Adiel Teixeira de Almeida-Filho

Multicriteria and Multiobjective Models for Risk, Reliability and Maintenance Decision Analysis

Multiobjective and Multicriteria Problems and Decision Models.- Multiobjective and Multicriteria Decision Processes and Methods.- Basic Concepts on Risk Analysis, Reliability and Maintenance.- Multidimensional Risk Analysis.- Preventive Maintenance Decisions.- Decision Making in Condition-Based Maintenance.- Decision on Maintenance Outsourcing.- Spare Parts Planning Decisions.- Decision on Redundancy Allocation.- Design Selection Decisions.- Decisions on Priority Assignment for Maintenance Planning.- Other Risk, Reliability and Maintenance Decision Problems.

A new method for elicitation of criteria weights in additive models: Flexible and interactive tradeoff

This paper proposes the Flexible and Interactive Tradeoff (FITradeoff) method, for eliciting scaling constants or weights of criteria. The FITradeoff uses partial information about decision maker (DM) preferences to determine the most preferred in a specified set of alternatives, according to an additive model in MAVT (Multi-Attribute Value Theory) scope. This method uses the concept of flexible elicitation for improving the applicability of the traditional tradeoff elicitation procedure. FITradeoff offers two main benefits: the information required from the DM is reduced and the DM does not have to make adjustments for the indifference between two consequences (trade-off), which is a critical issue on the traditional tradeoff procedure. It is easier for the DM to make comparisons of consequences (or outcomes) based on strict preference rather than on indifference. The method is built into a decision support system and applied to two cases on supplier selection, already published in the literature.

A multicriteria approach for analysis of conflicts in evidence theory

We propose a new approach to model the conflict on DST as multi-criteria problem.We introduce a new subjective notion of conflict through a pseudo-criteria analysis.We present a numerical application of the proposed model based on 2 and 3 conflict measures.Therefore, it is showed that is possible to capture conflict in DST using more that to 2 criteria. Dempster-Shafer Theory (DST) or Evidence Theory is regarded as one of the leading theories for modeling uncertainty in imprecise situations. The main advantage of this theory arises from the possibility of combining different bodies of evidence originally developed by using Dempsters combination rule. However, this rule leads to counter-intuitive results when the bodies of evidence conflict with each other to a high degree. Thus, different combinations of conflict management rules have been developed over the years where, regardless of the method used, what should be identified first of all is the level of conflict between the bodies of evidence. Therefore, different metrics were used to classify or quantify the conflict but no single one of these was successful because it is impracticable to represent all situations of conflict in this theory by using only one metric. Therefore, the contribution of this article is to analyze conflict within DST by using a multi-criteria analysis, for which the Multi Criteria Decision Making (MCDM) method considered was ELECTRE TRI. On modeling the problem, three classes of conflict (low, medium and high), were considered. To validate the model, a numerical analysis was conducted that included the use of a method to point up conflict so as to infer parameters.

A flexible elicitation procedure for additive model scale constants

This paper contributes to the process of eliciting additive model scale constants in order to support choice problems, thereby reducing the effort a decision maker (DM) needs to make since partial information with regard to DM preferences can be used. Procedures related to eliciting weights without a tradeoff interpretation of weights are justified based on assumptions that DM is not able to specify fixed weight values or if DM is able to do so, this would not be reliable information. As long as partial information is provided, the flexible elicitation procedure performs dominance tests based on a linear programming problem to explore the DM’s preferences as a vector space which is built using the DM’s partial information. To provide evidence of the satisfactory performance of the flexible elicitation procedure, an empirical test is presented with results that indicate that this procedure requires less effort from DMs.

A Multicriteria Approach Using MAUT to Assist the Maintenance of a Water Supply System Located in a Low-Income Community

Public services for low-income communities located in developing countries have shortcomings in their infrastructure, which very often result from the lack of capacity to supply water to all clients simultaneously. Because of this, in these regions maintenance activity becomes more important because besides the limited capacity, the losses of supply or availability due to system failures further increase the number of hours during which service to the whole community is interrupted. Given this problem, this article proposes a multi-criteria decision model based on MAUT for the planning of preventive maintenance in cases like this, and considers that besides costs, keeping the system available is an objective. The contribution of this paper is to present a model for a real and common problem in developing countries, besides the results found through the case study. By applying the model proposed in a case study in a low-income community in Brazil, it was possible to identify a maintenance policy that seeks to increase the efficiency of the system which can be achieved by making the increase in availability and minimizing cost compatible.

A DSS for Resolving Evaluation of Criteria by Interactive Flexible Elicitation Procedure

This paper describes the implementation of a Decision Support System (DSS) with a flexible elicitation procedure to support a decision maker (DM) to establish weights for a multi criteria choice decision problem, based on the concept of a flexible procedure. The DSS not only presents an elicitation interface to establish weights, but computes the performance of alternatives in an additive model, thus enabling the DM to reduce the set of non-dominated alternatives since partial information about preferences is given. When the flexible elicitation procedure implemented in this DSS is performed less information is required from the DM, which implies that less cognitive effort is required. The DSS contributes by providing a user friendly interface which enables the elicitation protocol to be performed using graphics and by computing alternatives performance quickly so that the elicitation process can be finished quickly with the minimum number of questions needed to solve the choice problem. The flexible elicitation procedure proposed is illustrated with a numerical application.

Allocation of Power-Quality Monitors Using the P-Median to Identify Nontechnical Losses

This study sets out to develop a procedure that enables monitors to be allocated that are able to estimate values of voltage and current of a circuit. Taking the importance of the load as a starting point, this paper puts forward a model for allocating power-quality monitors in distribution systems based on the P-median model. As a first step, by using an improved model that already exists, the fewest number of monitors which guarantee the observability of the system is obtained. Then, the modified P-median model is applied which incorporates a constraint in order to allocate monitors in accordance with the importance of the load, which is considered as the volume of the load to tackle nontechnical losses. A model such as this is required especially when dealing with nontechnical losses, which can be one of the main causes of undervoltage, thereby affecting the quality of the service. The proposed model was applied in a real distribution network considering the importance of the load throughout the distribution system.

A method for elicitation and combination of imprecise probabilities: A mathematical programming approach

In this paper, a method of elicitation and expert aggregation, supported by existing models in the literature, is developed. To avoid cognitive problems resulting from numerical judgment, comparative judgment of probability is used in the elicitation process, where the best way to represent the data is admitted to be by imprecise probability models. In this sense, a linear programming model is used to convert such judgment into probability intervals. For the aggregation of experts, a quadratic programming model is used, where the use of different metrics to represent the weights will be discussed.

A DSS Based on Multiple Criteria Decision Making for Maintenance Planning in an Electrical Power Distributor

This paper presents a real world application that lead to the construction of a decision support system built to support maintenance planning in an electrical power distribution company. The outcome of the DSS is to determine the order of priority in which potential failure repair orders should be conducted. This company-specific problem has been structured and is similar to that of most electrical power distribution companies, since this company has a complex distribution network with different types of customers which leads it to considering different criteria such as income losses, the probability of service interruptions, national regulation criteria, and so forth. There is a specific budget for potential failure repair orders, while functional failure repairs are not modeled at this managerial level since once the service should not be interrupted and involves long term planning. During periodical inspections the entire distribution network is inspected and a database with all this data is maintained. Thereafter, the potential failures in the distribution network are converted into maintenance orders which typically amount to more than 4 times the annual sum allocated for potential failure repairs budget. After having structured the problem, a DSS was built to support maintenance planning that will result in an annual maintenance plan based on the company´s strategic criteria and national regulations.

A new PROMETHEE-based approach applied within a framework for conflict analysis in Evidence Theory integrating three conflict measures

Abstract This article seeks to expand how conflict is classified in Evidence Theory, which is also known as Dempster-Shafer Theory (DST). Therefore, alternatives for classifying conflict under DST are discussed and a new approach is proposed based on the PROMETHEE method. To illustrate the use of this approach proposed, a numerical example from the literature on this topic is considered. The results obtained by comparing this proposal and the approach suggested in the literature indicate consistency in the classification. Thus, this novel approach enables the analysis of conflict in DST to be broadened to include decision-makers (DMs) who are more familiar with the structure of the PROMETHEE method. This method includes several intra-criteria preference functions and considers different thresholds within the scale. In addition to a new approach for classifying conflict under DST, this article proposes an adaptation to the disaggregation model based on linear programming in the literature, from which the parameters of the model can be generated using examples of conflict defined by the DM. The approach proposed uses this aggregated information from the DM and disaggregates it in order to define the parameters of the model, and results in a multicriteria analysis of conflict in Evidence Theory.