Prasit Imtanavanich

Evolutionary computation with linear physical programming for solving a disassembly-to-order system

We solve the disassembly-to-order (DTO) problem by using Evolutionary Computation. DTO is a system where a variety of returned products are disassembled to fulfill the demand for specified numbers of components and materials. The main objective is to determine the optimal number of take-back EOL products for the DTO system which satisfy the desirable criteria of the system. One of the most widely used forms of Evolutionary Computation is Genetic Algorithm (GA). GA, which has the capability to improve a set of solutions over evolutionary steps, is used to generate optimal number of take-back EOL products. Moreover, linear physical programming (LPP), which has key features to entirely remove the decision maker (DM) from the process of choosing weights and to handle the vagueness of aspiration levels, is used to calculate fitness values in the GA process. A numerical example is considered to illustrate the methodology.

Calculating disassembly yields in a multi-criteria decision-making environment for a disassembly-to-order system

In this paper, we consider the disassembly-to-order (DTO) problem, where a variety of returned products are disassembled to fulfill the demand for specified numbers of components and materials. The objective is to determine the optimal numbers of returned products to disassemble so as to maximize profit and minimize costs. We model the DTO problem using a multi-criteria decision-making approach. Since the conditions of returned products are unknown, the yields from disassembly are considered to be stochastic. To solve the stochastic problem, we use one of the two heuristic approaches (viz., one-to-one approach or one-to-many approach) that converts the problem into a deterministic equivalent. We compare the performance of the two heuristic approaches using a case example.

Multi-criteria decision making for disassembly-to- order system under stochastic yields

In this paper, we consider the problem of determining the optimal number of returned products to disassemble to fulfill the demand for a specified number of parts. This is known as the disassembly-to-order (DTO) problem. The deterministic yield version of this problem has been addressed in the literature. Recently, the stochastic yield version of this problem with a single objective has also been reported in the literature. In this paper, we extend the methodology to include multiple objectives. To this end, we model the DTO problem using integer goal programming. The stochastic problem is solved by transforming it into its deterministic equivalent problem. This conversion is accomplished by considering the specific structures of the products with one core and one part (“one-to-one structure”) and apply it to handle the products with one core and multiple parts (“one-to-many structure”). For these special cases it is possible to solve the stochastic problem analytically so that valuable insights can be gained by comparing the stochastic and deterministic solutions. This will help us to determine effective deterministic yield equivalents. We present a case example to illustrate the methodology.

Solving a disassembly-to-order system by using Genetic Algorithm and weighted fuzzy goal programming

In this paper, Genetic Algorithm (GA) is used to solve the disassembly-to-order (DTO) problem. DTO is a system where a variety of returned products are disassembled to fulfill the demand for specified numbers of components and materials. The main objective is to determine the optimal number of take-back EOL (end-of-life) products for the DTO system which satisfy the desirable criteria of the system. We implement the Weighted Fuzzy Goal Programming (WFGP) to calculate the fitness values in GA process. We also consider product deterioration which affects the yield rates (e.g., older products tend to have lower yield rates for usable components) and use heuristic procedure to transform the stochastic disassembly yields into their deterministic equivalents. A numerical example is also considered.

GENERATING A COMPLETE DISASSEMBLY-TO-ORDER PLAN

Abstract We introduce techniques used to generate a complete disassembly-to-order (DTO) plan. DTO is a system where a variety of returned products are disassembled to fulfill the demand for specified numbers of components and materials. Complete DTO plan includes the optimal number of take-back products which maximize profit and minimize costs of the DTO system, the number of products to be sent through each disassembly station (non-destructive, destructive and mixed disassembly) including the type of disassembly to be performed (viz., complete or selective) and the optimal disassembly sequence of each product type to help minimize the complexities and the total disassembly time. Techniques implemented to solve the problem are Genetic Algorithm, Linear Physical Programming and refining algorithm. In this paper, we describe how these techniques can be combined to solve for a complete DTO plan. A numerical example is considered to illustrate the methodology.