Augmented Lagrangian Genetic Algorithm Research Articles

A Si IGBT/SiC MOSFET Hybrid Isolated Bidirectional DC–DC Converter for Reducing Losses and Costs of DC Solid State Transformer

The DC solid state transformer (DCSST) is a crucial component for connecting buses of different voltage levels in the DC distribution grid. This paper proposes a Si IGBT/SiC MOSFET hybrid isolated bidirectional DC–DC converter and an optimized modulation strategy (OMS) to reduce the losses and costs of DCSST. Based on the analysis of topology and operating principles, a duty-cycle modulation strategy is proposed and the converter is modeled by the time domain analysis (TDA) method. Through the analysis of switching characteristics, an optimization problem is established, which aims to reduce the conduction losses of switches while ensuring zero-voltage switching (ZVS) for all switches and low-current turn-off for IGBTs simultaneously. The optimization problem is solved by the augmented Lagrangian genetic algorithm (ALGA), and the OMS for the proposed converter is deduced. Finally, a 2 kW experimental prototype with the primary voltage of 405–495 V and the secondary voltage of 150 V is built to verify the effectiveness of the proposed topology and OMS. The switching costs of the proposed converter is reduced by 27.3% and the efficiency is improved by up to 4.04% compared to the existing method.

A truncated test scheme design method for success-failure in-orbit tests

Based on the success-failure test feature of in-orbit tests (IOTs) for typical space equipment, this paper presents a method for designing a truncated test scheme for success-failure in-orbit tests. With this method, a small upper boundary of the sample size for the IOT verification test can be obtained before the test starts. The method introduces the truncated Bayes-sequential mesh test (SMT) method into the design of the IOT verification test scheme and greatly compresses the continuous test area by incorporating optimization theory, resulting in a smaller upper limit of the IOT sample size. First, this paper derives a specific calculation formula for the Bayes-SMT critical line. Second, the Markov chain model is adopted to calculate the occurrence probabilities of each acceptance and rejection point through state transition. Finally, an optimal truncated test optimization algorithm based on the augmented lagrangian genetic algorithm is proposed. Simulation tests show that, compared with the classical single sampling method, the truncated sequential probability ratio test method, the truncated SMT method, and the truncated Bayes-SMT method based on step-by-step calculation, the method presented in this paper can be used to obtain a sequential test scheme with smaller truncated sample size.

Adaptive Protection for Active Distribution Networks: An Approach Based on Fuses and Relays With Multiple Setting Groups

Protection schemes are essential in active distribution networks and microgrids’ reliable, efficient, and flexible operation. However, the protection of these networks presents significant challenges due to operational changes, such as variations in topology, distributed energy resources connection/disconnection, and microgrid operating modes, among others. This paper proposes an adaptive protection scheme based on overcurrent devices with several setting groups based on artificial intelligence algorithms. The developed strategy is composed of two stages. In the off-line stage, a clustering technique is employed to group the active distribution network operating scenarios exhibiting similarities. The optimal settings for the protection devices are determined for each set of scenarios. On the other hand, in the on-line stage, the protection strategy’s implementation and operation, considering the active distribution network’s existing communication system, are defined. Furthermore, the approach formulates the overcurrent relay coordination as a mixed-integer non-linear optimization problem, and as a result, the optimal setting of the overcurrent protection devices is obtained. It aims to minimize the operating time, considering the transformers’ thermal limits, fuse operating curves, and overcurrent relay settings. The solution is determined by using an Augmented Lagrangian genetic algorithm. The presented protection scheme is validated on the modified IEEE 34 node test feeder, considering the main operating scenarios of the active distribution networks, such as topology changes, distributed energy resource connection/disconnection, and microgrid operating modes (on-grid and off-grid). The results obtained and its easy implementation indicates the high potential for real-life applications.

Optimization of cutting parameters for machining time in turning process

This paper describes the most effective methods for nonlinear constraint optimization of cutting parameters in the turning process. Among them are Linearization Programming Method with Dual-Simplex algorithm, Interior Point method, and Augmented Lagrangian Genetic Algorithm (ALGA). Every each of them is tested on an actual example – the minimization of production rate in turning process. The computation was conducted in the MATLAB environment. The comparative results obtained from the application of these methods show: The optimal value of the linearized objective and the original function are the same. ALGA gives sufficiently accurate values, however, when the algorithm uses the Hybrid function with Interior Point algorithm, the resulted values have the maximal accuracy.

Optimization of Real Power Loss and Voltage Stability Index for Distribution Systems with Distributed Generation

Integrating Distributed Generation (DG) units into distribution systems can have an impact on the voltage profile, power flow, power losses, and voltage stability. In this paper, a new methodology for DG location and sizing are developed to minimize system losses and maximize voltage stability index (VSI). A proper allocation of DG has to be determined using the fuzzy ranking method to verify best compromised solutions and achieve maximum benefits. Synchronous machines are utilized and its power factor is optimally determined via genetic optimization to inject reactive power to decrease system losses and improve voltage profile and VSI. The Augmented Lagrangian Genetic Algorithm with nonlinear mixed-integer variables and Non-dominated Sorting Genetic Algorithm have been implemented to solve both single/multi-objective function optimization problems. For proposed methodology effectiveness verification, it is tested on 33-bus and 69-bus radial distribution systems then compared with previous works.

Exergy optimization of cooling tower for HGSHP and HVAC applications

In the present work, a constrained inverse optimization method for building cooling applications is proposed to control the mechanical draft wet cooling tower by minimizing the exergy destruction and satisfying an imposed heat load under varying environmental conditions. The optimization problem is formulated considering the cooling dominated heating, ventilation and air conditioning (HVAC) and hybrid ground source heat pump (HGSHP). As per the requirement, new second degree correlations for the tower outlet parameters (water temperature, air dry and wet-bulb temperatures) with five inlet parameters (dry-bulb temperature, relative humidity, water inlet temperature, water and air mass flow rates) are developed. The Box–Behnken design response surface method is implemented for developing the correlations. Subsequently, the constrained optimization problem is solved using augmented Lagrangian genetic algorithm. This work further developed optimum inlet parameters operating curves for the HGSHP and the HVAC systems under varying environmental conditions aimed at minimizing the exergy destruction along with the fulfillment of the required heat load.

Wireless Distributive Personal Communication for Early Detection of Collateral Cancer Using Optimized Machine Learning Methodology

Computer-aided diagnosis plays a vital role in many medical analysis especially with CT Images. The diagnosis of colorectal cancer detection becomes more challenging since the perfect detection of polyps (An opacified fluid lie on the colon walls) is difficult even for advanced segmentation methods. Many methods have been already developed, which lacks in segmentation of structure. In-order to achieve an accurate segmentation, a machine learning based algorithm such as regression neural network enhanced with the augmented Lagrangian genetic algorithm (RNN-ALGA). Since the segmentation was processed for multispectral image, the optimization strategy was included in-order to process all the 2D slices with reduced time span. The segmentation was done in-order to process the differentiation of colons, partial volume effect and bowels. By using this algorithm (RNN-ALGA) it is possible to achieve 97 % accuracy with some minor error. The comparison results shows that this algorithms best suitable abdominal slices of CT image is due to its enhanced nonlinearity nature. This methodology was developed especially for the hospital applications. RNN-ALGA works virtually in the common cloud server with an effective wireless distributive system accessed by many clients placed at the remote regions (medical centers, hospitals etc.).

Unconstrained shape optimisation of singly-symmetric and open cold-formed steel beams and beam-columns

This study aims to optimise the cross-sectional shape of singly-symmetric, open-section and simply-supported cold-formed steel (CFS) beams and beam-columns. No manufacturing or assembly constraints are considered. The previously developed augmented Lagrangian Genetic Algorithm (GA), referred to as the “self-shape” optimisation algorithm, is used herein. Fully restrained and unrestrained beams against lateral deflection and twist, as well as unrestrained beam-columns are optimised. Various combinations of axial compressive load and bending moment are analysed for the beam-columns. The Direct Strength Method (DSM) is used to evaluate the nominal member compressive and bending capacities. The accuracy of the automated rules, developed in the literature to determine the elastic local and distortional axial buckling stresses from Finite Strip signature curves, is verified herein to estimate the elastic bending buckling stresses. The optimised cross-sectional shapes are presented for all cases and the evolution of the unrestrained shapes from pure axial compression to pure bending is discussed.

A new hybrid method for size and topology optimization of truss structures using modified ALGA and QPGA

Modified Augmented Lagrangian Genetic Algorithm (ALGA) and Quadratic Penalty Function Genetic Algo­rithm (QPGA) optimization methods are proposed to obtain truss structures with minimum structural weight using both continuous and discrete design variables. To achieve robust solutions, Compressed Sparse Row (CSR) with reordering of Cholesky factorization and Moore Penrose Pseudoinverse are used in case of non-singular and singular stiffness matrix, respectively. The efficiency of the proposed nonlinear optimization methods is demonstrated on several practical exam­ples. The results obtained from the Pratt truss bridge show that the optimum design solution using discrete parameters is 21% lighter than the traditional design with uniform cross sections. Similarly, the results obtained from the 57-bar planar tower truss indicate that the proposed design method using continuous and discrete design parameters can be up to 29% and 9% lighter than traditional design solutions, respectively. Through sensitivity analysis, it is shown that the proposed methodology is robust and leads to significant improvements in convergence rates, which should prove useful in large-scale applications.

Augmented Lagrangian Genetic Algorithm Based Decentralized Control Configuration Design for Fluid Catalytic Cracking Units

In this work, three decentralized control configuration designs—independent, sequential and simultaneous designs—were used in multivariable feedback configurations for PI control of the riser and regenerator temperatures of FCCU in order to compare their performances. Control design was formulated as optimization problem to minimize infinity norm of weighted sensitivity functions subject to μ-interaction measure bound on diagonal complementary functions of the closed loop system. The optimization problem was solved using augmented Lagrangian genetic algorithm. Simulation results show that simultaneous and independent designs give good response with less overshoot and with no oscillation. Bound on μ-interaction measure is satisfied for both designs meaning that their nominal stabilities are guaranteed; however, it is marginal for simultaneous design. Simultaneous design outperforms independent design in term of robust performance while independent design gives the best performance in terms of robust stability. Sequential design gives the worst performance out of the three designs.

Genetic single objective optimisation for sizing and allocation of renewable DG systems

The optimal design of renewable-based distributed generations (DGs) is a challenging issue in order to maximise their benefits and to overcome power quality problems. Therefore, this paper proposes a methodology for optimal allocation and sizing of renewable DG units to minimise total power losses over radial distribution systems. The planning problem is formulated as a single objective nonlinear mixed integer-constrained optimisation problem and is solved by using the augmented Lagrangian genetic algorithm (ALGA) by combining the objective function and the nonlinear constraints. In that case, the ALGA solves a sequence of sub-problems where the objective function penalises the constraints violation in order to obtain the best solution. The proposed technique is applied to IEEE radial test systems including 33-bus, 69-bus and 119-bus and is implemented with different scenarios including all possible combinations and various types of renewable DG units to prove the effectiveness of the proposed methodology.

Single-stage Kanban system with deterioration failures and condition-based preventive maintenance

Despite the fact that the fields of pull type production control policies and condition-based preventive maintenance have much in common contextually, they have evolved independently up to now. In this investigation, an attempt is made to bridge the gap between these two branches of knowledge by introducing the single-stage Kanban system with deterioration failures and condition-based preventive maintenance. The formalism of continuous time Markov chains is used to model the system and expressions for eight performance metrics are derived. Two important, from a managerial perspective, constrained optimization problems for the proposed model are defined where the objective is the simultaneous optimization of the Kanban policy, the preventive maintenance policy and the inspection schedule under conflicting performance criteria. Multiple instances of each optimization problem are solved by means of the augmented Lagrangian genetic algorithm. The results from the optimization trials coupled by the results from extensive numerical examples facilitate the thorough investigation of the system’s behaviour.

Interactive decision procedure for watershed nutrient load reduction: An integrated chance-constrained programming model with risk–cost tradeoff

Nutrient load reduction is a well-recognized requirement for aquatic ecosystem restoration. However, decision making is difficult due to challenges related to uncertainty and the interaction between decision makers and modelers, including (a) the quantitative relationship between risks arising from different aspects and the fact that cost is not usually revealed and (b) the fact that decision makers are not significantly involved in the modeling process. In this study, an interactive optimal-decision procedure with risk–cost tradeoff is proposed to overcome these limitations. It consists of chance-constrained programming (CCP) models, risk scenario analysis using the Taguchi method, risk–cost tradeoff and feedback for model adaption. A hybrid intelligent algorithm (HIA) integrating Monte Carlo simulation, artificial neural networks, and an augmented Lagrangian genetic algorithm was developed and applied to solve the CCP model. The proposed decision procedure and HIA are illustrated through a case study of uncertainty-based optimal nutrient load reduction in the Lake Qionghai Watershed, China. The CCP model has four constraints associated with risk levels indicating the possibility of constraint violation. Sixteen risk scenarios were designed with the Taguchi method to recognize the interactions between multiple constraint risks and total cost. The results were analyzed using the signal-to-noise ratio, analysis of variance, and multivariate regression. The model results demonstrate how cost is affected by risk for the four constraints and show that the proposed approach can provide effective support for decision making on risk–cost tradeoffs.

An optimal modeling of multidimensional wave digital filtering network for free vibration analysis of symmetrically laminated composite FSDT plates

The technique of multidimensional wave digital filtering (MDWDF) that builds on traveling wave formulation of lumped electrical elements, is successfully implemented on the study of dynamic responses of symmetrically laminated composite plate based on the first order shear deformation theory. The philosophy applied for the first time in this laminate mechanics relies on integration of certain principles involving modeling and simulation, circuit theory, and MD digital signal processing to provide a great variety of outstanding features. Especially benefited by the conservation of passivity gives rise to a nonlinear programming problem (NLP) for the issue of numerical stability of a MD discrete system. Adopting the augmented Lagrangian genetic algorithm, an effective optimization technique for rapidly achieving solution spaces of NLP models, numerical stability of the MDWDF network is well received at all time by the satisfaction of the Courant–Friedrichs–Levy stability criterion with the least restriction. In particular, optimum of the NLP has led to the optimality of the network in terms of effectively and accurately predicting the desired fundamental frequency, and thus to give an insight into the robustness of the network by looking at the distribution of system energies. To further explore the application of the optimum network, more numerical examples are engaged in efforts to achieve a qualitative understanding of the behavior of the laminar system. These are carried out by investigating various effects based on different stacking sequences, stiffness and span-to-thickness ratios, mode shapes and boundary conditions. Results are scrupulously validated by cross referencing with early published works, which show that the present method is in excellent agreement with other numerical and analytical methods.

Use of Genetic Algorithms in Resource Scheduling of Construction Projects

This paper presents an augmented Lagrangian genetic algorithm model for resource scheduling. The algorithm considers scheduling characteristics that were ignored in prior research. Previous resource scheduling formulations have primarily focused on project duration minimization. Furthermore, resource leveling and resource-constrained scheduling have traditionally been solved independently. The model presented here considers all precedence relationships, multiple crew strategies, total project cost minimization, and time-cost trade-off. In the new formulation, resource leveling and resource-constrained scheduling are performed simultaneously. The model presented uses the quadratic penalty function to transform the resource-scheduling problem to an unconstrained one. The algorithm is general and can be applied to a broad class of optimization problems. An illustrative example is presented to demonstrate the performance of the proposed method.

Fuzzy Genetic Algorithm for Optimization of Steel Structures

In the traditional optimization algorithms, constraints are satisfied within a tolerance defined by a crisp number. In actual engineering practice, constraint evaluation involves many sources of imprecision and approximation. When an optimization algorithm is forced to satisfy the design constraints exactly, it can miss the global optimum solution within the confine of commonly acceptable approximations. Extending the augmented Lagrangian genetic algorithm (GA) of Adeli and Cheng, a fuzzy augmented Lagrangian GA is presented for optimization of steel structures subjected to the constraints of the AISC allowable stress design specifications taking into account the fuzziness in the constraints. The membership function for the fuzzy domain is found by the intersection of the fuzzy membership function for the objective function and the constraints using the max-min procedure of Bellman and Zadeh. Nonlinear quadratic fuzzy membership functions are used for objective function and constraints. The algorithm is applied to two space axial-load structures including a large 37-story structure with 1,310 members. The features and advantages of the new fuzzy GA include acknowledging the imprecision and fuzziness in the code-based design constraints, increased likelihood of obtaining the global optimum solution, improved convergence, and reduced total computer processing time.

Concurrent Genetic Algorithms for Optimization of Large Structures

In a recent article, the writers presented an augmented Lagrangian genetic algorithm for optimization of structures. The optimization of large structures such as high‐rise building structures and space stations with several hundred members by the hybrid genetic algorithm requires the creation of thousands of strings in the population and the corresponding large number of structural analyses. In this paper, the writers extend their previous work by presenting two concurrent augmented Lagrangian genetic algorithms for optimization of large structures utilizing the multiprocessing capabilities of high‐performance computers such as the Cray Y‐MP 8/864 supercomputer. Efficiency of the algorithms has been investigated by applying them to four space structures including two high‐rise building structures. It is observed that the performance of both algorithms improves with the size of the structure, making them particularly suitable for optimization of large structures. A maximum parallel processing speed of 7.7 is achieved for a 35‐story tower (with 1,262 elements and 936 degrees of freedom), using eight processors.

Augmented Lagrangian Genetic Algorithm for Structural Optimization

This paper presents a robust hybrid genetic algorithm for optimization of space structures using the augmented Lagrangian method. An attractive characteristic of genetic algorithm is that there is no line search and the problem of computation of derivatives of the objective function and constraints is avoided. This feature of genetic algorithms is maintained in the hybrid genetic algorithm presented in this paper. Compared with the penalty function‐based genetic algorithm, only a few additional simple function evaluations are needed in the new algorithm. Furthermore, the trial and error approach for the starting penalty function coefficient and the process of arbitrary adjustments are avoided. There is no need to perform extensive numerical experiments to find a suitable value for the penalty function coefficient for each type or class of optimization problem. The algorithm is general and can be applied to a broad class of optimization problems.