Multiobjective concurrent engineering: a goal programming approach

Concurrent engineering (CE) is a commonly used engineering/management technique for product development. Multiple and sometimes conflicting objectives can be experienced when using CE as an aid in new product development. One management science technique that commonly used to analyze multiple objectives in a decision making environment is goal programming (GP). In this paper, the authors demonstrate how goal programming can be used to enhance CE and improve decision making by considering conflicting objective criteria of cost and time. The procedure by which CE problems can be modeled using GP is presented along with an illustrative example from the literature. Results from a real-world ongoing study are also reported     

Augmenting Shewhart quality control and Delphi method for multi-person,multi-objective decision-making

The aim of this paper is to augment the Delphi method with Shewhart’s quality control technique to propose a systematic and flexible methodology for group decision making in the presence of multiple objectives. To illustrate the application of the proposed model, goal programming (GP) is used. GP is perhaps one of the most powerful and most widely used multiple criteria decision making (MCDM) techniques. However, major drawbacks of GP are the inherent subjectivity introduced into problem formulation by requiring the decision-maker to specify the objectives and the trade-off among them as well as a target level for each objective. The methodology proposed in this paper attempts to reduce/eliminate this subjectivity.     

Integrated Genetic Algorithm and Goal Programming for Network Topology Design Problem With Multiple Objectives and Multiple Criteria

Network topology design (NTD) with multiple objectives has been presented by many researchers. However, no work in the literature has addressed this issue with both multiple objectives and multiple criteria. In order to suit real-world situations, this paper presents a new idea integrating genetic algorithm and goal programming to establish a model for solving the NTD problem with multiple objectives and multiple criteria taken into consideration. In addition, the proposed model can also solve both construct and extend network topology problems under shared risk link group (SRLG) constraints. Finally, illustrative examples are included to demonstrate the superiority and usefulness of the proposed method.     

An integrated multicriteria model for warranty cost estimation andproduction

Warranty cost estimation for multiple products is considered. The decision variables include the price, warranty time, production quantity, and lot size. System constraints on the above decision variables are based on absolute minimum and maximum values between which those variables should lie. Several goals are considered at different levels of priorities. Some of these goals may be conflicting. The goals considered included operation within a limited resource capacity, the achievement of a specified market share for each product, limitation on the total warranty cost as a given proportion of total sales, limitation of the warranty reserve for a given product as a proportion of the total warranty reserve for all products, and attainment of a minimum level of warranty reserve for a given product as a proportion of the total warranty reserve cost. A goal programming approach is used to solve the formulated problem. An example is illustrated using the proposed formulation, and goal achievements are discussed. Sensitivity analysis is also conducted for some of the model parameters     

Multi‐objective optimization for channel allocation in mobile computing using NSGA‐II

Efficient channel allocation to mobile hosts aims to minimize the number of blocked hosts and is of utmost importance in a mobile computing network. Also, to achieve highly reliable data transmission, wireless mobile networks require efficient and reliable link connectivity regardless of terminal mobility, and thus reliable traffic performance. A mobile network consists of mobile nodes, base stations, links, etc. that are often prone to failure. The multi‐objective optimization problem (MOP), does not offer one best solution with respect to all the objectives. The aim is to determine the trade‐off surface, which is a set of non‐dominated solution points known as Pareto‐optimal. The two objectives addressed in this paper are to minimize the number of blocked hosts while maximizing the reliability of the data transmission. A multi‐objective optimization is carried out to optimize both objectives simultaneously. The elitist NSGA‐II (non‐dominated sorting genetic algorithm) has been used as an evolutionary optimization technique to solve this problem. A population of efficient solutions results when the termination condition is satisfied. Also the Pareto‐optimal fronts obtained provide a wide range of trade‐off operating conditions from which an appropriate operating point may be selected by the decision maker. The experimental results are presented and analyzed for overall evaluation of the performance of the proposed work. Further, comparison of the results with the two recent earlier models reveals that the proposed work performs better in serving mobile hosts as it caters to two objectives simultaneously. Copyright © 2010 John Wiley & Sons, Ltd.     

Multiobjective water resources investment planning under budgetaryand socio-technical uncertainties

An integration of stochastic dynamic programming (SDP) and integer goal programming (IGP) modeling framework is proposed to handle problems of multiobjective-multicriteria sequential decision making under budgetary and socio-technical uncertainties inherent in water resources investment planning. In the proposed SDP model, probabilities of the funding levels in any time period that are generated using a subjective model are employed to handle budgetary fluctuations. This subjective model consists of historical data as a basic rate, functional relationships among inter-related parameters of the SDP model, scenarios of future budget availability, and subjective inputs elicited from a group of decision makers through a collective opinion technique. Application of the SDP model primarily yields an optimal investment planning policy that recognizes the possibility that actual funding received maybe less than that anticipated, and therefore the projects being implemented under the anticipated budget would be interrupted. In this integrated model, economic return of each level of investment decision together with its associated project portfolio is determined by the IGP model based on goals and criteria preferences with explicit consideration to socio-technical uncertainty     

Conditional stochastic dominance in R&D portfolio selection

This paper describes a methodology for the selection of research and development (R&D) projects to add to or remove from an existing R&D portfolio. The analysis uses the criterion of conditional stochastic dominance to make selection recommendations. This criterion takes into account the effect of a given project on the risk and return of the existing portfolio. The authors use a methodology previously employed to analyze stock portfolios; however, they apply it using simulation in an R&D portfolio context. They apply the methodology to the portfolios of two actual companies and find that it generates priorities very close to those developed by internal company heuristics. They conclude that this methodology can be applied appropriately in these circumstances and that its recommendations are consistent with observed decision maker behavior. Their results suggest that an R&D manager should not consider project selection decisions in isolation, but, following this methodology, should take into account the context of the existing portfolio     

Markov decision process (MDP) framework for software power optimization using call profiles on mobile phones

We present an optimization framework for delay-tolerant data applications on mobile phones based on the Markov decision process (MDP). This process maximizes an application specific reward or utility metric, specified by the user, while still meeting a talk-time constraint, under limited resources such as battery life. This approach is novel for two reasons. First, it is user profile driven, which means that the user’s history is an input to help predict and reserve resources for future talk-time. It is also dynamic: an application will adapt its behavior to current phone conditions such as battery level or time before the next recharge period. We propose efficient techniques to solve the optimization problem based on dynamic programming and illustrate how it can be used to optimize realistic applications. We also present a heuristic based on the MDP framework that performs well and is highly scalable for multiple applications. This approach is demonstrated using two applications: Email and Twitter synchronization with different priorities. We present experimental results based on Google’s Android platform running on an Android Develepor Phone 1 (HTC Dream) mobile phone.     

Fuzzy analytic network process and its application to the development of decision support systems

In this paper we propose a fuzzy extension of the analytic network process (ANP) that uses uncertain human preferences as input information in the decision-making process. Instead of the classical Eigenvector prioritization method, employed in the prioritization stage of the ANP, a new fuzzy preference programming method, which obtains crisp priorities from inconsistent interval and fuzzy judgments is applied. The resulting fuzzy ANP enhances the potential of the ANP for dealing with imprecise and uncertain human comparison judgments. It allows for multiple representations of uncertain human preferences, as crisp, interval, and fuzzy judgments and can find a solution from incomplete sets of pairwise comparisons. An important feature of the proposed method is that it measures the inconsistency of the uncertain human preferences by an appropriate consistency index. A prototype decision support system realizing the proposed method is developed, and its performance is illustrated by examples.     

Incorporating Quality Considerations Into Project Time/Cost Tradeoff Analysis and Decision Making

Existing models and methods of project scheduling implicitly assume uniform quality when evaluating time/cost tradeoffs, but do not model quality explicitly. For a project manager such as a general contractor who subcontracts most tasks of a project, or other project managers who face decisions concerning the level of quality to perform for each task, this is not a realistic assumption. In this paper, we extend the standard discrete time/cost tradeoff problem by assuming that each option for each task is evaluated for its duration, cost, and also its quality. We then provide a mixed integer linear programming model for solving one version of this problem, and illustrate with an example. We next formulate a goal programming mixed integer linear program for a very general version of the problem. This formulation allows for many different definitions of quality, and many combinations of time, cost, and quality goals with preemptive and/or weighted priorities, and is also illustrated with an example. We show how these models can be used to generate quality level curves to illustrate the tradeoffs among time, cost, and quality. These level curves can then be used by project managers to make project scheduling decisions that explicitly model and consider quality as well as time and cost, so that better and more appropriate decisions can be made for a particular situation     

Design of transmit FIR filters for FDM data transmission systems

The design of transmit finite-impulse response (FIR) filters with few coefficients in frequency division multiplexing data transmission systems is considered. In these systems, quality objectives are imposed for transmission over channels affected by additive white Gaussian noise (AWGN) and over channels affected by AWGN plus adjacent channel interference (ACI). The goal of this letter is: given that an adaptive receive filter, possibly cooperating with an adaptive decision feedback equalizer, is used for the AWGN channel and for the AWGN plus ACI channel, what is the best fixed FIR transmit filter to use for both channel cases? This goal is achieved by optimizing the compromise between the performance of the system on the two mentioned channels. Also, the advantages of the proposed design over a rival method based on a fixed receive filter are demonstrated.     

Increasing the Robustness of Boosting Algorithms within the Linear-programming Framework

AdaBoost has been successfully used in many signal classification systems. However, it has been observed that on highly noisy data AdaBoost easily leads to overfitting, which seriously constrains its applicability. In this paper, we address this problem by proposing a new regularized boosting algorithm LP_norm2-AdaBoost (LPNA). This algorithm arises from a close connection between AdaBoost and linear programming. In the algorithm, skewness of the data distribution is controlled during the training to prevent outliers from spoiling decision boundaries. To this end, a smooth convex penalty function (l ₂ norm) is introduced in the objective function of a minimax problem. A stabilized column generation technique is used to transform the optimization problem into a simple linear programming problem. The effectiveness of the proposed algorithm is demonstrated through experiments on many diverse datasets.     

Optimized bandwidth allocation with fairness and service differentiation in multimedia wireless networks

In this article we present an optimal Markov Decision‐based Call Admission Control (MD‐CAC) policy for the multimedia services that characterize the next generation of wireless cellular networks. A Markov decision process (MDP) is used to represent the CAC policy. The MD‐CAC is formulated as a linear programming problem with the objectives of maximizing the system utilization while ensuring class differentiation and providing quantitative fairness guarantees among different classes of users. Through simulation, we show that the MD‐CAC policy potentially achieves the optimal decisions. Hence our proposed MD‐CAC policy satisfies its design goals in terms of call‐class‐differentiation, fairness and system utilization. Copyright © 2006 John Wiley & Sons, Ltd.     

Optimal scheduling of handoffs in cellular networks

The phenomenon of hard handoffs (as applicable to FDMA- and TDMA-based networks) as well as soft handoffs (as applicable to DS/CDMA-based networks) is formulated as stochastic optimization problems. The signals received by a mobile user are treated as stochastic processes with associated rewards, which are functions of some measurable characteristics of the received signals, while the handoff is associated with a switching penalty. This formulation captures the trade-offs involved in handoffs in a flexible manner and captures many facets of popular cellular communication systems in use currently. Using dynamic programming, necessary and sufficient conditions for determining the optimal base station(s) the mobile should be associated with during each decision epoch are derived. For the cases where the above-mentioned necessary and sufficient conditions fail to determine an optimal decision, “limited lookahead” arguments are used for determining handoff decisions. The decisions are taken in a decentralized manner, which makes its implementation easier compared to centralized algorithms. Simulation results show that for the hard handoffs, performance gain by the proposed algorithm over the simpler threshold algorithms proposed in the literature is small; however, for the case of soft handoffs, the proposed algorithm offers considerable improvement over the algorithm proposed in the IS-95 standard     

基于层次分析模型的产品多属性综合排序

为了解决产品的多属性综合排序问题,提出了一种基于层次分析模型的建模方法。通过将决策问题按目标层、准则层和方案层等分解为不同的层次结构,然后求得每一层次各元素对上一层次某元素的优先权重,再用加权和的方法归并各参评方案对总目标的最终权重,从而完成产品的综合排序任务。实验结果证明所提方法在产品综合排序任务上的有效性。     

Multiobjective programming using uniform design and genetic algorithm

The notion of Pareto-optimality is one of the major approaches to multiobjective programming. While it is desirable to find more Pareto-optimal solutions, it is also desirable to find the ones scattered uniformly over the Pareto frontier in order to provide a variety of compromise solutions to the decision maker. We design a genetic algorithm for this purpose. We compose multiple fitness functions to guide the search, where each fitness function is equal to a weighted sum of the normalized objective functions and we apply an experimental design method called uniform design to select the weights. As a result, the search directions guided by these fitness functions are scattered uniformly toward the Pareto frontier in the objective space. With multiple fitness functions, we design a selection scheme to maintain a good and diverse population. In addition, we apply the uniform design to generate a good initial population and design a new crossover operator for searching the Pareto-optimal solutions. The numerical results demonstrate that the proposed algorithm can find the Pareto-optimal solutions scattered uniformly over the Pareto frontier.     

Bio-inspired Patrolling Scheme Design in Wireless and Mobile Sensor and Robot Networks

Environment monitoring is an essential application in wireless sensor networks. We consider a large area surveillance in which a monitored region is divided into small sub areas where sensors, sinks, and mobile robots are responsible for surveillance jobs. Sensors collect data from the environment for specific monitoring purposes. Sinks are usually stationary and responsible for collecting regional information, while mobile robots are always patrolling the monitored areas to handle events. Under the above scenario, we propose anti-inspired digital pheromone to classify priorities of events used in a proposed patrolling algorithm, in which, mobile robots are patrolling and handling events with priorities along the roads. Events with higher priorities are handled with higher priorities, while productive event handling, i.e., handling more events, is also the goal. Extensive simulations are conducted to study the efficiency of the proposed method.

     

The linear multi-objective R&D project selection problem

A multi-objective model of the project-selection problem is described. The model departs from an earlier goal-programming formulation of the problem, which suggested Delphic methods for selection of priorities and aspiration levels. It is shown that the multiobjective formulation yields multiple nondominated solutions for the same problem solved by goal programming, whereas the goal-programming formulation revealed only one solution, and that the goal-programming solution is sensitive to choice of aspiration levels. The multiobjective model is recommended as a more general approach to the research and development project-selection problem, since it will develop a set of all nondominated solution. Subjective methods (such as the Delphic technique) can then be called for at a later point in the analysis to choose among alternative nondominated solutions     

Array pattern synthesis using semidefinite programming and a bisection method

In this paper, we propose an array pattern synthesis scheme using semidefinite programming (SDP) under array excitation power constraints. When an array pattern synthesis problem is formulated as an SDP problem, it is known that an additional rank‐one constraint is generated inevitably and relaxed via semidefinite relaxation. If the solution to the relaxed SDP problem is not of rank one, then conventional SDP‐based array pattern synthesis approaches fail to obtain optimal solutions because the additional rank‐one constraint is not handled appropriately. To overcome this drawback, we adopted a bisection technique combined with a penalty function method. Numerical applications are presented to demonstrate the validity of the proposed scheme.