Genetic Algorithm-based Optimization Research Articles

Multi-criterion optimization of a biologically-produced nanofluid flow inside tubes fitted with coaxial twisted tape based on first and second law viewpoints

The current study deals with a numerical simulation of heat transfer inside enhanced tubes having the coaxial double-twisted tape in the presence of an eco-friendly nanofluid containing graphene nanoplatelets (GNPs). The results unveil that when the particle concentration ( ϕ ) escalates, the thermal entropy generation rate ( S ̇ h ) tends to decline, whereas the heat transfer coefficient enhances. It was found that around 21.6% reduction in the S ̇ h is attained by a decrease of the twisted ratio ( ω ) from 3.5 to 2.5 at the concentration of 0.075% and Reynolds number of 10,000. Adding more GNPs to the water yields an increment in either the pressure drop or the frictional entropy generation. The findings refer to this issue that a multi-criterion optimization to find the best optimum state is required. A genetic algorithm-based optimization with the entropy generation rate, heat transfer, and pressure loss results in the optimum values at Re = 16,995.4, ϕ = 0.093%, and ω = 3.41. • Employing an eco-friendly nanofluid containing graphene nanosheets. • Second law analysis of a novel coaxial double twisted tape inside tubes. • Smaller twisted ratio causes further decrement in total entropy generation rate. • The higher Reynolds numbers weaken the positive effect of the nanofluid. • Multi-criterion optimization is done to determine the best optimum state.

High-performance permanent magnet array design by a fast genetic algorithm (GA)-based optimization for low-field portable MRI

Lightweight and compact permanent magnet arrays (PMAs) are suitable for portable dedicated magnetic resonance imaging (MRI). It is worth exploring different PMA design possibilities and optimization methods with an adequate balance between weight, size, and performance, in addition to Halbach arrays and C-shaped/H-shaped magnets which are widely used. In this paper, the design and optimization of a sparse high-performance inward-outward ring-pair PMA consisting of magnet cuboids is presented for portable imaging of the brain. The design is lightweight (151kg) and compact (inner bore diameter: 270mm, outer diameter: 616mm, length: 480mm, 5-Gauss range: 1840×1840×2340mm3). The optimization framework is based on the genetic algorithm with a consideration of both field properties and simulated image quality. The resulting PMA design has an average field strength of 101.5 mT and a field pattern with a built-in linear readout gradient. Subtracting the best fit to the linear gradient target resulted in a residual deviation from the target field of 0.76mT and an average linear regression coefficient of 0.85 to the linear gradient. The required radiofrequency bandwidth is 6.9% within a field of view (FoV) with a diameter of 200mm and a length of 125mm. It has a magnetic field generation efficiency of 0.67mT/kg, which is high among the sparse PMAs that were designed for an FoV with a diameter of 200mm. The field can be used to supply gradients in one direction working with gradient coils in the other two directions, or can be rotated to encode signals for imaging with axial slice selection. The encoding capability of the designed PMA was examined through the simulated reconstructed images. The force experienced by each magnet in the design was calculated, and the feasibility of a physical implementation was confirmed. The design can offer an increased field strength, and thus, an increased signal-to-noise ratio. It has a longitudinal field direction that allows the application of technologies developed for solenoidal magnets. This proposed design can be a promising alternative to supplying the main and gradient fields in combination for dedicated portable MRI. Lastly, the design is resulted from a fast genetic algorithm-based optimization in which fast magnetic field calculation was applied and high design flexibility was feasible. Within optimization iterations, image quality metrics were used for the encoding field of a magnet configuration to guide the design of the magnet array.

Design Methodology and Mission Assessment of Parallel Hybrid Electric Propulsion Systems

Abstract An integrated engine cycle design methodology and mission assessment for parallel hybrid electric propulsion architectures are presented in this paper. The aircraft case study considered is inspired by Fokker 100, boosted by an electric motor on the low-pressure shaft of the gas turbine. The fuel burn benefits arising from boosting the low-pressure shaft are discussed for two different baseline engine technologies. A three-point engine cycle design method is developed to redesign the engine cycle according to the degree of hybridization. The integrated cycle design and power management optimization method is employed to identify potential fuel burn benefits from hybridization for multiple mission ranges. The sensitivity of these mission results has also been analyzed for different assumptions on the electric powertrain. With 1 MW motor power and a battery pack of 2307 kg, a 3% fuel burn benefit can be obtained by retrofitting the gas turbine for 400 nm range. Optimizing the power management strategy improves this fuel burn benefit by 0.2-0.3%. Redesigning the gas turbine and optimizing the power management strategy, provides a 4.2% fuel benefit on 400 nm. The results suggest that a high hybridization by power, low hybridization by energy, and ranges below 700 nm are the only cases where the redesigned hybrid electric aircraft has benefits in fuel burn and energy consumption relative to the baseline aircraft. Finally, it is found that the percentage of fuel burn benefits from the hybrid electric configuration increases with the improvement in engine technology.

FEM-Based Power Transformer Model for Superconducting and Conventional Power Transformer Optimization

There were many promising superconducting materials discovered in the last decades that can significantly increase the efficiency of large power transformers. However, these large machines are generally custom-made and tailored to the given application. During the design process the most economical design should be selected from thousands of applicable solutions in a short design period. Due to the nonlinearity of the task, the cost-optimal transformer design, which has the smallest costs during the transformers’ planned lifetime, is usually not the design with the highest efficiency. Due to the topic’s importance, many simplified transformer models were published in the literature to resolve this problem. However, only a few papers considered this preliminary design optimization problem in the case of superconducting transformers and none of them made a comparison with a validated conventional transformer optimization model. This paper proposes a novel FEM-based two-winding transformer model, which can be used to calculate the main dimension of conventional and superconducting transformer designs. The models are stored in a unified JSON-file format, which can be easily integrated into an evolutionary or genetic algorithm-based optimization. The paper shows the used methods and their accuracy on conventional 10 MVA and superconducting 1.2 MVA transformer designs. Moreover, a simple cost optimization with the 10 MVA transformer was performed for two realistic economic scenarios. The results show that in some cases the cheaper, but less efficient, transformer can be the more economic.

Fault Tolerance Optimization of a Lithium Battery Pack Having a Damaged Unit.

As a kind of green and sustainable technology, electric vehicles are continuously highlighted for solving the significant problems of energy and air pollution. In this paper, fault tolerance optimization of an air-cooled lithium battery pack having a damaged unit was considered to improve the heat dissipation performance. For constructing the relationship between the objective function and the design variables, a quadratic polynomial response function was set up based on the experimental points, which were selected by the Latin-hypercube design of experiment method. Then, a multi-island genetic algorithm-based optimization of the battery pack having a damaged unit was performed under the condition of adjusting the flow velocity of each inlet. The results show that the acceptable temperature differences were obtained under the condition of fixing the total mass flow rate of the inlets. The simulation and fault tolerance optimization methods recommended in this study can be widely used to improve the real safety level of the systems.

Multi-objective optimization of turbocharger turbines for low carbon vehicles using meanline and neural network models

• Optimization framework for engine-turbocharger match and turbine design is described. • Framework couples engine cycle, turbine meanline, and neural network inertia models. • Aerodynamic and inertia changes reflect in steady and transient engine performance. • Turbine is optimized for the Worldwide harmonized Light vehicles Test Procedure. • We estimate resulting global CO2 saving of millions of tonnes over vehicle lifetimes. Due to slow turnover of the global vehicle parc internal combustion engines will remain a primary means of motive power for decades, so the automotive industry must continue to improve engine thermal efficiency to reduce CO 2 emissions, since savings will be compounded over the long lifetime of millions of vehicles. Turbochargers are a proven efficiency technology (most new vehicles are turbocharged) but are not optimally designed for real-world driving. The aim of this study was to develop a framework to optimize turbocharger turbine design for competing customer objectives: minimizing fuel consumption (and thus CO 2 emissions) over a representative drive cycle, while minimizing transient response time. This is achieved by coupling engine cycle, turbine meanline, and neural network inertia models within a genetic algorithm-based optimizer, allowing aerodynamic and inertia changes to be accurately reflected in drive cycle fuel consumption and transient performance. Exercising the framework for the average new passenger car across a drive cycle representing the Worldwide harmonized Light vehicles Test Procedure reveals the trade-off between competing objectives and a turbine design that maintains transient response while minimizing fuel consumption due to a 3 percentage-point improvement in turbine peak efficiency, validated by experiment. This optimization framework is fast to execute, requiring only eight turbine geometric parameters, making it a commercially viable procedure that can refine existing or optimize tailor-made turbines for any turbocharged application (whether gasoline, diesel, or alternatively fuelled), but if applied to turbocharged gasoline cars in the EU would lead to lifetime CO 2 savings of > 290,000 tonnes per production year, and millions of tonnes if deployed worldwide.

Derivation of unit hydrograph using genetic algorithm-based optimization model

Derivation of unit hydrograph using genetic algorithm-based optimization model

Genetic Algorithm-Based Optimization for Color Point Cloud Registration.

Point cloud registration is an important technique for 3D environment map construction. Traditional point cloud registration algorithms rely on color features or geometric features, which leave problems such as color affected by environmental lighting. This article introduced a color point cloud registration algorithm optimized by a genetic algorithm, which has good robustness for different lighting environments. We extracted the HSV color data from the point cloud color information and made the HSV distribution of the tangent plane continuous, and we used the genetic algorithm to optimize the point cloud color information consistently. The Gauss–Newton method was utilized to realize the optimal registration of color point clouds for the joint error function of color and geometry. The contribution of this study was that the genetic algorithm was used to optimize HSV color information of the point cloud and was applied to the point cloud registration algorithm, which reduces the influence of illumination on color information and improves the accuracy of registration. The experimental results showed that the square error of color information saturation and lightness optimized by the genetic algorithm was reduced by 14.07% and 37.16%, respectively. The color point cloud registration algorithm in this article was reduced by 12.53% on average compared with the optimal result algorithm RMSE.

TextHoaxer: Budgeted Hard-Label Adversarial Attacks on Text

This paper focuses on a newly challenging setting in hard-label adversarial attacks on text data by taking the budget information into account. Although existing approaches can successfully generate adversarial examples in the hard-label setting, they follow an ideal assumption that the victim model does not restrict the number of queries. However, in real-world applications the query budget is usually tight or limited. Moreover, existing hard-label adversarial attack techniques use the genetic algorithm to optimize discrete text data by maintaining a number of adversarial candidates during optimization, which can lead to the problem of generating low-quality adversarial examples in the tight-budget setting. To solve this problem, in this paper, we propose a new method named TextHoaxer by formulating the budgeted hard-label adversarial attack task on text data as a gradient-based optimization problem of perturbation matrix in the continuous word embedding space. Compared with the genetic algorithm-based optimization, our solution only uses a single initialized adversarial example as the adversarial candidate for optimization, which significantly reduces the number of queries. The optimization is guided by a new objective function consisting of three terms, i.e., semantic similarity term, pair-wise perturbation constraint, and sparsity constraint. Semantic similarity term and pair-wise perturbation constraint can ensure the high semantic similarity of adversarial examples from both comprehensive text-level and individual word-level, while the sparsity constraint explicitly restricts the number of perturbed words, which is also helpful for enhancing the quality of generated text. We conduct extensive experiments on eight text datasets against three representative natural language models, and experimental results show that TextHoaxer can generate high-quality adversarial examples with higher semantic similarity and lower perturbation rate under the tight-budget setting.

Efficient Heat Exchange Configuration for Sub-Cooling Cycle of Hydrogen Liquefaction Process

The hydrogen liquefaction process is highly energy-intensive owing to its cryogenic characteristics, and a large proportion of the total energy is consumed in the subcooling cycle. This study aimed to develop an efficient configuration for the subcooling cycle in the hydrogen liquefaction process. The He-Ne Brayton cycle is one of the most energy-efficient cycles of the various proposed hydrogen liquefaction processes, and it was selected as the base case configuration. To improve its efficiency and economic potential, two different process configurations were proposed: (configuration 1) a dual-pressure cycle that simplified the process configuration, and (configuration 2) a split triple-pressure cycle that decreased the flow rate of the medium- and high-pressure compressors. The ortho–para conversion heat of hydrogen is considered by using heat capacity data of equilibrium hydrogen. Genetic algorithm-based optimization was also conducted to minimize the energy consumption of each configuration, and the optimization results showed that the performance of configuration 1 was worse than that of the base case configuration. In this respect, although less equipment was used, the compression load on each compressor was very intensive, which increased the energy requirements and costs. Configuration 2 provided the best results with a specific energy consumption of 5.69 kWh/kg (3.2% lower than the base case configuration). The total expense of configuration 2 shows the lowest value which is USD 720 million. The process performance improvements were analyzed based on the association between the refrigerant composition and the heat exchange efficiency. The analysis demonstrated that energy efficiency and costs were both improved by dividing the pressure levels and splitting the refrigerant flow rate in configuration 2.

Multi-Objective Optimization Design of Ladle Refractory Lining Based on Genetic Algorithm.

Genetic algorithm is widely used in multi-objective mechanical structure optimization. In this paper, a genetic algorithm-based optimization method for ladle refractory lining structure is proposed. First, the parametric finite element model of the new ladle refractory lining is established by using ANSYS Workbench software. The refractory lining is mainly composed of insulating layer, permanent layer and working layer. Secondly, a mathematical model for multi-objective optimization is established to reveal the functional relationship between the maximum equivalent force on the ladle lining, the maximum temperature on the ladle shell, the total mass of the ladle and the structural parameters of the ladle refractory lining. Genetic algorithm translates the optimization process of ladle refractory lining into natural evolution and selection. The optimization results show that, compared with the unoptimized ladle refractory lining structure (insulation layer thickness of 0 mm, permanent layer thickness of 81 mm, and working layer thickness of 152 mm), the refractory lining with insulation layer thickness of 8.02 mm, permanent layer thickness of 76.20 mm, and working layer thickness of 148.61 mm has the best thermal insulation performance and longer service life within the variation of ladle refractory lining structure parameters. Finally, the results of the optimization are verified and analyzed in this paper. The study found that by optimizing the design of the ladle refractory lining, the maximum equivalent force on the ladle lining, the maximum temperature on the ladle shell and the ladle mass were reduced. The thermal insulation performance and the lightweight performance of the ladle are improved, which is very important for improving the service life of the ladle.

Multiagent Reinforcement Learning for Community Energy Management to Mitigate Peak Rebounds Under Renewable Energy Uncertainty

Price-based demand response (DR) can aid power grid management, but an uncoordinated response may lead to peak rebounds during low-price periods. This article proposes a community energy management system based on multiagent reinforcement learning. The scheme consists of a community aggregator that optimizes the total community electricity cost for multiple residential users. A home requires energy management for home appliances, electric vehicles, energy storage systems, and renewable energy generation. The appliance scheduling problem is decomposed into smaller sequential decision problems that are easier to solve. Renewable generation is predicted and used to mitigate the influence of energy generation uncertainty. As indicated in numerical analyses, the proposed approach can handle the uncertainty in renewable energy and leads to more economical energy usage relative to existing energy management methods. The method outperforms conventional algorithms, such as centralized mixed-integer nonlinear programming and genetic algorithm-based optimization, in terms of mitigating peak rebounds and addressing the uncertainty of renewable energy generation.

Hygrothermal calibration and validation of vernacular dwellings: A genetic algorithm-based optimisation methodology

Heritage model calibration and validation are crucial for decreasing uncertainty and enhancing the robustness of simulation results and conservation interventions. Yet, hygrothermal modelling methodologies are marked by significant heterogeneity and lack of robustness. Aiming to provide a solution for the drawbacks identified, this study puts forth a comprehensive hygrothermal modelling methodology. Following the in situ data collection of a subset of heritage buildings, 22 vernacular dwellings in Southern Portugal, a three-step method was developed, consisting of: Morris sensitivity analysis, optimisation-based calibration, and validation and multi-criteria decision-making (MCDM). A genetic algorithm multi-objective optimisation-based calibration with NSGA-II was implemented for simultaneously minimising the statistical indicators RMSE and MAE for the indoor air temperature of the winter and summer models. The validation and MCDM were conducted by means of threshold compliance and Compromise Programming. NSGA-II found Pareto frontiers composed of nine and six optimal solutions for the summer and winter models, respectively, in nearly 3 h each. All optimal solutions significantly decreased the RMSE and MAE, especially in the summer model, regarding the baseline data. The final solutions selected after the MCDM resulted in an accuracy improvement of 51% and 54% for the RMSE and MAE for the winter model and 80% and 81% for the RMSE and MAE in the summer model, compared to the baseline models. The strong correlation found between the calibrated models and the measured data along with the enhancement of calibrated data regarding the baseline model, highlighted the potential of using GAs to obtain calibrated vernacular models that robustly predict real building performance and foster better retrofitting decision-making.

Enhanced Cellular Materials through Multiscale, Variable-Section Inner Designs: Mechanical Attributes and Neural Network Modeling.

In the current work, the mechanical response of multiscale cellular materials with hollow variable-section inner elements is analyzed, combining experimental, numerical and machine learning techniques. At first, the effect of multiscale designs on the macroscale material attributes is quantified as a function of their inner structure. To that scope, analytical, closed-form expressions for the axial and bending inner element-scale stiffness are elaborated. The multiscale metamaterial performance is numerically probed for variable-section, multiscale honeycomb, square and re-entrant star-shaped lattice architectures. It is observed that a substantial normal, bulk and shear specific stiffness increase can be achieved, which differs depending on the upper-scale lattice pattern. Subsequently, extended mechanical datasets are created for the training of machine learning models of the metamaterial performance. Thereupon, neural network (NN) architectures and modeling parameters that can robustly capture the multiscale material response are identified. It is demonstrated that rather low-numerical-cost NN models can assess the complete set of elastic properties with substantial accuracy, providing a direct link between the underlying design parameters and the macroscale metamaterial performance. Moreover, inverse, multi-objective engineering tasks become feasible. It is shown that unified machine-learning-based representation allows for the inverse identification of the inner multiscale structural topology and base material parameters that optimally meet multiple macroscale performance objectives, coupling the NN metamaterial models with genetic algorithm-based optimization schemes.

A new federated genetic algorithm-based optimization technique for multi-criteria vehicle route planning using ArcGIS network analyst

PurposeUsing a real-time road network combined with historical traffic data for Al-Salt city, the paper aims to propose a new federated genetic algorithm (GA)-based optimization technique to solve the dynamic vehicle routing problem. Using a GA solver, the estimated routing time for 300 chromosomes (routes) was the shortest and most efficient over 30 generations.Design/methodology/approachIn transportation systems, the objective of route planning techniques has been revised from focusing on road directors to road users. As a result, the new transportation systems use advanced technologies to support drivers and provide them with the road information they need and the services they require to reduce traffic congestion and improve routing problems. In recent decades, numerous studies have been conducted on how to find an efficient and suitable route for vehicles, known as the vehicle routing problem (VRP). To identify the best route, VRP uses real-time information-acquired geographical information systems (GIS) tools.FindingsThis study aims to develop a route planning tool using ArcGIS network analyst to enhance both cost and service quality measures, taking into account several factors to determine the best route based on the users’ preferences.Originality/valueFurthermore, developing a route planning tool using ArcGIS network analyst to enhance both cost and service quality measures, taking into account several factors to determine the best route based on the users’ preferences. An adaptive genetic algorithm (GA) is used to determine the optimal time route, taking into account factors that affect vehicle arrival times and cause delays. In addition, ArcGIS' Network Analyst tool is used to determine the best route based on the user's preferences using a real-time map.

A novel consumer-friendly electric vehicle charging scheme with vehicle to grid provision supported by genetic algorithm based optimization

Due to escalating demand for electric vehicles (EVs) in the worldwide transportation sector, the charging facilities supported by an efficient charging scheme across the power network have a significant impact on the operation of the power system. The unsupervised and decentralized charging of the EVs have several adverse effects on various operational features of the power system. On the other hand, the centralized charging system with proper charging schemes offers less complexity in power system operation to avoid unnecessary power system network stress. This article proposes a valley-filling technique-based EV charging scheme for residential consumers facilitated by a centralized charging system. The charging scheme comprises a genetic algorithm-based optimization for EV acceptance to utilize the available energy of the supply network in the best way. The charging scheme also includes vehicle to grid (V2G) facilitation and the reallocation technique for shifting EVs to resolve any overloading of charging slots. Simulation results of Genetic Algorithm (GA) based optimization show the attainment of near-optimum EV data set with less than 50 iteration numbers and usage of more than 99.5% of available kWh for charging. Other result segments indicate the efficient use of available supply energy by incorporating EVs in the valley time slab of the demand curve. Also, the findings point out that the charging scheme successfully encounters unwanted network stress issues during EV integration by V2G provision and reallocation technique, prioritizing consumer satisfaction at the overloaded charging time slots. The test case simulation results indicate that the ‘average to peak’ demand ratio in the demand curve is increased to 90% from 68% by incorporating the proposed charging scheme for EV charging. Moreover, the outcomes indicate the novelty of the proposed methodology in allocating charging slots considering complex driving parameters like network stress and customer prioritization.

Analysis and Performance Optimization of Supercritical CO2 Recompression Brayton Cycle Coupled Organic Rankine Cycle Based on Solar Tower

Abstract Supercritical carbon dioxide (SCO2) Brayton cycle has been proved to be an efficient power cycle to replace the traditional steam Rankine cycle. The thermal efficiency of SCO2 cycle can be further improved by coupling another type of cycle (the bottom cycle) at the waste heat end. A supercritical carbon dioxide recompression Brayton cycle (SCRBC) coupled organic Rankine cycle (ORC) based on solar tower is designed and established. According to the requirements of the waste heat temperature range of the top cycle, R600 is selected as the working medium of ORC. Under the design conditions, the effects of split ratio on the net power, the thermal efficiency, and the exergy loss of the combined cycle are studied. The variation of thermal efficiency of each part of the system with split ratio under different turbine inlet pressures and temperatures is further analyzed, and the influence of turbine inlet pressure and working fluid mass flow ratio ε (mass flow ratio of CO2 to R600) on the system performance is analyzed. Genetic algorithm-based multiobjective optimization is used to obtain the Pareto solution set for the thermal performance and unit investment cost of the system. The results show that the thermal efficiency of the combined cycle can be increased by more than 2% compared with that of a single top cycle. There is an optimal split ratio to maximize the thermal efficiency of the combined cycle, and the positions of the optimal split ratio are different for different turbine inlet pressures. Finally, through the multiobjective optimization method, several groups of Pareto solutions can be found, which can provide some reference for engineering design.

Development of an optimized thermodynamic model for VVER-1200 reactor-based nuclear power plants using genetic algorithm

In this work, an optimized thermodynamic modelforthe secondary coolant loop of a VVER-1200 reactor-based nuclear power plant isdeveloped. The values of some of the thermo-physical parameters such as coolant flow rate, coolant inlet and outlet temperatures from the steam generator and the turbines, etc. are obtained from the available literature. On the other hand, the unknown plant parameters such as feedwater heater pressures are predicted using a Genetic Algorithm-based optimization process. Three crossover rates, 0.5, 0.7 and 0.9, and three mutation rates, 0.01, 0.05 and 0.1, are considered. Results reveal that the value of the optimization process of the thermodynamic model is not significantly affected by the change in the crossover rate or mutation rate. The optimized model is compared with two theoretical models, a model available in the literature and experimental data. Results reveal that the predicted values of the optimized model are more accurate compared to the other available options. The optimized model also predicts that the maximum overall efficiency and the maximum output power of a plant situated in tropical region should be around 35.87% and 1152.16MWe respectively for 28 °C tertiary coolant temperature, down by 3.94% from the rated values in cold weather. However, the power output at 35 °C tertiary coolant temperature is calculated to be 1117.88 MWe, almost 6.69% lower than the rated value for VVER-1200. Thus, the efficiency of the plant is highly sensitive to the weather variation of a region.

Learnable Wavelet Scattering Networks: Applications to Fault Diagnosis of Analog Circuits and Rotating Machinery

Analog circuits are a critical part of industrial electronics and systems. Estimates in the literature show that, even though analog circuits comprise less than 20% of all circuits, they are responsible for more than 80% of faults. Hence, analog circuit fault diagnosis and isolation can be a valuable means of ensuring the reliability of circuits. This paper introduces a novel technique of learning time–frequency representations, using learnable wavelet scattering networks, for the fault diagnosis of circuits and rotating machinery. Wavelet scattering networks, which are fixed time–frequency representations based on existing wavelets, are modified to be learnable so that they can learn features that are optimal for fault diagnosis. The learnable wavelet scattering networks are developed using the genetic algorithm-based optimization of second-generation wavelet transform operators. The simulation and experimental results for the diagnosis of analog circuit faults demonstrates that the developed diagnosis scheme achieves greater fault diagnosis accuracy than other methods in the literature, even while considering a larger number of fault classes. The performance of the diagnosis scheme on benchmark datasets of bearing faults and gear faults shows that the developed method generalizes well to fault diagnosis in multiple domains and has good transfer learning performance, too.

True Gaussian mixture regression and genetic algorithm-based optimization with constraints for direct inverse analysis

True Gaussian mixture regression and genetic algorithm-based optimization with constraints for direct inverse analysis