Multi-objective Evolutionary Design Research Articles

Gaussian Process-Accelerated Multiobjective Evolutionary Design of Charging Process Considering Multiple User Preferences

Gaussian Process-Accelerated Multiobjective Evolutionary Design of Charging Process Considering Multiple User Preferences

Evolutionary multi-objective design of autoencoders for compact representation of histopathology whole slide images

The recent success of deep learning coincides with the surge of digital pathology as an effect of the pandemic. However, processing a massive volume of gigapixel histopathology images is daunting. Hence, establishing compact image representations for efficient computational pathology becomes paramount. Deep autoencoders have been frequently used to compress feature vectors extracted from pre-trained networks. However, the topology of autoencoders is generally set heuristically. In this paper, we propose a multi-objective evolutionary framework as a bi-level optimization scheme in which the outer level evolves autoencoders for the goal of dimension reduction while the inner level is to optimize their weights for training. Throughout the evolution process, multiple minimization objectives such as complexity, classification error, and code size are considered. These objectives can obtain compressed models to generate a very small set of deep features with the highest classification accuracy. We use images from The Cancer Genome Atlas (TCGA) repository provided by the National Cancer Institute to train and validate our optimization framework. We increased the classification accuracy by 8% while the image representation is 46,000 times shorter. We apply the framework to demonstrate higher accuracy, substantial memory reduction and faster processing for image retrieval compared to SOTA.

Evolutionary Multiobjective Design Approach for Robust Balancing of the Shaking Force, Shaking Moment, and Torque under Uncertainties: Application to Robotic Manipulators

In this paper, the environmental uncertainties are taken into account when designing a robotic manipulator to balance the shaking force, shaking moment, and torque. The proposed robust balancing design approach does not consider the probability distributions of the uncertainties and is addressed without dependence on specific trajectories. This is expressed as a nonlinear constrained multiobjective optimization problem in which the nominal performance in the time-independent terms of the shaking force balancing, the shaking moment balancing, and the torque delivery, as well as their three sensitivities to uncertainties, are simultaneously optimized to provide a set of link shapes that match link mass distributions in a single stage. The proposal is applied to a three-degree-of-freedom serial-parallel manipulator, and the Non-Dominated Sorting Genetic Algorithm II (NSGA-II) is used to solve the associated problem. Comparative results with other design approaches reveal that the selected design achieves a suitable tradeoff in balancing the shaking force balancing, the shaking moment balancing, and the torque delivery and their sensitivities, leading to a reduction in their values and variations under mass changes in the manipulator end-effector with different operating conditions (tasks).

Novel ABC—Variable Based Closed-Form Analytical Model of Surface Mounted Permanent Magnet AC Machines With Sinusoidal and Non-Sinusoidal Back-EMF

This work presents a novel <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">abc</i> -based model applicable to surface-mounted permanent magnet AC (SM-PMAC) machines with sinusoidal and non-sinusoidal back-emf. It is capable of predicting the electromagnetic performance metrics such as torque waveforms, machine inductances, flux linkages and back-emf. The closed form expressions of the model, which can be evaluated with a high computational efficiency, are derived from basic geometric and winding parameters. Validation of the model is carried out numerically and experimentally with a very good match in results. Finally, the computational efficiency of the model is highlighted by considering a multi-objective evolutionary optimization design of SM-PMAC machine with a relatively large number of design parameters, where results are presented and discussed.

Multiobjective Evolutionary Design of Deep Convolutional Neural Networks for Image Classification

Convolutional neural networks (CNNs) are the backbones of deep learning paradigms for numerous vision tasks. Early advancements in CNN architectures are primarily driven by human expertise and by elaborate design processes. Recently, neural architecture search was proposed with the aim of automating the network design process and generating task-dependent architectures. While existing approaches have achieved competitive performance in image classification, they are not well suited to problems where the computational budget is limited for two reasons: 1) the obtained architectures are either solely optimized for classification performance, or only for one deployment scenario and 2) the search process requires vast computational resources in most approaches. To overcome these limitations, we propose an evolutionary algorithm for searching neural architectures under multiple objectives, such as classification performance and floating point operations (FLOPs). The proposed method addresses the first shortcoming by populating a set of architectures to approximate the entire Pareto frontier through genetic operations that recombine and modify architectural components progressively. Our approach improves computational efficiency by carefully down-scaling the architectures during the search as well as reinforcing the patterns commonly shared among past successful architectures through Bayesian model learning. The integration of these two main contributions allows an efficient design of architectures that are competitive and in most cases outperform both manually and automatically designed architectures on benchmark image classification datasets: CIFAR, ImageNet, and human chest X-ray. The flexibility provided from simultaneously obtaining multiple architecture choices for different compute requirements further differentiates our approach from other methods in the literature.

Multi-Objective Evolutionary Design of an Electric Vehicle Chassis.

An iterative algorithm is proposed for determining the optimal chassis design of an electric vehicle, given a path and a reference time. The proposed algorithm balances the capacity of the battery pack and the dynamic properties of the chassis, seeking to optimize the tradeoff between the mass of the vehicle, its energy consumption, and the travel time. The design variables of the chassis include geometrical and inertial values, as well as the characteristics of the powertrain. The optimization is constrained by the slopes, curves, grip, and posted speeds of the different sections of the track. Particular service constraints are also considered, such as limiting accelerations due to passenger comfort or cargo safety. This methodology is applicable to any vehicle whose route and travel time are known in advance, such as delivery vehicles, buses, and race cars, and has been validated using telemetry data from an internal combustion rear-wheel drive race car designed for hill climb competitions. The implementation of the proposed methodology allows to reduce the weight of the battery pack by up to 20%, compared to traditional design methods.

Efficient Surrogate-Based NVH Optimization of a Full Vehicle Using FRF Based Substructuring

&lt;div class="section abstract"&gt;&lt;div class="htmlview paragraph"&gt;The computer simulation with the Finite Element (FE) code for the structural dynamics becomes more attractive in the industry. However, it normally takes a prohibitive amount of computation time when design optimization is performed with running a large-scale FE simulation many times. Exploiting Dynamic Structuring (DS) leads to alleviating the computational complexity since DS necessities iterative reanalysis of only the substructure(s) to be optimally designed. In this research, Frequency Response Function (FRF) based substructuring is implemented to realize the benefits of DS for fast single- and multi-objective evolutionary design optimization. Also, Differential Evolution (DE) is first combined with two sorting approaches of Non-dominated Sorting Genetic Algorithm II (NSGA-II) and Infeasibility Driven Evolutionary Algorithm (IDEA) for effective constrained single- and multi-objective evolutionary optimization. The effectiveness of the proposed algorithm (NSGA-II/DE-IDEA) is verified using several test functions for constrained single- and multi-objective optimization. To circumvent the need for frequent time-consuming simulation runs, Kriging surrogate models are established by interpolating the responses simulated at the sample points, which are generated by executing an Optimal LHS algorithm. Besides, the Morris method is implemented to leave out unimportant design variables. A constrained single-objective and a constrained multi-objective NVH design optimization of a truck are carried out to demonstrate the surrogate-based design optimization process involving FRF based substructuring and the proposed algorithm.&lt;/div&gt;&lt;/div&gt;

Multi-objective evolutionary design of antibiotic treatments.

Antibiotic resistance is one of the major challenges we face in modern times. Antibiotic use, especially their overuse, is the single most important driver of antibiotic resistance. Efforts have been made to reduce unnecessary drug prescriptions, but limited work is devoted to optimising dosage regimes when they are prescribed. The design of antibiotic treatments can be formulated as an optimisation problem where candidate solutions are encoded as vectors of dosages per day. The formulation naturally gives rise to competing objectives, as we want to maximise the treatment effectiveness while minimising the total drug use, the treatment duration and the concentration of antibiotic experienced by the patient. This article combines a recent mathematical model of bacterial growth including both susceptible and resistant bacteria, with a multi-objective evolutionary algorithm in order to automatically design successful antibiotic treatments. We consider alternative formulations combining relevant objectives and constraints. Our approach obtains shorter treatments, with improved success rates and smaller amounts of drug than the standard practice of administering daily fixed doses. These new treatments consistently involve a higher initial dose followed by lower tapered doses.

A Multi-Objective Evolutionary Algorithm Model for Product Form Design Based on Improved SPEA2

As a Kansei engineering design expert system, the product form design multi-objective evolutionary algorithm model (PFDMOEAM) contains various methods. Among them, the multi-objective evolutionary algorithm (MOEA) is the key to determine the performance of the model. Due to the deficiency of MOEA, the traditional PFDMOEAM has limited innovation and application value for designers. In this paper, we propose a novel PFDMOEAM with an improved strength Pareto evolutionary algorithm 2 (ISPEA2) as the core and combining the elliptic Fourier analysis (EFA) and the entropy weight and technique for order preference by similarity to ideal solution (entropy-TOPSIS) methods. Based on the improvement of the original operators in SPEA2 and the introduction of a new operator, ISPEA2 outperforms SPEA2 in convergence and diversity simultaneously. The proposed model takes full advantage of this superiority, and further combines the EFA method’s high accuracy and degree of multi-method integration, as well as the entropy-TOPSIS method’s good objectivity and operability, so it has excellent comprehensive performance and innovative application value. The feasibility and effectiveness of the model are verified by a case study of a car form design. The simulation system of the model is developed, and the simulation results demonstrate that the model can provide a universal and effective tool for designers to carry out multi-objective evolutionary design of product form.

Multi-objective framework for cost-effective OTN switch placement using NSGA-II with embedded domain knowledge

The continuous growth of traffic coming from a plethora of bandwidth-hungry applications will drive network operators to further pursue strategies to cost-efficiently plan and dimension their transport networks. In view of this trend, this paper presents an evolutionary multi-objective design framework for routing a set of services in an optical transport network such that the key resources impacting capital expenditures (CapEx) – line interfaces and optical transport network (OTN) switches – are minimized. Particularly, the multi-objective problem is customized to select the most cost-effective network nodes to place OTN switches, while at the same time keeping the number of line interfaces required to a minimum. To solve the multi-objective design problem, different strategies were considered to produce the Pareto front of non-dominated solutions, using the Non-dominated Sorting Genetic Algorithm (NSGA-II). These strategies differ on how mutation and crossover solutions are generated: randomly or exploiting prior knowledge. The solution quality obtained with both strategies after a fixed number of generations is compared. The results indicate that embedding expert knowledge within the genetic algorithm leads to better convergence results. Moreover, the knowledge-based implementation of the genetic algorithm presents on average a 59% increase in the hyper volume rate when compared to the purely random evolutionary algorithm for the same number of generations.

Self-Adaptive Solution-Space Reduction Algorithm for Multi-Objective Evolutionary Design Optimization of Water Distribution Networks

An effective way to improve the computational efficiency of evolutionary algorithms is to make the solution space of the optimization problem under consideration smaller. A new reliability-based algorithm that does this was developed for water distribution networks. The objectives considered in the formulation of the optimization problem were minimization of the initial construction cost and maximization of the flow entropy as a resilience surrogate. After achieving feasible solutions, the active solution space of the optimization problem was re-set for each pipe in each generation until the end of the optimization. The algorithm re-sets the active solution space by reducing the number of pipe diameter options for each pipe, based on the most likely flow distribution. The main components of the methodology include an optimizer, a hydraulic simulator and an algorithm that calculates the flow entropy for any given network configuration. The methodology developed is generic and self-adaptive, and prior setting of the reduced solution space is not required. A benchmark network in the literature was investigated, and the results showed that the algorithm improved the computational efficiency and quality of the solutions achieved by a considerable margin.

Group Contribution Method-based Multi-objective Evolutionary Molecular Design

Abstract The search for compounds exhibiting desired physical and chemical properties is an essential, yet complex problem in the chemical, petrochemical, and pharmaceutical industries. During the formulation of this optimization-based design problem two tasks must be taken into consideration: the automated generation of feasible molecular structures and the estimation of macroscopic properties based on the resultant structures. For this structural characteristic-based property prediction task numerous methods are available. However, the inverse problem, the design of a chemical compound exhibiting a set of desired properties from a given set of fragments is not so well studied. Since in general design problems molecular structures exhibiting several and sometimes conflicting properties should be optimized, we proposed a methodology based on the modification of the multi-objective Non-dominated Sorting Genetic Algorithm-II (NSGA-II). The originally huge chemical search space is conveniently described by the Joback estimation method. The efficiency of the algorithm was enhanced by soft and hard structural constraints, which expedite the search for feasible molecules. These constraints are related to the number of available groups (fragments), the octet rule and the validity of the branches in the molecule. These constraints are also used to introduce a special genetic operator that improves the individuals of the populations to ensure the estimation of the properties is based on only reliable structures. The applicability of the proposed method is tested on several benchmark problems.

Multi-objective evolutionary design of granular rule-based classifiers

In the last years, rule-based systems have been widely employed in several different application domains. The performance of these systems is strongly affected by the process of information granulation, which defines in terms of specific information granules such as sets, fuzzy sets and rough sets, the labels used in the rules. Generally, information granules are either provided by an expert, when possible, or extracted from the available data. In the framework of rule-based classifiers, we investigate the importance of determining an effective information granulation from data, preserving the comprehensibility of the granules. We show how the accuracies of rule-based classifiers can be increased by learning number and parameters of the granules, which partition the involved variables. To perform this analysis, we exploit a multi-objective evolutionary approach to the classifier generation we have recently proposed. We discuss different levels of information granulation optimization employing both the learning of the number of granules per variable and the tuning of each granule during the evolutionary process. We show and discuss the results obtained on several classification benchmark datasets using fuzzy sets and intervals as types of information granules.

Repeated double cross-validation for choosing a single solution in evolutionary multi-objective fuzzy classifier design

The main advantage of multi-objective genetic fuzzy systems (MoGFS) is that a number of non-dominated fuzzy rule-based systems are obtained along the tradeoff surface among conflicting objectives. ...

Multi-objective evolutionary design of robust controllers on the grid

Coupling conventional controller design methods, model based controller synthesis and simulation, and multi-objective evolutionary optimisation methods frequently results in an extremely computationally expensive design process. However, the emerging paradigm of grid computing provides a powerful platform for the solution of such problems by providing transparent access to large-scale distributed high-performance compute resources. As well as substantially speeding up the time taken to find a single controller design satisfying a set of performance requirements this grid-enabled design process allows a designer to effectively explore the solution space of potential candidate solutions. An example of this is in the multi-objective evolutionary design of robust controllers, where each candidate controller design has to be synthesised and the resulting performance of the compensated system evaluated by computer simulation. This paper introduces a grid-enabled framework for the multi-objective optimisation of computationally expensive problems which will then be demonstrated using and example of the multi-objective evolutionary design of a robust lateral stability controller for a real-world aircraft using H∞ loop shaping.

Conceptual interplanetary space mission design using multi-objective evolutionary optimization and design grammars

Conceptual design optimization (CDO) is a technique proposed for the structured evaluation of different design concepts. Design grammars provide a flexible modular modelling architecture. The model is generated by a compiler based on predefined components and rules. The rules describe the composition of the model; thus, different models can be optimized by the CDO in one run. This allows considering a mission design including the mission analysis and the system design. The combination of a CDO approach with a model based on design grammars is shown for the concept study of a near-Earth asteroid mission. The mission objective is to investigate two asteroids of different kinds. The CDO reveals that a mission concept using two identical spacecrafts flying to one target each is better than a mission concept with one spacecraft flying to two asteroids consecutively.

Efficient Hybrid-Game Strategies Coupled to Evolutionary Algorithms for Robust Multidisciplinary Design Optimization in Aerospace Engineering

A number of game strategies have been developed in past decades and used in the fields of economics, engineering, computer science, and biology due to their efficiency in solving design optimization problems. In addition, research in multiobjective and multidisciplinary design optimization has focused on developing a robust and efficient optimization method so it can produce a set of high quality solutions with less computational time. In this paper, two optimization techniques are considered; the first optimization method uses multifidelity hierarchical Pareto-optimality. The second optimization method uses the combination of game strategies Nash-equilibrium and Pareto-optimality. This paper shows how game strategies can be coupled to multiobjective evolutionary algorithms and robust design techniques to produce a set of high quality solutions. Numerical results obtained from both optimization methods are compared in terms of computational expense and model quality. The benefits of using Hybrid and non-Hybrid-Game strategies are demonstrated.

Multi-objective Evolutionary Design of Robust Controllers on the Grid

AbstractThe emerging paradigm of grid computing provides a powerful platform for the solution of complex and computationally expensive problems. An example of this is the multi-objective evolutionary design of robust controllers, where each candidate controller design has to be synthesised and the resulting performance of the compensated system evaluated by computer simulation. This paper introduces a grid-enabled framework for the multi-objective optimisation of computationally expensive problems, before using the framework in the multi-objective evolutionary design of a robust lateral stability controller for a real-world aircraft using H∞ loop shaping.

Comparison of four reliability surrogate measures for water distribution systems design

Water distribution system (WDS) design naturally involves a trade‐off between the cost and reliability or robustness of a design. Traditionally, WDS reliability quantification schemes have employed graph theoretic techniques or probabilistic schemes such as Monte Carlo simulation. In recent decades there has been increased interest in the application of so‐called reliability surrogate measures, such as flow entropy, the resilience index, and network resilience, because of their ease of use and dramatically reduced computational burden. In this paper, these surrogate measures (as well as a mixed reliability surrogate) are employed in the multiobjective evolutionary design of a set of WDS benchmarks from the literature. The resulting Pareto‐optimal sets in cost‐reliability space for each surrogate measure are analyzed in terms of their ability to handle demand uncertainty and pipe failure, and a regression analysis is conducted in order to determine whether the surrogate measures are correlated to stochastic reliability and failure reliability as expressed by demand satisfaction measures. It is found that the resilience index demonstrates the best performance under pure stochastic demand variation. However, it lags when compared to the network resilience and mixed reliability measures in terms of reliability under pipe failure conditions. These are recommended as the most practical reliability surrogate measures for use in general WDS design, since they also produce designs that minimize size discontinuities between adjacent pipes. Flow entropy performs relatively poorly in terms of correlation to both stochastic reliability and failure reliability.

Multi-Objective Evolutionary Seismic Design with Passive Energy Dissipation Systems

The problem of multi-objective seismic design optimization is examined within the context of passive energy dissipation systems. In particular, a genetic algorithm approach is developed to enable the evaluation of the Pareto front, where maximum inter-story drifts and maximum total accelerations, both important measures for damage, serve as objectives. Here the cost of the passive system is considered as a constraint, although it could be included instead as a third objective. Hysteretic, viscoelastic and viscous dampers are all considered as possible design strategies, as well as the weakening plus damping concept. Since different types of passive systems are included, diversity of the Pareto front becomes a key issue, which is addressed successfully through an innovative definition of fitness. The multi-objective framework enables the evaluation of trade-offs between the two objectives and, consequently, provides vital information for the decision maker. Furthermore, the results presented offer valuable insight into the characteristics of optimal passive designs for the different objectives. Some of these characteristics confirm results reported elsewhere, while others are presented here for the first time.