Browse Prior Art Database

High performance design for deploying features

IP.com Disclosure Number: IPCOM000129120D
Original Publication Date: 2005-Sep-28
Included in the Prior Art Database: 2005-Sep-28
Document File: 3 page(s) / 45K

Publishing Venue

IBM

Abstract

Disclosed is a software design for realizing high performance of deploying and expanding optional equipment items of car models in a specification matrix. This new design incorporates backtracking algorithm into the mechanism of generating possible combinations, so that it can thereby reduce execution time considerably from hours to several minutes.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 50% of the total text.

Page 1 of 3

High performance design for deploying features

Disclosed is a software design for realizing high performance of deploying and expanding optional equipment items of car models in a specification matrix. This new design incorporates backtracking algorithm into the mechanism of generating possible combinations, so that it can thereby reduce execution time considerably from hours to several minutes.

 An automotive product requires huge amounts of parts because of plethora of its variations to meet varying needs of fastidious customers. These circumstances lead to the increase in complexity and execution time for managing specifications of various car models.

 The equipment specifications of family car models are generally managed in the form of a matrix which contains many marks representing applications of equipment items, part of which are optional (see fig.1). In order to determine a unique specification of each car model in the family, every optional mark in the matrix must be deployed, whose logical combinations may usually result in extravagantly huge volume, because every optional pattern of a car must be multiplied.

 However, real combinations for the production of a car are ordinarily restricted by other data of combinatorial conditions. The deployment of optional items is performed basically with these two kinds of data.

Fig.1

 In contrast with conventional design that generates all logically possible combinations of optional equipment items and selects only realizable ones afterwards, the new design, by applying the algorithm of branch pruning through backtracking, generates only those combinations which are needed to check if they are allowed by combination restriction conditions, aiming at reducing the execution time. The points of the new design are as follows.

1. Decompose every optional mark of equipment groups in a specification matrix within the range of mutually exclusive items, and make linked lists of partial mark columns.
2. Combine every partial mark column of equipment groups in a specification matrix, in order to generate a unique combination of application marks for a car model.
3. Check conditions for restricting combinations, if necessary, during the combination of partial mark columns and move immediately to generate the next possible combination, if the conditions prohibit the current one.

Decompose Every Optional Mark within Equipment Groups

 Every optional mark is first decomposed within the range in which every optional item is mutually exclusive. For example, because high grade, standard, and minimum audio are mutually exclusive within the equipment group of audio kit, their optional marks are decomposed exhaustively as illustrated in fig.2.

Fig.2

 All optional marks in a specification matrix of a car model are decomposed in the same way in advance. Thus each equipment group has its own partial mark columns, which are implemented in memory as nodes of a linked list pointed by their equipment group (see fig.3)...