Multi-objective hierarchical genetic algorithms for multilevel redundancy allocation optimization

Multilevel redundancy allocation optimization problems (MRAOPs) occur frequently when attempting to maximize the system reliability of a hierarchical system, and almost all complex engineering systems are hierarchical. Despite their practical significance, limited research has been done concerning the solving of simple MRAOPs. These problems are not only NP hard but also involve hierarchical design variables. Genetic algorithms (GAs) have been applied in solving MRAOPs, since they are computationally efficient in solving such problems, unlike exact methods, but their applications has been confined to single-objective formulation of MRAOPs. This paper proposes a multi-objective formulation of MRAOPs and a methodology for solving such problems. In this methodology, a hierarchical GA framework for multi-objective optimization is proposed by introducing hierarchical genotype encoding for design variables. In addition, we implement the proposed approach by integrating the hierarchical genotype encoding scheme with two popular multi-objective genetic algorithms (MOGAs)—the strength Pareto evolutionary genetic algorithm (SPEA2) and the non-dominated sorting genetic algorithm (NSGA-II). In the provided numerical examples, the proposed multi-objective hierarchical approach is applied to solve two hierarchical MRAOPs, a 4- and a 3-level problems. The proposed method is compared with a single-objective optimization method that uses a hierarchical genetic algorithm (HGA), also applied to solve the 3- and 4-level problems. The results show that a multi-objective hierarchical GA (MOHGA) that includes elitism and mechanism for diversity preserving performed better than a single-objective GA that only uses elitism, when solving large-scale MRAOPs. Additionally, the experimental results show that the proposed method with NSGA-II outperformed the proposed method with SPEA2 in finding useful Pareto optimal solution sets.     

Alternatives and challenges in optimizing industrial safety using genetic algorithms

Safety (S) improvement of industrial installations leans on the optimal allocation of designs that use more reliable equipment and testing and maintenance activities to assure a high level of reliability, availability and maintainability (RAM) for their safety-related systems. However, this also requires assigning a certain amount of resources (C) that are usually limited. Therefore, the decision-maker in this context faces in general a multiple-objective optimization problem (MOP) based on RAMS+C criteria where the parameters of design, testing and maintenance act as decision variables. Solutions to the MOP can be obtained by solving the problem directly, or by transforming it into several single-objective problems. A general framework for such MOP based on RAMS+C criteria is proposed in this paper. Then, problem formulation and fundamentals of two major groups of resolution alternatives are presented. Next, both alternatives are implemented in this paper using genetic algorithms (GAs), named single-objective GA and multi-objective GA, respectively, which are then used in the case of application to solve the problem of testing and maintenance optimization based on unavailability and cost criteria. The results show the capabilities and limitations of both approaches. Based on them, future challenges are identified in this field and guidelines provided for further research.     

Optimization of multi-product batch plant design under uncertainty with environmental considerations

This paper presents a framework for the design and optimization of multi-product batch processes under uncertainty with environmental considerations. The uncertainties and environmental impacts are discussed. The profit and environmental impacts are considered as bi-objectives for batch plant design. The problem, thus, is formulated as a multi-objective stochastic programming problem. It can be converted into a single-objective two-stage stochastic linear programming problem using the weighted aggregation method. To solve the two-stage stochastic programming, we introduce both Monte Carlo sampling for the entire domain of the distribution function and the feasibility cut method based on dual theory in Benders’ decomposition. The detailed algorithm for problem-solving is presented. A numerical example is presented to illustrate the proposed framework.     

Multi-objective integrated acyclic crew rostering and vehicle assignment problem in public bus transportation

In this study, we try to solve a real planning problem faced in public bus transportation. It is a multi-objective integrated crew rostering and vehicle assignment problem. We model this problem as a multi-objective set partitioning problem. Most of the time, crew rostering problem with a single-objective function is considered, and the output may not satisfy some transport companies. To minimize the cost and maximize the fairness of the workload among the drivers, we define many criteria. Although crew rostering problem and its integrated versions appear in the literature, it is the first time these two problems are integrated. We propose a new multi-objective tabu search algorithm to obtain near Pareto-optimal solutions. The algorithm works with a set of solutions using parallel search. We test our algorithm for the case with ten objectives and define a method to choose solutions from the approximated efficient frontier to present to the user. We discuss the performance of our meta-heuristic approach.

     

Multi-objective optimization of space station logistics strategies using physical programming

This study extends a previously proposed single-objective optimization formulation of space station logistics strategies to multi-objective optimization. The four-objective model seeks to maximize the mean utilization capacity index, total utilization capacity index, logistics robustness index and flight independency index, aiming to improve both the utilization benefit and the operational robustness of a space station operational scenario. Physical programming is employed to convert the four-objective optimization problem into a single-objective problem. A genetic algorithm is proposed to solve the resulting physical programming-based optimization problem. Moreover, the non-dominated sorting genetic algorithm-II is tested to obtain the Pareto-optimal solution set and verify the Pareto optimality of the physical programming-based solution. The proposed approach is demonstrated with a notional one-year scenario of China's future space station. It is shown that the designer-preferred compromise solution improving both the utilization benefit and the operational robustness is successfully obtained.     

Optimisation of a multi-objective two-dimensional strip packing problem based on evolutionary algorithms

This paper considers a real-world two-dimensional strip packing problem involving specific machinery constraints and actual cutting production industry requirements. To adapt the problem to a wider range of machinery characteristics, the design objective considers the minimisation of material length and the total number of cuts for guillotinable-type patterns. The number of cuts required for the cutting process is crucial for the life of the industrial machines and is an important aspect in determining the cost and efficiency of the cutting operation. In this paper we propose the application of evolutionary algorithms to address the multi-objective problem, for which numerous approaches to its single-objective formulation exist, but for which multi-objective approaches are almost non-existent. The multi-objective evolutionary algorithms applied provide a set of solutions offering a range of trade-offs between the two objectives from which clients can choose according to their needs. By considering both the length and number of cuts, they derive solutions with wastage levels similar to most previous approximations which just seek to optimise the overall length.     

Dual‐based adaptive FEM for inelastic problems with standard FE implementations

Goal‐oriented error estimation allows to refine meshes in space and time with respect to arbitrary quantities. The required dual problems that need to be solved usually require weak formulations and the Galerkin method in space and time to be established. Unfortunately, this does not obviously leads to structures of standard finite element implementations for solid mechanics. These are characterized by a combination of variables at nodes (e.g. displacements) and at integration points (e.g. internal variables) and are solved with a two‐level Newton method because of local uncoupled and global coupled equations. Therefore, we propose an approach to approximate the dual problem while maintaining these structures. The primal and the dual problems are derived from a multifield formulation. Discretization in time and space with appropriate shape functions and rearrangement yields the desired result. Details on practical implementation as well as applications to elasto‐plasticity are given. Numerical examples demonstrate the effectiveness of the procedure. Copyright © 2015 John Wiley & Sons, Ltd.     

SSPH basis functions for meshless methods, and comparison of solutions with strong and weak formulations

We propose a new and simple technique called the Symmetric Smoothed Particle Hydrodynamics (SSPH) method to construct basis functions for meshless methods that use only locations of particles. These basis functions are found to be similar to those in the Finite Element Method (FEM) except that the basis for the derivatives of a function need not be obtained by differentiating those for the function. Of course, the basis for the derivatives of a function can be obtained by differentiating the basis for the function as in the FEM and meshless methods. These basis functions are used to numerically solve two plane stress/strain elasto-static problems by using either the collocation method or a weak formulation of the problem defined over a subregion of the region occupied by the body; the latter is usually called the Meshless Local Petrov–Galerkin (MLPG) method. For the two boundary-value problems studied, it is found that the weak formulation in which the basis for the first order derivatives of the trial solution are derived directly in the SSPH method (i.e., not obtained by differentiating the basis function for the trial solution) gives the least value of the L²-error norm in the displacements while the collocation method employing the strong formulation of the boundary-value problem has the largest value of the L²-error norm. The numerical solution using the weak formulation requires more CPU time than the solution with the strong formulation of the problem. We have also computed the L²-error norm of displacements by varying the number of particles, the number of Gauss points used to numerically evaluate domain integrals appearing in the weak formulation of the problem, the radius of the compact support of the kernel function used to generate the SSPH basis, the order of complete monomials employed for constructing the SSPH basis, and boundary conditions used at a point on a corner of the rectangular problem domain. It is recommended that for solving two-dimensional elasto-static problems, the MLPG formulation in which derivatives of the trial solution are found without differentiating the SSPH basis function be adopted.     

Second‐order optimal control algorithm for complex systems

The solution to large‐scale optimal control problems, characterized by complex dynamics and extended time periods, is often computationally demanding. We present a solution algorithm with favourable local convergence properties as a way to reduce simulation times. This method is based on using a trapezoidal direct collocation to convert the differential equations into algebraic constraints. The resulting constrained minimization problem is then solved with an augmented Lagrangian formulation to accommodate both equality and inequality constraints. In contrast to the prevalent optimal control software implementations, we calculate analytical first and second derivatives. We then apply a generalized Newton method to the augmented Lagrangian formulation, solving for all unknowns simultaneously. The computational costs of the Hessian formation and matrix solution remain manageable as the system size increases due to the sparsity of all tensor quantities. Likewise, the total iterations for convergence scale well due to the local quadratic convergence of the generalized Newton method. We demonstrate the method with an inverted pendulum problem and a neuromuscular control problem with complex dynamics and 18 forcing functions. The optimal control solutions are successfully found. In both examples, we obtain quadratic convergence rates in the neighbourhood of the solution. Copyright © 2002 John Wiley & Sons, Ltd.     

Frictional mortar contact for finite deformation problems with synthetic contact kinematics

In this paper we present a mortar based method, for frictional two dimensional contact problems. It is based on the work by Tur et al. (Comput Methods Appl Mech Eng 198(37–40):2860–2873, 2009) and uses the same concentrated integration scheme as well as a non regularized tangential contact formulation based on Lagrange multipliers only. We abstract the contact kinematics to a rather synthetic formulation. Therefore we are able to use two different methods of defining the normal field on the discretized surface normal: The popular method of averaged non-mortar side normal and the rather simple non continuous mortar side normal field. The problem is solved with a fixed point Newton–Raphson procedure and for both normal fields the full linearizations are derived. With numerical examples we show the performance of the more concise formulation of the non averaged non continuous mortar side normal field.     

A comparison of mixed integer programming formulations of the capacitated lot-sizing problem

We propose a novel mixed integer programming formulation for the capacitated lot-sizing problem with set-up times and set-up carryover. We compare our formulation to two earlier formulations, the Classical and Modified formulations, and a more recent formulation due to Suerie and Stadtler. Extensive computational experiments show that our formulation consistently outperforms the Classical and Modified formulations in terms of CPU time and solution quality. It is competitive with the Suerie–Stadtler (S&S) formulation, but outperforms all other formulations on the most challenging instances, those with low-capacity slack and a dense jobs matrix. We show that some of the differences in the performance of these various formulations arise from their different use of binary variables to represent production or set-up states. We also show that the LP relaxation of our Novel formulation provides a tighter lower bound than that of the Modified formulation. Our experiments demonstrate that, while the S&S formulation provides a much tighter LP bound, the Novel formulation is better able to exploit the intelligence of the CPLEX solution engine.     

Integrating fleet replacement and pilot training schedules for airborne service helicopters

In this study, we examine an integrated problem involving fleet replacement and pilot training scheduling for airborne service helicopters. Two objectives, minimizing the total cost and the number of batches of pilot training, are considered based on the practical requirements. Multi-objective programming combined with a time–space network technique is applied to solve the problem. Numerical tests, in reference to the operations of a Taiwanese organization, are carried out to demonstrate the performance of the proposed model. A comparison of the proposed model and two single-objective models corresponding to the two objectives is also presented.     

Multi-objective due-date setting in a make-to-order environment

This paper presents a new dual-objective problem of due-date setting over a rolling planning horizon in make-to-order manufacturing and proposes a bi-criterion integer programming formulation for its solution. In the proposed model the due-date setting decisions are directly linked with available capacity. A simple critical load index is introduced to quickly identify the system bottleneck and the overloaded periods. The problem objective is to select a maximal subset of orders that can be completed by the customer requested dates and to quote delayed due dates for the remaining acceptable orders to minimise the number of delayed orders or the total number of delayed products as a primary optimality criterion and to minimise total or maximum delay of orders as a secondary criterion. A weighted-sum program based on a scalarisation approach is compared with a two-level due-date setting formulation based on the lexicographic approach. In addition, a mixed-integer programming model is provided for scheduling customer orders over a rolling planning horizon to minimise the maximum inventory level. Numerical examples modeled after a real-world make-to-order flexible flowshop environment in the electronics industry are provided and, for comparison, the single-objective solutions that maximise total revenue subject to service level constraints are reported.     

A meshfree method: meshfree weak–strong (MWS) form method,for 2-D solids

A novel meshfree weak–strong (MWS) form method is proposed based on a combined formulation of both the strong-form and the local weak-form. In the MWS method, the problem domain and its boundary is represented by a set of distributed points or nodes. The strong form or the collocation method is used for all nodes whose local quadrature domains do not intersect with natural (Neumann) boundaries. Therefore, no numerical integration is required for these nodes. The local weak-form, which needs the local numerical integration, is only used for nodes on or near the natural boundaries. The locally supported radial point interpolation method and the moving least squares approximation are used to construct the meshfree shape functions. The final system matrix will be sparse and banded for computational efficiency. Numerical examples of two-dimensional solids are presented to demonstrate the efficiency, stability, accuracy and convergence of the proposed meshfree method.     

A new fast multipole boundary element method for solving large‐scale two‐dimensional elastostatic problems

A new fast multipole boundary element method (BEM) is presented in this paper for large‐scale analysis of two‐dimensional (2‐D) elastostatic problems based on the direct boundary integral equation (BIE) formulation. In this new formulation, the fundamental solution for 2‐D elasticity is written in a complex form using the two complex potential functions in 2‐D elasticity. In this way, the multipole and local expansions for 2‐D elasticity BIE are directly linked to those for 2‐D potential problems. Furthermore, their translations (moment to moment, moment to local, and local to local) turn out to be exactly the same as those in the 2‐D potential case. This formulation is thus very compact and more efficient than other fast multipole approaches for 2‐D elastostatic problems using Taylor series expansions of the fundamental solution in its original form. Several numerical examples are presented to study the accuracy and efficiency of the developed fast multipole BEM formulation and code. BEM models with more than one million equations have been solved successfully on a laptop computer. These results clearly demonstrate the potential of the developed fast multipole BEM for solving large‐scale 2‐D elastostatic problems. Copyright © 2005 John Wiley & Sons, Ltd.     

Fracture analysis of cracks in magneto-electro-elastic solids by the MLPG

A meshless method based on the local Petrov–Galerkin approach is proposed for crack analysis in two-dimensional (2-D) and three-dimensional (3-D) axisymmetric magneto-electric-elastic solids with continuously varying material properties. Axial symmetry of geometry and boundary conditions reduces the original 3-D boundary value problem into a 2-D problem in axial cross section. Stationary and transient dynamic problems are considered in this paper. The local weak formulation is employed on circular subdomains where surrounding nodes randomly spread over the analyzed domain. The test functions are taken as unit step functions in derivation of the local integral equations (LIEs). The moving least-squares (MLS) method is adopted for the approximation of the physical quantities in the LIEs. The accuracy of the present method for computing the stress intensity factors (SIF), electrical displacement intensity factors (EDIF) and magnetic induction intensity factors (MIIF) are discussed by comparison with numerical solutions for homogeneous materials.     

Local search for nonpreemptive multi-mode resource-constrained project scheduling

This paper addresses a general class of nonpreemptive resource-constrained project scheduling problems in which activity durations are discrete functions of committed renewable and nonrenewable resources. We provide a well known 0-1 problem formulation and stress the importance of the model by giving applications within production and operations management. Furthermore, we prove that the feasibility problem is already NP-complete. Solution procedures proposed so far have the following shortcomings: exact methods can solve only very small instances to optimality; heuristic solution approaches fail to generate feasible solutions when problems become highly resource-constrained. Hence, we propose a new local search method that first tries to find a feasible solution and secondly performs a single-neighborhood search on the set of feasible mode assignments. To evaluate the new procedure we perform a rigorous computational study on two benchmark sets. The experiment includes a comparison of our procedure with other heuristics.     

New approaches for error estimation and adaptivity for 2D potential boundary element methods

This work presents two new error estimation approaches for the BEM applied to 2D potential problems. The first approach involves a local error estimator based on a gradient recovery procedure in which the error function is generated from differences between smoothed and non‐smoothed rates of change of boundary variables in the local tangential direction. The second approach involves the external problem formulation and gives both local and global measures of error, depending on a choice of the external evaluation point. These approaches are post‐processing procedures. Both estimators show consistency with mesh refinement and give similar qualitative results. The error estimator using the gradient recovery approach is more general, as this formulation does not rely on an ‘optimal’ choice of an external parameter. This work presents also the use of a local error estimator in an adaptive mesh refinement procedure. This r‐refinement approach is based on the minimization of the standard deviation of the local error estimate. A non‐linear programming procedure using a feasible‐point method is employed using Lagrange multipliers and a set of active constraints. The optimization procedure produces finer meshes close to a singularity and results that are consistent with the problem physics. Copyright © 2002 John Wiley & Sons, Ltd.     

Availability allocation to repairable systems with genetic algorithms: a multi-objective formulation

This paper describes a methodology based on genetic algorithms (GA) and experiments plan to optimize the availability and the cost of reparable parallel-series systems. It is a NP-hard problem of multi-objective combinatorial optimization, modeled with continuous and discrete variables. By using the weighting technique, the problem is transformed into a single-objective optimization problem whose constraints are then relaxed by the exterior penalty technique. We then propose a search of solution through GA, whose parameters are adjusted using experiments plan technique. A numerical example is used to assess the method.     

A comparison of outlet boundary treatments for prevention of backflow divergence with relevance to blood flow simulations

Simulation divergence due to backflow is a common, but not fully addressed, problem in three-dimensional simulations of blood flow in the large vessels. Because backflow is a naturally occurring physiologic phenomenon, careful treatment is necessary to realistically model backflow without artificially altering the local flow dynamics. In this study, we quantitatively compare three available methods for treatment of outlets to prevent backflow divergence in finite element Navier–Stokes solvers. The methods examined are (1) adding a stabilization term to the boundary nodes formulation, (2) constraining the velocity to be normal to the outlet, and (3) using Lagrange multipliers to constrain the velocity profile at all or some of the outlets. A modification to the stabilization method is also discussed. Three model problems, a short and long cylinder with an expansion, a right-angle bend, and a patient-specific aorta model, are used to evaluate and quantitatively compare these methods. Detailed comparisons are made to evaluate robustness, stability characteristics, impact on local and global flow physics, computational cost, implementation effort, and ease-of-use. The results show that the stabilization method offers a promising alternative to previous methods, with reduced effect on both local and global hemodynamics, improved stability, little-to-no increase in computational cost, and elimination of the need for tunable parameters.