Computer-aided Design Optimization Research Articles

Task-Similarity is a Crucial Factor for Few-Shot Meta-Learning of Structure-Activity Relationships.

Machine learning models support computer-aided molecular design and compound optimization. However, the initial phases of drug discovery often face a scarcity of training data for these models. Meta-learning has emerged as a potentially promising strategy, harnessing the wealth of structure-activity data available for known targets to facilitate efficient few-shot model training for the specific target of interest. In this study, we assessed the effectiveness of two different meta-learning methods, namely model-agnostic meta-learning (MAML) and adaptive deep kernel fitting (ADKF), specifically in the regression setting. We investigated how factors such as dataset size and the similarity of training tasks impact predictability. The results indicate that ADKF significantly outperformed both MAML and a single-task baseline model on the inhibition data. However, the performance of ADKF varied across different test tasks. Our findings suggest that considerable enhancements in performance can be anticipated primarily when the task of interest is similar to the tasks incorporated in the meta-learning process.

Synthesis of Antenna Array Radiation Pattern at Large Scanning Angles Using Genetic Algorithm

At present, in most modern communication systems, for example, in modern satellite terminals, the use of scanning beam antennas, i. e. antenna arrays is assumed. At the same time, at large scanning angles, the side lobe level (SLL) increases strongly and decrease in the gain is observed. In this regard, the problem of finding a procedure for synthesizing an amplitude-phase distribution (APD) with low SLL and high gain (G) at large scanning angles comes up. One of the ways to reduce SLL and compensate for the decrease in G is to synthesize the optimal APD (in terms of the maximum G and minimum SLL) using optimization algorithms. At the same time, taking into account the characteristics of the radiation pattern of the array emitters requires numerical electrodynamic calculation. The goal of this paper is to develop a procedure for the synthesis of APD with low SLL for linear and rectangular antenna arrays at various, including large, scanning angles and compensation for G reduction using a genetic algorithm and numerical electrodynamic calculation. The methods for studying the characteristics of antenna radiators are numerical electrodynamic modeling by the finite element method (FEM) in Ansys HFSS computer-aided design system and optimization of the APD for a given radiation pattern(RP) by a random search method using partial diagrams of antenna elements. The novelty is the combination of accurate numerical electrodynamic calculation of the RP of antenna elements and optimization search for APD for the synthesis of the required RP using partial diagrams. As a result, a procedure for APD synthesis of linear and uniform rectangular equidistant (for example, 8- and 64-element) antenna arrays has been developed, taking into account the exact electrodynamic characteristics of antenna elements and their mutual resistance. Radiation patterns were obtained taking into account the effect of neighboring elements, with the help of which, using a genetic algorithm, APDs on emitters were found at different scanning angles. The change in SLL and G of the antenna array is analyzed at different scanning angles using different APDs. The proposed algorithm allows to synthesize APD for a RP with low SLL and high G at scanning angles up to 40° for linear antenna array and up to 80° in the case of a uniform rectangular antenna array. The results of this work are relevant in the problems of radiation pattern synthesis, since the proposed solution provides a significant gain in the radiation pattern synthesis rate of APD of linear and rectangular antenna arrays, especially for systems with a large number of antenna elements. At the same time, it is possible to maintain a high G at large scanning angles, and achieve a significant reduction of SLL.

Computer-Aided Time-Domain Operation and Design Optimization of Resonant Converters

This paper presents a method for computer-aided operation and design optimization of resonant converters using time domain analysis. Resonant converters can operate in different operation stage trajectories (OSTs) with the change of two degrees of freedom (2DOFs), namely switching frequency <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">f_s</i> and duty ratio <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">d</i> . The paper calculates and analyzes the operation characteristics, zero voltage switching-on (ZVS-on), and zero current switching-off (ZCS-off) regions when the 2DOFs vary by solving the OST differential equation sets. Based on this analysis, an algorithm is proposed to find the operation and design optimizations of a specific resonant converter while considering 2DOFs by traversing all possible control coordinates. The paper provides several case studies to demonstrate the design trade-offs in different scenarios, which are verified through experimental verifications.

Radioharvester Research

The object of the research is an alternative energy source, energy control and distribution devices in relation to intelligent power sources of wireless sensor networks. The aim of the work is to develop a technical solution for an intelligent power supply to improve cybersecurity and energy efficiency in telecommunications. Results of the work: a schematic diagram of an intelligent power supply for wireless sensor networks in a virtual environment has been developed; simulation modeling of physical processes using a computer-aided design system, parameterization and optimization of the power supply source has been performed. The given behavioral models allow the use of any software environment for circuit modeling. The developed behavioral model of the power supply circuit of wireless sensor networks based on the LT3980 chip can be used to create similar devices. Modeling makes it possible to increase the reliability of choosing the optimal solution for the structure, increase reliability even at the design stages, optimize the technical characteristics, design of the device being developed and reduce its cost.

Computer-aided architectural design optimization based on BIM Technology

This article addresses the problem of the non-circulation of information in each stage of architectural design. This paper explores the architectural design process based on the BIM platform and puts forward the structural design method based on the BIM platform. It carries out the seismic analysis of a high-rise building with a transfer floor structure and compares the analysis results with the structural analysis software commonly used by the current design institute. The results obtained for experimentation show that the period ratio, displacement ratio, and the first six modes calculated by the two methods in the modal analysis are consistent. The error between calculation results and PKPM calculation results is within a reasonable range. In the analysis of the mode decomposition response spectrum method, the seismic forces in X and Y directions, floor shear, overturning moment, floor average displacement, and displacement angle obtained by the two models are compared respectively. The analysis results of the two methods accord with the mechanical characteristics of the transfer floor structure, and the calculation error is within the allowable range. The structural design based on the BIM platform has the advantages of high visualization, parameter-driven component size, and high model accuracy, improving design drawing efficiency.

Computer-aided efficient design and performance optimization of cutting head for roadheader

The cutting head is the core working mechanism of the roadheader for coal-rock materials cutting. The efficient and high performance design of cutting head is the key to improve the road head digging and mining technology. In this paper, based on cutting head design theory and virtual prototype technology, we propose a computer-aided structure design and performance optimization method for cutting head. We compile the calculation code and realize the reading and storing of relevant data through Excel. In particular, to obtain more realistic cutting performance data of the cutting head, we construct a coupling model of cutting head cutting rock wall based on virtual prototype technology, and then establish a database matching structural parameters, working parameters, coal-rock properties and cutting performance through extensive simulations. Based on the method, we complete the design of EBZ220 roadheader cutting head. We show that our method can realize the fast and efficient design of cutting head, and the designed cutting head has good working performance.

Computer-aided optimal design for flexible cable in aerospace products based on dynamic analogy modeling

Due to large, complex deformations, the accurate design of cables has become a major problem in the manufacturing of aerospace products. The current design method often leads to large products, uncertain centroids, and poor reliability. To solve these problems, a computer-aided optimal design method for flexible cables was proposed based on dynamic analogy modeling. A nonlinear optimization model was established by combining Cosserat theory and the minimum potential energy principle. The total deformation energy was considered as the optimization object, and Euler parameters were used as control variables to describe the cable geometric shape. Considering the length and bending radius requirements, the normalized form of the cable constraints was expressed by the cross-section position and orientation matrix. An efficient method to solve this problem using finite element discretization and the primitive dual interior point method was proposed. A digital wiring module was developed based on an open source geometry kernel system, and a cable geometry test bench was built. To verify our model, a satellite wiring simulation example was implemented using the module, SolidWorks, and the test bench. Our method achieved the optimal design for the cable length and geometric shape. A theoretical and technical foundation for effectively solving the problem of large cable manufacturing errors and realizing the lightweight design of aerospace products was outlined.

Компьютерное проектирование больших трубопроводных сетей высокого ранга оптимальности

The paper presents a computer-aided optimal design of hydraulic systems. The method is based on dividing the network synthesis problem into two stages resulting in a dimensional reduction. This allows the optimal networks design of the higher rank. The proposed technique is intended for automated design of large pipeline networks for rural and interregional water supply.

Computer-Aided Efficient Design and Performance Optimization of Cutting Head for Roadheader

Computer-Aided Efficient Design and Performance Optimization of Cutting Head for Roadheader

Interactive Architectural Design with Diverse Solution Exploration.

In architectural design, architects explore a vast amount of design options to maximize various performance criteria, while adhering to specific constraints. In an effort to assist architects in such a complex endeavour, we propose IDOME, an interactive system for computer-aided design optimization. Our approach balances automation and control by efficiently exploring, analyzing, and filtering space layouts to inform architects' decision-making better. At each design iteration, IDOME provides a set of alternative building layouts which satisfy user-defined constraints and optimality criteria concerning a user-defined space parametrization. When the user selects a design generated by IDOME, the system performs a similar optimization process with the same (or different) parameters and objectives. A user may iterate this exploration process as many times as needed. In this work, we focus on optimizing built environments using architectural metrics by improving the degree of visibility, accessibility, and information gaining for navigating a proposed space. This approach, however, can be extended to support other kinds of analysis as well. We demonstrate the capabilities of IDOME through a series of examples, performance analysis, user studies, and a usability test. The results indicate that IDOME successfully optimizes the proposed designs concerning the chosen metrics and offers a satisfactory experience for users with minimal training.

Design of an Optimum Single Phase Inverter for a Grid Tie PV System

Power converter optimization by genetic algorithm (GA) is used to provide simpler and more reliable converter design for high efficiency, small size and low cost. This paper presents a Computer-Aided Design Optimization Tool based on GA to determine the optimal structure of single-phase voltage source inverter devoted to grid-connected photovoltaic applications. An accurate non-linear averaged model was used to model the power converter. The hysteresis technique was used to control the output sine wave current of the inverter while the Elitist Non-dominated Sorting Genetic Algorithm NSGA-II was used to search the Pareto optimal front and the best design in terms of efficiency, volume and cost under electrical constraints. The converter model and the NSGA-II algorithm are developed in the MATLAB/Simulink environment. The problem formulation was detailed. It was shown that the optimization of a power converter, working in a given application without the need of tedious and expensive experimental tests classically used to build this converter, is possible by mean of simulation. This will decrease time to market phase for manufacturers.

Analysis of a combined cycle propulsion system for STRATOFLY hypersonic vehicle over an extended trajectory

Hypersonic civil aviation is an important enabler for extremely shorter flight durations for long-haul routes and using unexploited flight altitudes. Combined cycle engine concepts providing extended flight capabilities, i.e. propelling the aircraft from take-off to hypersonic speeds, are proposed to achieve high-speed civil air transportation. STRATOFLY project is a continuation of former European efforts in hypersonic research and aims at developing a commercial reusablevehicle for cruise speed of Mach 8 at stratospheric altitudes as high as 35 km above ground level. The propulsion plant of STRATOFLY aircraft consists of combination of two different type of engines: an array of air turbo rockets and a dualmode ramjet/scramjet. In the present study, 1D transient thermodynamic simulations for this combined cycle propulsion plant have been conducted between Mach 0 to 8 by utilizing 1D inviscid flow transport relations, numerical tools availablein EcosimPro software platform and the European Space Propulsion System Simulation libraries. The optimized engine parameters are achieved by coupling EcosimPro software with Computer Aided Design Optimization which is a differential evolution algorithm developed at the von Karman Institute.

A new overhang constraint for topology optimization of self-supporting structures in additive manufacturing

This work falls within the scope of computer-aided optimal design, and aims to integrate the topology optimization procedures and recent additive manufacturing technologies (AM). The elimination of scaffold supports at the topology optimization stage has been recognized and pursued by many authors recently. The present paper focuses on implementing a novel and specific overhang constraint that is introduced inside the topology optimization problem formulation along with the regular volume constraint. The proposed procedure joins the design and manufacturing processes into a integrated workflow where any component can directly be manufactured with no requirement of any sacrificial support material right after the topology optimization process. The overhang constraint presented in this work is defined by the maximum allowable inclination angle, where the inclination of any member is computed by the Smallest Univalue Segment Assimilating Nucleus (SUSAN), an edge detection algorithm developed in the field of image analysis and processing. Numerical results on some benchmark examples, along with the numerical performances of the proposed method, are introduced to demonstrate the capacities of the presented approach.

A Pipe Routing Method Considering Vibration for Aero-Engine Using Kriging Model and NSGA-II

The computer-aided design optimization and vibration analysis, especially avoiding resonance are fundamental challenge to piping system design of complex products. Conventional pipe design methods commonly focus on either automatic geometry layout or vibration analysis, which usually leads to repeated modifications between both disciplines. In this paper, a multi-objective pipe routing algorithm considering vibration performance for aero-engine is presented based on Kriging model and non-dominated sorting genetic algorithm-II (NSGA-II), where pipe length, smoothness, and natural frequency are taken into account. Pipe node positions are optimized by using NSGA-II subject to several constraints. In order to improve computational efficiency, a Kriging model reflecting the relationship between the pipe node position and natural frequency is constructed, and then is applied instead of computer-aided engineering analysis program to evaluate the individual fitness during evolutionary computations. Several numerical computations are performed on a simplified piping model, and the results show that the constructed Kriging model can guarantee high modeling accuracy and the presented routing algorithm is able to obtain a set of non-dominated solutions of pipe layouts, in particular, the natural frequency of pipes are far from engine excitation frequency, which guarantees to avoid possible resonance.

Multiobjective Optimization of Medium-Frequency Transformers for Isolated Soft-Switching Converters Using a Genetic Algorithm

The main challenge of medium-frequency transformers is the number of design parameters, constraints and objectives, and the difficulty of handling them on a particular design. This paper presents a novel computer-aided optimal design for MF transformers using a multiobjective genetic algorithm, in particular the nondominated sorting genetic algorithm II. The proposed methodology has the aim of reaching the best MF transformer for a given power converter topology, by optimizing transformer efficiency, weight, and also, transformer leakage and magnetizing inductances at the same time. The proposed methodology and the optimal solutions are validated with the design and the development of two 10-kVA/500-V transformers considering two different topologies. Finally, some experimental measurements are presented so as to demonstrate the proposed models and the performance of built transformers.

Fatigue-based structural optimisation of vehicle components

This paper deals with a new product design framework for the optimal structural design applications. The proposed approach called hybrid charged system search (CSS) and Nelder-Mead algorithm (HCSSNM) offers significant improvements in performance over existing methods in the field of computer-aided design optimisation. The HCSSNM based on the CSS and Nelder-Mead algorithm is used for optimisation of a vehicle suspension arm which is critically important for the vehicle safety. As observed in the results, the HCSSNM can be used effectively to reduce the mass of the vehicle components while maintaining its fatigue durability.

통계적 기계학습 기술을 이용한 시뮬레이션 결과 예측 시스템 개발

컴퓨터 시뮬레이션은 전산유쳬역학, 나노 물리, 계산화학, 구조 동역학, 전산설계 등 여러 계산과학공학 분야에서 시스템의 움직임을 예측하기 위해 널리 사용되고 있다. 하지만 시뮬레이션의 정밀도와 복잡도가 점점 증가함에 따라 시뮬레이션을 수행하는 비용 역시 크게 증가하고 있다. 따라서 시뮬레이션의 수행비용을 줄이는 것은 특히 입력 변수들의 값을 변화시켜가며 시뮬레이션을 반복적으로 수행하는 경우, 시뮬레이션 수행 시간 단축을 위해 매우 중요하다. 본 논문은 어떤 시뮬레이션의 수행이 요청되었을 때, 해당 시뮬레이션을 실제로 수행하지 않고도 기존에 수행된 시뮬레이션의 결과를 저장하여 이전에 획득되거나 혹은 예측된 결과를 반환하는 시스템을 개발한다. 이를 위해 본 논문에서 개발된 시스템은 크게 다음 2가지 기능을 제공한다: (1) 수행이 완료된 시뮬레이션의 결과를 데이터베이스에 저장하는 기능, (2) 사용자가 요청한 시뮬레이션의 결과를 통계적 기계학습 기술을 사용하여 예측하는 기능. 본 논문에서 개발한 예측 시스템의 예측 성능을 실제 유체역학 시뮬레이션 데이터를 사용하여 평가한 결과, 출력변수에 따라 0.9%의 매우 낮은 평균 예측 오차율을 보였다. 본 논문에서 개발한 시스템을 통하여 사용자들은 계산 및 저장 자원에 큰 부하를 주는 시뮬레이션을 실제 수행하지 않고도, 수행을 원하는 시뮬레이션의 결과를 빠르게 예측해 볼 수 있다. Computer simulation is widely used in a variety of computational science and engineering fields, including computational fluid dynamics, nano physics, computational chemistry, structural dynamics, and computer-aided optimal design, to simulate the behavior of a system. As the demand for the accuracy and complexity of the simulation grows, however, the cost of executing the simulation is rapidly increasing. It, therefore, is very important to lower the total execution time of the simulation especially when that simulation makes a huge number of repetitions with varying values of input parameters. In this paper we develop a simulation service system that provides the ability to predict the result of the requested simulation without actual execution for that simulation: by recording and then returning previously obtained or predicted results of that simulation. To achieve the goal of avoiding repetitive simulation, the system provides two main functionalities: (1) storing simulation-result records into database and (2) predicting from the database the result of a requested simulation using statistical machine learning techniques. In our experiments we evaluate the prediction performance of the system using real airfoil simulation result data. Our system on average showed a very low error rate at a minimum of 0.9% for a certain output variable. Using the system any user can receive the predicted outcome of her simulation promptly without actually running it, which would otherwise impose a heavy burden on computing and storage resources.

Design of a dual slot antenna for small animal microwave ablation studies.

This study presents the development of a dual-slot antenna for small animal tumor ablation. By using a dual-slot design at 8 GHz, it was hypothesized that smaller and more spherical ablations can be produced. After computer-aided design optimization, antennas were fabricated and ablations performed at 5-20 W for 15-120 s with the objective of creating ablations with a diameter/length aspect ratio of at least 0.9. The new dual-slot design at 8 GHz created significantly more spherical ablations than a commercial antenna at 2.45 GHz in ex vivo liver tissue (Average Aspect Ratio 0.8081 vs. 0.4532, p <;<; 0.05). In vivo studies confirmed the highly spherical results ex vivo. Initial testing shows that the dual-slot antenna and 8 GHz generator can be used to ablate tumors in mice.

Computer-Aided Thermal Design Optimisation

The aim of this paper is to demonstrate by means of relevant examples the usefulness of Excel as a tool for introducing computer-aided design optimisation to engineering students.The performance of thermal systems is strongly influenced by the cost of energy which constitutes a major part of their running cost.Thermal design, which must take into consideration operating costs as well as initial costs, offers good examples of design optimisation. Moreover, design optimisation of thermal systems can easilybe performed by using Excel. The examples considered in the paper deal with insulated conduits carrying hot or cold air with respect to initial and energy costs. Unlike analytical optimisation that can only be used for simple situations with a single design parameter, Excel can deal with thermal designs that involve multiple design factors and elaborate analytical models.

Bi-objective sizing optimization of power converter using genetic algorithms

Purpose – This paper aims to focus on the trade-off between losses and converter cost. Design/methodology/approach – The continual development of power electronic converters, for a wide range of applications such as renewable energy systems (interfacing photovoltaic panels via power converters), is characterized by the requirements for higher efficiency and lower production costs. To achieve such challenging objectives, a computer-aided design optimization based on genetic algorithms is developed in Matlab environment. The elitist non-dominated sorting genetic algorithm is used to perform search and optimization, whereas averaged models are used to estimate power losses in different semiconductors devices. The design problem requires minimizing the losses and cost of the boost converter under electrical constraints. The optimization variables are, as for them, the switching frequency, the boost inductor, the DC capacitor and the types of semiconductor devices (IGBT and MOSFET). It should be pointed out that boost topology is considered in this paper but the proposed methodology is easily applicable to other topologies. Findings – The results show that such design methodology for DC-DC converters presents several advantages. In particular, it proposes to the designer a set of solutions – as an alternative of a single one – so that the authors can choose a posteriori the adequate solution for the application under consideration. This then allows the possibility of finding the best design among all the available choices. Furthermore, the design values for the selected solution were obtainable components. Originality/value – The authors focus on the general aspect of the discrete optimization approach proposed here. It can also be used by power electronics designers with the help of additional constraints in accordance with their specific applications. Furthermore, the use of such non-ideal average models with the multi-objective optimization is the original contribution of the paper and it has not been suggested so far.