关键词: 海上风电工程, 复杂海上环境, 多级供应网, 集成优化, MC-SAA

Abstract:

With the deployment and implementation of China's 'dual carbon' and 'green development' strategies， wind power in China has entered a period of rapid development， and the vigorous development of offshore wind power projects has also become an important measure in the transformation of the energy structure.Offshore wind power project has the characteristics of distributed factory construction， all kinds of components in several factories distributed in different areas to complete the prefabricated production， transportation， and assembly in stages， and then transported by sea to the offshore construction site for on-site installation. Based on the engineering physical process of integrating the physical entities of wind turbine units from parts to whole （parts → components → assemblies → wind turbines）， suppliers and manufacturers of each component level under the influence of supply-demand relationships， form a multi-level supply network. This network， in conjunction with offshore construction sites， constitutes a coherent and orderly operating whole. However， the long supply chain of offshore wind power projects and the close connections between each stage， along with the complex circulation of physical goods and information exchange within the supply network， pose significant management challenges. Moreover， as a type of major infrastructure project facing the open and complex maritime environment， the installation of wind turbines is greatly affected by uncertain factors such as sea winds， waves， tides， and undercurrents， resulting in high environmental complexity. The construction progress and assembly resource planning of offshore wind power projects are significantly influenced by real-time environmental changes and impacts， which also affect the overall supply network at the back end， thus increasing the complexity of overall supply decision-making. Therefore， the construction progress and resource demand planning for offshore wind power projects are significantly influenced by real-time environmental changes. The challenge addressed in this paper is how to accurately predict changes in maritime wind and wave conditions， and then integrate these predictions with their impacts on construction operations and progress to formulate a multi-level supply integration decision-making strategy， thereby ensuring the effective operation of the overall supply network.Firstly， this paper constructs a scenario-based multi-level supply integration optimization model for offshore wind power projects to provide support for subsequent algorithm design. This model， based on the uncertain characteristics of the real offshore environment， proposes a multi-level supply integration optimization model that addresses different environmental scenarios. The model can devise optimal multi-level supply integration plans for the entered |S| scenarios. Additionally， drawing from a case study of a wind power engineering supply network， the model considers the multi-level supply relationships corresponding to the physical process of integrating the physical entities of offshore wind power projects from parts to whole （parts → components → assemblies → wind turbines）， which involves more complex elements and interrelations. For example， some parts are supplied by suppliers and transported to manufacturers to be made into components， then shipped to the dock， forming the supply chain： Supplier （parts） → Manufacturer （components） → Dock （components）； some components may be supplied by suppliers and pre-assembled at the dock into assemblies， then transported to the wind farm site for final installation， forming the supply chain： Supplier （components） → Dock （assemblies） → Wind farm site （turbines）. Furthermore， there are two types of turbines hoisting schemes in the construction process： one involves pre-assembling blades and hubs into rotors at the dock， then transporting them by ship for installation at the wind farm； the other involves transporting blades and hubs directly to the wind farm site for onsite assembly. Based on the above physical processes， it is necessary to integrate and collaboratively consider decisions on production， storage， and other logistics at each node along the supply chains， including decisions on matching supply with demand and choosing transportation modes. The decision variables of the model include （1） dynamic demand， production， and inventory levels of parts， components， or assemblies at various supply nodes， including suppliers， manufacturers， and docks； （2） the volumes of various components transported between supply nodes and the choices of sea or land transportation modes.Subsequently， a Markov chain-sample average approximation two-stage algorithm （MC-SAA） is designed for a solution. Initially， the first stage uses the MC model to simulate possible regional maritime environmental scenarios based on the historical environmental data of the target sea area. Then， based on the construction process requirements of the wind power project and the various offshore wind and wave scenarios simulated by the MC， a set of scenarios for construction progress and resource demand plans are further generated. In the second stage， using the SAA algorithm， multiple cycles of random scenario sampling are conducted， with each scenario set being input into the model described in section 2.2 to solve for the optimal solution. Then， the best multi-level supply integration decision-making plan is selected through comparisons of the solutions under all sampled scenarios.Finally， a real case of a wind power project in the southern maritime area of China is taken as its background， verifying the effectiveness of the model and methods through computational experiments， and exploring several managerial insights. It indicates that， overall， compared to two empirical methods， the algorithm proposed in this paper not only effectively achieves better supply decision-making outcomes but also ensures the adequate supply of wind power project components across a broader range of maritime environmental conditions， while minimizing costs associated with resource idleness， such as vessel leasing and storage， thereby demonstrating stronger scenario adaptability.The research on multi-level supply optimization for engineering projects under uncertain environments is extended and practical guidance and decision support are provided for supply decision-making in offshore wind power projects.

Key words: offshore wind power project, complex maritime environment, multi-level supply network, integrated optimization, MC-SAA