مدل سازی چندهدفه فازی زمان بندی پروژه با محدودیت منابع چندمهارته: با قابلیت تغییر سطح مهارت ها و انقطاع فعالیت ها (مقاله علمی وزارت علوم)
درجه علمی: نشریه علمی (وزارت علوم)
آرشیو
چکیده
مسئله زمان بندی پروژه، یکی از مهم ترین و کاربردی ترین مفاهیم مدیریت پروژه است. بسیاری از شرکت ها و سازمان هایی که پروژه محورند، استراتژی کاهش هزینه های متغییر را در اجرای پروژه دنبال می کنند. با توجه به محیط کسب وکار کنونی، بسیاری از شرکت ها علاوه بر پایین آوردن هزینه های خود، به دنبال پیشگیری از تأخیر در اتمام پروژه اند. در این پژوهش، یک مدل ریاضی چندهدفه فازی زمان بندی پروژه با محدودیت منابع چندمهارته، با قابلیت تغییر سطح مهارت ها ارائه شد که هدف آن بهینه کردن سیاست زمان بندی پروژه و استخدام مهارت هاست. با توجه به چند هدفه بودن مدل، از یک رویکرد برنامه ریزی آرمانی استفاده شده است که مدل تک هدفه معادل حاصل می شود. نظر به اینکه مسئله زمان بندی پروژه چندمهارته جزء مسائل ان پی سخت محسوب می شود و مسئله پیشنهادی نیز حالت توسعه یافته مسئله مذکور است، درنتیجه آن نیز جزء مسائل ان پی سخت است. به همین سبب برای حل مسئله پیشنهادی، روش فرا ابتکاری ژنتیک چندهدفه ژنتیک و فاخته انتخاب و برای حل مسئله از آن استفاده شد. در ادامه، مقدار بهینه پارامترهای الگوریتم های پیشنهادی با استفاده از رویکرد تاگوچی تعیین و سپس نتایج محاسباتی برای مجموعه ای از مسائل نمونه تولیدشده توسط نرم افزار رنجن 1، ارائه و عملکرد الگوریتم ها ارزیابی و آنالیز شد. نتایج نشان می دهد الگوریتم ژنتیک چندهدفه عملکرد بهتری نسبت به الگوریتم فاخته چندهدفه دارد. در پایان نیز یافته ها جمع بندی و پیشنهادهایی به منظور تحقیقات آتی ارائه شد.Fuzzy multi-objective modeling of project scheduling with multi-skill resource constraints with the ability to change the level of skills and interrupt activities
Purpose: Time and cost are significant factors in every project. By reducing the resources allocated to the project, project costs are reduced, while the reduction of available resources means the inability to simultaneously implement activities or activities in the shortest possible time, which in turn increases the duration of the project. This is although, in all projects, the completion of the projects in the earliest time is considered one of the important parameters of the project. Considering the highly practical application of the examined problem, project scheduling by investing multi-skill resources with the possibility of changing the skill level in fuzzy conditions can be considered a positive step towards creating project scheduling problems.
Design/methodology/approach: In this paper, the proposed mathematical model of meta-heuristic genetic algorithms to solve the proposed model is discussed and explained in detail. Several skills are needed to perform each activity. The goal is to optimally determine resource availability and find the best schedule by minimizing investment in resources.
Findings: Considering the activities' need for different skills as well as the expertise of the project members in different skills, it seems obvious that each activity can be done with several different situations in terms of human resources allocation, which might be only for one activity, reaching more than 10 modes. As a result, compared to MRCPSP, this issue has a much higher complexity. The Resource Investment Problem (RIP) is a variant of RCPSP where renewable resource constraints are considered decision variables. In many projects, managers, in addition to making decisions about the time of implementation of activities, should determine the number of resources allocated to activities in each period of the implementation of activities according to the status of the project, which means ignoring the constant pattern of resource consumption for activities during their implementation.
Practical implications: By comparing the algorithms with the indicators of maximum extension, distance from the ideal solution, distance, and several Pareto solutions, it was found that the multi-objective genetic algorithm performs far better than the multi-objective Cuckoo algorithm regarding the criteria, distance from the ideal solution, and the largest expansion. However, in terms of the number of Pareto solutions, the algorithm is not superior to the other algorithms. Therefore, it can be concluded that the multi-objective genetic algorithm has relatively a better performance than the multi-objective Cuckoo algorithm.
Social implications: In this research, each activity can be performed with several different situations in terms of human resource allocation, which may reach more than 10 situations just for one activity. As a result, compared to MRCPSP, this issue has a higher level of complexity. Literature review indicates that being multi-skilled increases the productivity, quality, and consistency of work and gives managers more flexibility in work allocation.
Originality/value: One of the most important branches of project scheduling knowledge is the problem of project scheduling with limited resources. This new concept has led to the development of one of the most general modes of scheduling problems under the title of multi-mode project scheduling with limited resources, which solves many real problems and can be modeled for application.
تبلیغات
