At present, Evolutionary Algorithms that Genetic Algorithm represents are widely applicable to the optimization problem of the various engineering fields. Then, the effectiveness is reported. Authors already applied to the topology optimization of structure problem and reported the effectiveness of it too. But, Evolutionary Algorithms require a large amount of computational cost to obtain the optimal solution steadily. In this paper, Authors developed new method, Adaptive Plan system with Genetic Algorithm (APGA) for solving this cost. This is synthesis method of global search ability of GA and the Adaptive Plan with excellent local search ability. The Adaptive Plan is defined a numerical algorithm or an experiential method. The standard test functions examines to APGA, through number of design variable. A performance of APGA is discussed.

View full abstract

Simulated annealing (SA) is an effective general heuristic method for solving many combinatorial optimization problems. This paper deals with two problems in SA. One is the long computational time of the numerical annealings. The other one is the determination of the appropriate temperature schedule in SA. The proposed method is based on the concept of the Equilibrium Binary Search which is proposed by Jonathan, and controls the temperature schedule adaptively. In addition, It can reduce the computational time of the numerical annealings maintaining high searching ability. The proposed method is applied to solve many TSPs (Traveling Salesman Problems), and it is found that the method is very useful and effective.

View full abstract

The optimization of the various processes from the development to production of the product is carried out. However, there are many cases that are the approximation solution by using RSM(response surface modeling) and also request a local optimum solution by utilizing the mathematical programming. And generally most of the design variables in optimum object include the continuous numbers and discrete numbers. Thereupon, hi addition to the former method of zone division for the continuous numbers, we propose the new method that deal with the continuous and discrete design variables. The effectiveness of proposed method has been confirmed by several numerical examples and a real design problem.

View full abstract

We propose a parallel distributed interactive genetic algorithm(PDIGA) as a new design collaboration method. PDIGA uses an IGA and a parallel distributed genetic algorithm, which combines several IGA systems, and it has a scheme that best ones among the design solutions based on the subjective evaluation of each user are shared among users for every generation. A collaboration system using PDIGA is developed to make good design solutions among several people at difference locations. To verify a validity of the PDIGA system, we conducted experiments for comparing IGA and PDIGA. In particular, we examined each user's evaluation to the final design, compared the averages of the individual evaluation value, and examined the similarity of the design solution. These showed that the collaboration system using PDIGA become a consensus building system, and users' design solutions are unified in a group.

View full abstract

This paper presents a method to find some local minima by branching at local minimum, based on the Generalized Random Tunneling Algorithm (GRTA). We call this method as Branching GRTA( BGRTA). BGRTA also consists of three phases, that is, the minimization phase, the tunneling and branching phase, and constraint phase. In the minimization phase, local search technique is used. In the tunneling and branching phase, some local minima can be found by branching at local minimum obtained in the minimization phase. The next branching point is chosen from them. In the constraint phase, the feasibility is examined. We apply BGRTA to the traffic road design problem. Through numerical examples, we examine the validity of BGRTA.

View full abstract

Simulated Annealing Method is used as an effective solution method of combinatorial optimization problems, but it has the problem that the settlement of the temperature which is an optimization parameter and scheduling are difficult. So, How to schedule temperature is proposed as an approach for this problem by using Q-Learning which is one of the reinforcement learning. With this technique, every time cooling is done, the state of the solution in a metropolis loop is observed, scheduling of temperature is performed from the state and the result of learning and it learns from that result again. By this technique, the loss in search is reduced by performing temperature scheduling using the history of search and even if a parameter is not adjusted in accordance with the problem, it is expected that it finds an answer firmly. From the result of this numerical experiment, the expected effect was acquired and the usefulness of this technique has been checked. A performance of the proposal methodology is discussed.

View full abstract

Simulated annealing (SA) is an effective general heuristic method for solving many optimization problems. This paper deals with two problems in SA. One is the long computational time of the numerical annealings, and the solution to it is the parallel processing of SA. The other one is the determination of the appropriate neighborhood range in SA, and the solution to it is the introduction of an adaptive mechanism for changing the neighborhood range. The multiple SA processes are performed in multiple processors, and the neighborhood ranges in the SA processes are determined by the evolutionary computation. The proposed method is applied to solve many continuous optimization problems, and it is found that the method is very useful and effective.

View full abstract

SEPA2+, one of the multi objective genetic algorithms, has two archives: one of them is for restoring the Pareto solutions that have the wide variety in the objective space and the other is for archiving the Pareto solutions that have the variety in the design variable space. In this paper, we examined the effect of two archives of SPEA2+ on solution search through the numerical examples. The results showed the following two tendencies: the archive for design variable helps to keep the diversity not only in the design variable space but also in the objective space and the diversity in the design variable space begins to maintain after the solutions are converged into local Pareto solutions. Inspiring these results, we proposed the new way how to use these two archives. In the proposed method, the population in the archive for objective function space is used for genetic operations until the number of the non-dominated solutions becomes bigger than the archive size. After that, the population in the archive for design variable space is used for genetic operations. In the comparison of the proposed method with SPEA2 and SPEA2+, the results of the proposed method showed the same precision of the derived Pareto solutions with the higher diversity in the design variable space.

View full abstract

Most practical problems are formulated as multi-objective optimization problem with conflicting objectives so as to meet diversified demands of a decision maker. Generally, there is no solution to optimize all objective functions simultaneously in multi-objective optimization, thus Pareto optimal solution is introduced. Recently, many methods using evolutionary algorithms have been developed for generating Pareto optimal solutions. In the cases with two or three objective functions, especially, those methods been shown to be useful for visualizing the set of Pareto optimal solutions, which is called Pareto frontier. This paper proposes a multi-objective optimization technique by using computational intelligence in order to generate well approximate Pareto frontiers. By combining a machine learning algorithm and an evolutionary algorithm, it will be shown that the proposed method can find effectively, actively and fast the close Pareto frontier to the real one, through several numerical and practical examples.

View full abstract

Getting adequate information for determining a final design solution is not an easy task especially if there arc many optimum and quasi-optimum solutions in a multi-dimensional space. To overcome the difficulty, a synchronous three dimensional visualization is proposed Through the interactive operation of the present visualization system, engineers can explore and understand the correlation among multi-dimensional function and parameter spaces and the real world space. Moreover to find distinct features of the solutions, an enhanced clustering algorithm is applied to the optimum and quasi-optimum solutions. As a practical engineering application, the proposed approaches arc applied to a multi- objective optimization of heat pipe design of artificial satellite. The synchronous visualization helps an engineer to understand the design effect of each solution out of the Pareto and quasi-Pareto solutions. The clustering of the optimum and cjuasi-optimum solutions guides the shape parameters those corresponds to worse or better design clearly. Thus it is confirmed that the information gleaned from the proposed methods can support engineering decision-making on the design of heat pipes effectively.

View full abstract

In this paper, we are going to propose trend analysis using the data envelopment analysis. We have been studying characteristic analysis of products using DEA in the past study, and showed that with publication data, we can analyze the characteristics products in the same categories, based on analyzing weight on each criteria. In this study, we will propose a method to measure the growth of frontier. Then by using history of the change in the shape of characters we will propose a method of looking at trends of the products that will fit to the customers.

View full abstract

This paper presents a new method for combinatorial optimization problems. Most of the actual problems that have discrete structure can be formulated as combinatorial optimization problems. It is experientially known that Proximate Optimality Principle (POP) holds in most of the actual combinatorial optimization problems. The concept of Proximate Optimality Principle says that good solutions of most real combinatorial optimization problems have the structural similarity in parts of solution. In this paper we propose a new optimization method based on Tabu Search. In the proposed algorithm, POP is taken into consideration. The proposed algorithm is applied to some knapsack problems, which are typical combinatorial optimization problems in order to verify the performance of the proposed algorithm.

View full abstract

The Particle Swarm Optimization (PSO) method is one of the most powerful methods for solving unconstrained and constrained global optimization problems. Little is, however, known about an adaptive strategy for tuning the parameters of the PSO method in order to apply the PSO method to large-scale nonlinear noncovex optimization problems. This paper deals with an adaptive strategy for tuning the parameters of the PSO method based on the analysis of the dynamics of PSO. While the relation between the dynamics of average velocity of the particles and successful search processes is analyzed, an adaptive tuning strategy for adaptive search is proposed based on the investigated relation. The feasibility and the advantage of the proposed adaptive PSO method is demonstrated through some numerical simulations using a typical global optimization test problem.

View full abstract

It is well known that Particle Swarm Optimization (PSO), which was originally proposed by J. Kennedy et al., is a powerful algorithm for solving unconstrained and constrained global optimization problems. Appropriate adjustment of its parameters, however, requires a lot of time and labor when PSO is applied to real optimization problems. This paper investigates the adaptive Particle Swarm Optimization from the viewpoint of search history. Some numerical simulations were carried out in order to examine the adaptability of the proposed approach.

View full abstract

As an essential part of recent business decision-making, this study has called attention on the entire transportation network associated with supply-chain management. To optimize such large-scale and complex problems, we proposed a hybrid method equipped with meta-heuristic method and graph algorithm after decomposing the problem in a hierarchical manner. Then we provide a novel adjusting method to obtain a consequent solution from each result solved individually. Through numerical experiments, we verified the performance of the proposed method.

View full abstract

Recently technological advances made systems more sophisticated, and the problems more complicated. To adapt to complex conditions, the behavior of a system should be optimized. For example, robots, aircrafts, spacecrafts and Aero-assited Orbital Transfer Vehicles (AOTV) have sophisticated trajectory programming problems. This study presents the proposal of two new general methods to solve optimum control problem. Applying Minimum Principle to parameterize optimum control problem by initial value of adjoint variables p_0. The problem can be solved by direct optimization using general Non-Linear Programing solutions. The proposed methods offer faster numerical solutions for optimum trajectory programming problems. In order to see whether the proposed methods are advantageous, the results of a comparative study with conventional methods are shown.

View full abstract

This paper proposes a multidisciplinary design approach of piezoelectric actuator and control system for smart structures that have a characteristics variation. The length of a beam-type smart structure changes and so the characteristics of the system is varied. The system is composed of piezoelectric actuator array and scheduled controller. The control system is designed according to the H_2 specification with a reduced-order modal model. An adaptive control is conducted by scheduling the actuators and the controllers to keep the stability and performance against the characteristics variation of the system. The design problem for improving the H_2 performance is defined, and then the piezoelectric array and the scheduled control system are simultaneously optimized by the presented approach, resulting in an enhanced performance for the vibration control.

View full abstract

This paper studies effects of layout of PZT patches on performance of vibration control in shunt damping. First, for a beam with multiple PZT patches a method to design the resistors and inductors of the shunt circuits is presented. Then, using the presented method to design the shunt circuits, effects of layout of the PZT patches such as position, length and division of the PZT patch on the performance of the vibration control is discussed. It is shown that.division of the PZT patch can increase the performance of the vibration control as well as that the optimal length and position of the PZT patch exists.

View full abstract

Fluctuation waves of pressure and flow rate in closed pipe-system are obtained by a pump and a valve controlled with cyclic and arbitrary operation of rotational speed and valve opening. A recurrent neural-network computer program learns response characteristics of the system. Rotational speed of pump and valve opening are obtained from this neural-network computing so that pressure and flow rate achieves required variation. This control method is studied in this report. And this method is investigated on computing simulation of the pipe system as preliminary step toward applying experimental apparatus. It made clear that, if variable time-series are measured from the operated system, target fluctuations can generate with high accuracy by using the recurrent neural network computing which has learned the response characteristics, without using characteristics of each device.

View full abstract

The purpose of this paper is to specify the positions and the sizes of cracks by a percussive sound. At first, a cracked element in FEM analysis is developed for both a surface crack and an internal crack for obtaining monotone decreasing eigenvalues for all degrees. Next, a converse theorem of Weyl's monotony theorem is constructed in order to inspect the accuracy of the crack elements. A method for searching and identifying various cracks is developed based on an efficient exhaustive search method composed of the hierarchical algorithm using the evaluation function. Finally, the validity of this method is verified from several calculation examples.

View full abstract

Spine structure is mainly constructed of many vertebrae and flexible intervertebral disks, and it is unstable by itself under standing condition. Supporting by erector muscles and ligaments keep the spine standing condition, so that boundary condition of each vertebra depend on the muscle and ligament forces. Moment applied to vertebra by intervertebral disk deformation also has to be considered for the boundary condition of vertebra Biomechanical simulation of musculoskeletal model taking account of muscle and ligament forces and intervertebral joint moments is necessary to determine the boundary condition of vertebral analysis. In this study, a musculoskeletal model around lumber vertebrae was developed, and loading condition of each vertebra was analyzed considering muscle and ligament forces and joint moments. Number of unknown variables, which were muscle and ligament forces and intervertebral joint moments, was greater than number of the equilibrium equations. That is, the problem to solve the force and moment was higher order indeterminate problem. In order to solve the problem, an optimization problem has to be formulated. We considered an objective function corresponds to work done by muscle force and intervertebral joint moment This is a typical quadratic programming problem, so the optimization problem was solved by quadratic programming method. Muscle forces and intervertebral joint moments were evaluated varying flexion or extension angle, and efficiency of the musculoskeletal model and proposed method to obtain the muscle forces and intervertebral joint moments was discussed.

View full abstract

The initial condition setting problem of physics base CG animation is considered as inverse problem and formulated as the minimization problem of error which is the difference of target and final stage. The response surface method based on the moving least square approximation is proposed, which enables the approximation of complicated non-smooth responses, and the efficient placement procedure of trial points is proposed. Several examples with several design variables and rigid body simulation with bouncing are shown.

View full abstract

Piezoceramics are piezoelectric materials that can convert electric energy into mechanical energy, and vice versa. When piezoceramics are used in combination with elastic materials, they can provide sophisticated actuations in a range of devices. This paper proposes a structural topology optimization method for obtaining multiple-actuation piezoelectric actuators composed of piezoceramics and flexible structures. First, a new type of homogenization design method is introduced, in which continuous material distributions are assumed using a continuous interpolation function at each node, in order to overcome numerical instabilities. Next, actuator design requirements are clarified and the corresponding objective functions are formulated. A new type of multi-objective function is presented to enable multiple-actuation designs and an optimization algorithm is developed using Sequential Linear Programming (SLP). Finally, several examples are provided in order to confirm the usefulness of the generated optimal designs for the multiple-actuation piezoelectric actuators.

View full abstract

In the complicated design problems, the design variables are large and to search the solutions is difficult. Therefore, a optimization technique without increasing the design variables is desirable. In this paper, such two techniques are proposed for the structural optimization problem as one of the example of the complicated design problems. The first is the method processing the design space in the multi stage, and the second is the method using the quad-trees data structure. The validity of these techniques is made clear to the design problems of continuity structure.

View full abstract

Topological structural optimization contributes to the optimum design of structures with the miriimum weight mamtaining a high stiffness. This paper presents a! method for achieving topological optimization involving the optimum layout of plural materials for maximizing structural buckling strength. This method is formulated based on both the constraint condition concerning the layout of materials with various characteristics in the topological optimization and the stability theory of an elastic body with initial stresses. Then, an asteroid'type virtual element is developed based on a substructuring method for the purposes of maintaining a high level of accuracy in a relatively short CPU time. This method is applied to the two-dimensional model of column, and the validity of this method is confirmed from both the geometrical shape and the material layout regarding the obtained optimization structures. Various optimization shapes and material layouts are illustrated on the columns supporting the compressive load in the eccentric axial directions of those.

View full abstract

This paper presents an effective method for designing structures using a cellular automata, representing a simple conceptual basis for the self-organization of structural systems. The proposed methods are sufficiently simple to solve topology optimization problems as pure 0-1 problems, and yet sufficiently complex to express a wide variety of complicated topologies. Local rules for the birth and death of cells are all that is required for this method. The effectiveness of the proposed method is demonstrated through numerical topology optimization problem examples. The method proposed in this paper offers a new approach to structural optimization, overcoming most of the problems associated with traditional techniques.

View full abstract

This paper suggests a new engineering design methodology by utilizing the optimality of branching network in nature, such as botanical tree and root systems, animal vascular systems, and so on. The growth mechanism of branching networks in nature is studied and the generation process is reproduced. By applying the generation method of the branching network, a branch-like cooling pipe system is constructed inside a plate with a certain heat flux distributed on its upper surface. Finite element analysis of temperature distribution is implemented, and the cooling performances of such cooling systems are discussed. The effectiveness of the generation method is also confirmed by comparing the result with that from conventional design method.

View full abstract

In this paper, a nonparametric boundary shape optimization technique based on traction method considering die drawing and draft angle is proposed. Shape variation in general traction method is obtained as a solution of a boundary value problem of linear elastic continuum loaded with traction in proportion to the shape gradient of objective functional or constraint functional. A penalty term, denotes a virtual work by draft angle variation of boundary surface, is introduced into the equation oi virtual work in the shape variation analysis. An example of shape optimization problems considering the die drawing was analyzed using the technique. Appropriate result shape is obtained without any remarkable additional cost such as memory area, calculation time and troublesome optimization data preparation.

View full abstract

In shape optimization, an optimal shape is generated by moving some nodes of an initial mesh. But there are many cases that meshes can not follow the movement of nodes. In this paper, the shape optimization technique using mesh superposition method is proposed. In the present technique, the shape of design domain is represented by local mesh to allow flexible shape control. The shape of the local mesh is controlled by basis vector method. Several examples which can not optimize in the conventional optimization technique are shown to demonstrate the effectivity of the present method.

View full abstract

This paper reports on a new multi-disciplinary shape optimization approach, which is aimed to use in conceptual design stage and enable quick reviews of numerous design alternatives with body in white (BIW) vehicle model. In the proposed method, shape model of BIW consist, of design lines of body frame and its cross-section, and does not needs CAD drawings fully controlled by optimization technology. Basis function method is used for representing the cross-section geometries and change them. Major difference between proposal method and conventional one is, in our method, the bases are reformulated based on contribution of performance characteristics before they are used in optimization. Due to the reformulation, each basis has physical meaning (ex. better for stiffness, better for crashworthiness and so on). Also it significantly decreases DOF of problem, because it describes problems in the space of performance characteristics. Finally, we will apply the proposal approach to a BIW model of real Honda vehicle and show its validity and effective ness comparing with current design with canonical optimization approach.

View full abstract

In this paper, a new type of optimization problem solving systems based on the Open Grid Services Architecture (OGSA) is proposed. The OGSA can define the Grid service aligned with Web service technologies. In the proposed system, the service data and notification function are used to integrate services. An end-user who wants to construct an optimization system decides how to integrate the services for optimization and analysis. The proposed system also uses the Grid Security Infrastracture (GSI), Index Service and Reliable File Transfer (RFT) Service to improve reliability and convenience. Therefore, the proposed system is expected to be an efficient optimization problem solving system on the Grid.

View full abstract

A huge amount of calculation is needed for optimizing design of actual structures. Various studies on parallel and distributed calculation have been performed in relation to a collection of computing resources distributed on a local area network. Recently researches on a highly efficient wide area computing system called the "Grid" have appeared. In order to realize design optimization of artifacts on the Grid, the following requirements have to be satisfied : (1) Computation with little dependency can be performed asynchronously, and (2) An optimal solution for required accuracy can be obtained robustly and promptly. The present study describes the development of a robust and efficient system for optimization of structural design of artifacts on the Grid. To develop the system, both the requirements are considered simultaneously. To satisfy the requirement (1), evaluation calculations such as finite element structural analyses are asynchronously performed on the Grid. Each FE analysis is solved using an Open Source parallel finite element system named ADVENTURE, developed by the authors. To satisfy the requirement (2), two kinds of efficient optimization algorithms based on real-coded genetic algorithms are developed. Both methods devise generation of new sample points. The first algorithm employs approximate quadratic programming problem, while the second one employs smoothing spline. In addition to these algorithmic works, robust processing mechanism is implemented to realize a whole computation on the Grid without interrupting. In the study, two kinds of Grid environments are employed. The one is simply a group of PC clusters distributed in remote sites. The other is one of Japanese national Grid projects named ITBL (IT-based Laboratory). Practical performances of the present system are demonstrated through some examples.

View full abstract

This paper proposes symbolic-numerical hybrid computing approach for evaluating derivates accurately. The derivatives are necessary (or at least very useful) in the computational solution of nonlinear problems. There are two approaches for evaluating derivatives in conventional scientific computation: analytically obtained by symbolic computing with symbolic algebra packages; or approximated by finite difference in numerical computing. The symbolic computing is surely reasonable in logicality and mathematics but is impossible in practice for large size problems because of asymptotically exponential or exponential computing complexity. On the other hand, the error analysis shows that finite difference approach only can provide limited accuracy of derivatives. Our symbolic-numerical hybrid computing approach establishes a bridge between symbolic computing and numerical computing, and achieves a good balance between computing complexity and accuracy for evaluating derivatives. The results of simulating computing have illustrated that the hybrid computing approach proposed in this paper is valid.

View full abstract

This paper proposes a new scheme for saving the computation cost of robust optimal design by using cumulative function approximation. Since robust optimal design seeks the optimal solution considering the variation of functions over distribution region, it requires much more computation cost than nominal optimal design because evaluation of each tentative solution requires iteration of system analysis at a certain number of reference points within a distribution region. The proposed scheme dramatically saves such computational cost by substituting system analysis at respective reference points with a cumulative function approximation that has an ability of gradually improving fidelity of approximation through gradual, adaptive and cumulative addition of sample points. In this paper, the outline of the scheme is discussed, the procedures of mini-max type robust optimal design and Voronoi diagram based cumulative function approximation are reviewed, and their hybridization for the purpose is described. Finally, numerical examples are demonstrated to ascertain its effectiveness.

View full abstract

CAE technology is indispensable to present design & development in every industries. But because it is the main purpose that the CAE tools are used for grasp the phenomenon, it owes a good deal to expert engineers that in order to do the better design. In this context, optimization tools have started to get a lot of publicity as the ways to improve the situations. The optimization tools are used for not only finding the optimal solutions but also improve the system of design & development of its own through the functions in optimization tool; Sampling method, Multi-objective robust design optimization, Response surface method and Multi criteria decision making. This report shows these technical functions through case studies.

View full abstract

As for the automotive engine cooling system of recent years, the various supplemental devices (EGR Cooler, ATF warmer and etc) are added, is a tendency of higher systematization. This is because decrease of the emission, improvement of fuel economy efficiency and improvement of reliability are made to actualize. This paper is introduced about the design approach which an optimization technique was used for calibration problem about automotive thermal management system.

View full abstract

Until now, the analysis method and the optimization method, which were used in design, were evolving independently. However, when designed combining them, coexistence of time and precision was difficult. Then, we aimed at constructing the high-speed first order design method by the modularizing model and integration of first order analysis and optimization method based on boundary conditions between modules as the dynamic behavior of the design object. It designed using the thermal layout problem (all 16 parameters, design space about 6*10^<24>) of four device modules as a virtual design object. As a result, for 600s, the design time could search about 2*10^7 counts of evaluation for the proper solution, and enabled the very high-speed first order design.

View full abstract

Use for the paper transport simulation has begun with a purpose of improvement in design efficiency and performance such as a copying machine and a printer. And, the technology of the optimization design using optimization toool is being put to practical use with the improvement in the design quality and robust design. So, we tried the optimization design of the stacker form for the copying machine using the paper transport simulation and the optimization toool. Approach by quality engineering and approach by the mathematical optimization technique were compared.

View full abstract

This paper describes an optimization design process of the rolling disc displacement of Trocoid tooth reducers. The reducers with plural row disc will have larger transmitted angle error than the reducers with single row disc. We analyze the mechanism that the reducer of the plural row disc causes larger angle error, and derive an optimal displacement angle of the plural row disc to minimize angle error. Simulation study and experimental trial show the strong effectiveness of this design process.

View full abstract