بهينه يابي تخصيص ريسك در پروژه هاي ساخت؛ با الگوريتم بهينه سازي جامعه‌ي مورچگان (ACO)

Risk Allocation Optimization in a Construction Project- Ant Colony Algorithm

هدف از تخصيص ريسك، انتقال ريسك ها به تواناترين عامل و تفاهم براي تسهيم متناسب سود قرارداد است. با توجه به تاثير تخصيص ريسك بر زمان و هزينه‌ي تمام‌شده‌ي پروژه، انتخاب مناسب‌ترين تخصيص ريسك ها براي كارفرما اهميتي حياتي دارد. در اين نوشتار، براي اولين بار انتخاب مناسب ترين تخصيص ريسك هاي پروژه، در قالب يك مسيله‌ي بهينه يابي به صورت كمّي مدل‌سازي شده است. مدل بهينه‌ي پيشنهادي با هدف دستيابي به بالاترين اطمينان در كسب اهداف پروژه با كمترين هزينه‌ي ممكن، يك پارامتر تصميم كاربردي (سود هر عامل براي پذيرش ريسك) را تعريف كرده و يك مدل بهينه يابي را براساس الگوريتم جامعه‌ي مورچگان توسعه داده است. همچنين آناليز حساسيت مدل، مي تواند بهترين سناريوي ممكن در ارايه‌ي ضمانت‌هاي مالي را توصيه كند. قابليت مدل پيشنهادي با پياده سازي آن براي يك پروژه‌ي موردي نيز نمايش داده شده است.

Risk allocation in construction project in order to transfer the risks to most capable party and provide a base for project profit sharing, has a strong influence on time and cost of project. In this paper for the first time, allocation of risk to most proper party is defined as an optimization problem and a quantitative model for risk allocation optimization introduced. Since the aim of risk allocation is defined as achieving to project objectives within maximum reliability and minimum cost, the proposed model is introduced an applicable and logical decision parameter and developed an optimization algorithm based on Ant Colony Optimization (ACO) method. The proposed model also provides a useful decision tools for project owner to select best insurance package by including the sensitivity analysis. In order to design the ACO optimization algorithm of the risk allocation optimization model; initially the objectives of owners within allocation process is identified and objective function is defined. Respect to the objective function, then, risk allocation problem is restructured as an optimization problem and decision parameter, constraints and the flowchart of model’s structure are defined. These parameters and constraints formulated in mathematical equation; and an optimization algorithm is designed based on Ant Colony Optimization (ACO). By receiving “profit requested for risk bearing by each participant” as input, the proposed model is calculating the cost of risk management for the owner and then minimize the objective function in an Ant Colony Optimization algorithm. Two constraints is defined and formulated in the proposed model in order to simulate real decision process of risk allocation. As follows: 1) the maximum financial credit of each party to compensate the consequence of any risk and 2) the financial ability of owner to ensure risk owner against risk events. Varying this constraint, sensitivity analysis would be available in the model to optimize the guaranty package for the owner. Therefore, this model could guide the decision maker to address most effective guaranties to best party. This model is applied in a case study to present its capability and usefulness. According to findings of applying the proposed model, it could be concluded that risk sharing in a project should be done based on party competency and willingness. It is also concluded that the owner shall be participate in risk allocation process if it wishes to achieve best reliability in project objectives within minimum cost.