approach. It is showed that the algorithm successfully builds models with a recurrent structure. This approach can be potentially improved by adding various recurrent links. [ABSTRACT FROM AUTHOR]

The Hybrid Flow Shop Scheduling (HFS) topic has received wide attention regarding problem modeling and solution techniques. The HFS problem has been extensively researched due to its importance and complexity. This paper presents an optimization scheme for a case study on minimizing makespan and energy consumption in a hybrid flow shop scheduling problem. The HFS consists of a few productions line with several parallel machines in one or more stages. The significant difficulty is assigning work to specific devices at different locations. A case study was implemented with twenty-seven jobs in four stages comprising bending machines, lathe machines, milling machines, and wire cut machines. The (EE-HFS) optimization has been carried out using Multi Objectives Dragonfly Algorithm (MODA). The optimization result was compared with well-established algorithms, the Pareto Envelope-based Selection Algorithm II (PESA2), Multi Objectives Evolutionary Algorithm Based on Decomposition (MOEA/D), and new algorithms, Multi Objectives Grasshopper Optimization Algorithm (MOGOA) and Multi-Objective Ant Lion Optimizer (MOALO). Five metrics were employed for comparison purposes, Error Ratio (ER), Spacing (SP), Maximum Spread (MS), computational speed (CS), and Generational Distance (GD). The MODA outperformed other algorithms based on the convergence metric and was on par with different algorithms for spacing and maximum spread metrics optimization results. In conclusion, these results are significant for Hybrid flow shop scheduling regarding the energy utilization model and MODA's excellent potential to be implemented further in other combinatorial scheduling problems. This case study considerably benefits the company by revising its time and minimizing its daily energy use. [ABSTRACT FROM AUTHOR]

Energy load forecasting is a systematic method of predicting future loads in advance to maintain a balance between energy demand and supply. Sustainable energy development is critical in energy design, planning, generation, transmission, and distribution, which leads to computational complexities in conventional methods when it is non-linear. Bio-inspired optimization algorithms are part of artificial intelligence that is inspired by the biological behavior and evolutionary principles of nature. Due to advancements in artificial intelligence, an increasing number of heuristic optimization techniques have resulted in the efficient handling of complex problems. The complexities in the accurate prediction of electricity can be achieved with evolutionary algorithms and swarm intelligence optimization algorithms. It avoids unnecessary expenditure, overestimation, and underestimation of electric energy and maximizes capacity utilization. This paper reviews recent studies on population-based optimization techniques to handle complexities in energy load forecasting and smart energy management. [ABSTRACT FROM AUTHOR]

Inventory models for stock-dependent demand have been developed continuously in recent years. In this paper, we develop inventory dependent model for multi-items by considering stock-dependent demand and stock-dependent holding rate for a rack that has a specific capacity. Two models are developed. The first model is an individual order quantity for each item and the second model is a joint replenishment model. The non-linear model with constraint is solved using an evolutionary algorithm. A numerical example and sensitivity analysis are conducted to show how the models work and to get some management insights. The sensitivity analysis shows that there is no dominant model. However, the joint replenishment order outperforms the individual order when the rack capacity is tight and there is a high ordering cost [ABSTRACT FROM AUTHOR]

The scheduling of power generating units is also recognized as Unit Commitment (UC); is an important and complex constrained mixed-integer nonlinear optimization problem. Traditionally UC was solved for thermal Units only, however, with the widespread adoption of Renewable energy sources like wind farms, UC is now being solved for hybrid generating units. This makes UC even more challenging due to the inherent uncertainty of the velocity of the wind. Evolutionary Algorithms (EAs) are used for solving UC as the traditional techniques employed for solving UC, like Dynamic Programming, suffer from the 'curse of dimensionality. However, EAs are unconstrained optimization techniques, and that is why they require special methods for handling constraints. Operator design and constraint handling in an EA is largely dependent on the problem representation. This paper proposes a novel adaptive quantum-inspired evolutionary algorithm with binary and real representation and a novel Repair heuristic for constraint handling in solving UC with thermal and wind farms. Initial testing has been performed on a well-known UC problem with ten thermal units and wind farm for twenty-four hours varying load model. The results achieved are competing as compared to other well-known techniques. [ABSTRACT FROM AUTHOR]

This paper demonstrates the application mechanism of a novel evolutionary algorithm to evaluate the financial performance of a power system while dealing with a combined economic and emission dispatch problem posed by various technical constraints. In particular, as a constrained objective function is associated with weighting factor scenarios, this problem considers the two dispatches for fuel and environmental aspects. Running simulations indicate that minimum costs are determined by weighting factors applied to the problem as a whole. Reducing total fuel costs with a dispatching priority and a pollutant target based on emission production has presented different implications as far as its contribution to the financial performance. Moreover, the increasing load demand results in generated power, costs, and emission discharges linked to the parameters and unit commitments. [ABSTRACT FROM AUTHOR]

Load balanced cluster formation in IoT based wireless sensor networks being reported in recent research literature as a well studied NP-hard problem has no trivial or obvious solution. Several protocols proposed in recent studies to resolve the load balancing problem mostly resort to communication intensive centralized algorithms or computation intensive memetic or evolutionary algorithms, that pose significant challenge on their applicability in pragmatic network deployment in real application scenario. This paper presents a simple yet effective distributed greedy algorithm for Energy efficient load balanced cluster assignment (EELBCA) for IoT based large scale sensor network. Our algorithm works in completely localized manner without the need of global topological information about the network. With simple and need based limited message passing the nodes self organize themselves to form well balanced clustered network solely based on their neighborhood information without the need of any computational intensive procedure. We evaluated our algorithm for various deployments viz. Random, Grid based and Gaussian networks and found our work seamlessly works on all such deployments. Simulation results shows our algorithm not only significantly minimizes on the standard deviation of loads among different cluster heads by 20-40% but also minimizes the energy requirement by 70-90% as compared to well documented existing algorithms for different network scenario and different node distribution in the network. [ABSTRACT FROM AUTHOR]

Fault location observability is an essential procedure among smart monitoring and outage management system issues used to realize self-healing networks, one of the most important characteristics of modern power systems. Further, the determination of the location of a faulty transmission line in a power grid is a vital topic to facilitate the self-healing network and to maintain the continuity of power supply. The self-healing procedure is to constantly keep the energy supplies to the customers in case of a power system configuration change such as a fault event happening on an electrical power transmission network. Self-healing is an operational aspect of electrical power networks for real-time identification and localization of faulty transmission lines for the whole power network. The implementation of PMUs as an accurate measurement device is a promising key entry for fault location and clearing utilizations. This paper proposes algorithmic models that determine the optimal number of PMUs and their placement sites to achieve a fault observable system whereas the state-estimation issue is examined. The optimization models discussed herein find a configuration of PMU locations with a least number of devices that locates any fault occurring in a transmission power network. An optimization process is presented within a nonlinear programming model coupling a synchronized phasor as well as conventional measurements. As a comparative optimization tool, a zero-one integer constraint linear program is implemented that satisfies global optimality with sufficient precision. Evolutionary algorithms such as a binary-particle swarm optimization and genetic algorithms have been adopted to characterize a fault observable power system. The algorithmic models are illustrated with an IEEE-14 bus system. Simulation results are tested on different size power systems to verify the efficacy of the proposed algorithmic approach without taking into account the fault classification, pre-fault loading conditions or considering the fault resistance. We present that this wide-area measuring implementation and its numerical results is an outstanding approach that improves the state-of-the-art in measurement driven methods regarding the finding of an optimal PMU set solution to pinpoint a fault event occurring anywhere in the power transmission grid. As proved via the numerical optimization, the mathematical and evolutionary algorithms are efficient and robust in finding some new benchmark optimal results for faulty conditions occurring in a modern power grid for any IEEE power system and case study. [ABSTRACT FROM AUTHOR]

In energy management, providing accurate findings to anticipate electrical load consumption is critical. The goal of this study is to develop methods for predicting electrical load, including artificial intelligence, neural network, ARIMA models, Bayesian models, and regression models. This paper proposes a new support vector regression application with the chaotic algorithms for electrical load forecasting in the southern region (Basra, Maysan, Dhi Qar, Muthanna), When choosing three parameters of a model (SVR) using evolutionary algorithms, you often encounter problems of early convergence, slowly reaching the solution of global optimization or falling into local optimization, and to overcome early local optimization in determining three parameters of the model (SVR) a chaotic algorithm is used. The 1980-2019 electrical load was used as a data set, and the results (CGASVR) were compared with (CPSOSVR) and (CIASVR) to choose the best form of electrical load forecasting where results show that the CGASVR model is more superior and efficient based on MSE, MAE, MAPE and MPE. [ABSTRACT FROM AUTHOR]

The main goal of the study is to build a fuzzy Sugeno model for diagnosing diseases in cattle based on the development of a knowledge base. To achieve this goal, an analytical review of mathematical methods for assessing the state of animals, soft computing methods for solving classification problems was carried out. The scientific novelty of the research is the improvement of the algorithm for constructing a Sugeno fuzzy model based on the development of a knowledge base for diagnosing diseases in cattle and solving classification problems based on the theory of fuzzy sets and interval methods, as well as developing improved algorithms and building models for diagnosing diseases in cattle based on Sugeno's fuzzy inference. When constructing a Sugeno fuzzy logical model for diagnosing diseases in cattle, the optimal number of rules and, accordingly, the values of their membership functions were found. Membership function parameters were obtained from experts in the veterinary field. In the future, the parameters of the membership function were tuned using neural networks and evolutionary algorithms to obtain the minimum number of fuzzy rules. The tasks of assessing, classifying and diagnosing a disease in cattle are solved based on training in seventeen signs. Types of diseases are diagnosed, such as ketosis, osteodystrophy, secondary osteodystrophy and hypomicroelementosis, which must be recognized. Based on the proposed approach, a program for diagnosing diseases in cattle has been developed. On the basis of the developed program, practical problems are solved that provide data training with the creation of a model and visualization of the results obtained and the reliability and accuracy in diagnosing at all stages of using the program. The ease of use of the software package lies in the fact that it does not require the user to have in-depth knowledge of its structure and the operating system of the computer. [ABSTRACT FROM AUTHOR]

Some possibilities of using convolutional artificial neural networks (ANNs) for powder diffraction structural analysis of crystalline substances are investigated. First, ANNs were used to classify crystal systems and space groups of symmetry based on the full-profile diffractograms calculated from the crystal structures of the ICSD 2017 database. The ICSD database contains 192004 structures, of which 80% were used for deep network learning, and 20% for independent testing of recognition accuracy. The accuracy of classification by a network of crystalline systems was 87.9%, and space groups was 77.2%. Second, ANN was applied for a similar classification of structural models generated by a stochastic evolutionary algorithm in the search for triclinic crystal structures of test compounds K4SnO4 and Rb4SnO4 from their full-profile diffraction patterns. The classification criterion was the hit of one or more atoms in their correct crystallographic positions in the structure of the substance. Independent deep learning of the network was performed on 100 thousand structural models of the triclinic structure of K4SnO4, generated in several runs of the evolutionary algorithm. The classification accuracy of the structural models of these substances was 72.2% and 76.9%, respectively. [ABSTRACT FROM AUTHOR]