Offshore engineering structures, including offshore wind turbines, offshore oil platforms and ships for maintenance and material transport, must work reliably over longer periods of time. Your reliability and security requirements are higher due to the considerable complexity of the offshore environment compared to the country. It must withstand a variety of complex loads, including wind and waves. This complex and random suggestion represent significant challenges for the dynamic reaction analysis and the design of offshore structures.
In order to ensure structural high performance and security, the dynamic reaction properties of offshore technology structures are examined with complex environmental pollution. Examples are the design of offshore wind support structures [1] and maintenance ship gangways [2]Present [3]. For the support structures, environmental analyzes have shown that the alignment of the jack structure has a major impact on the quality of the jacket and the maximum embarrassment depth [4]. Various support configurations were assessed [5]. Tran et al. [6] Developed a three -legged jacket structure and a four -legged coat structure to analyze the strength under ambient loads, and found that the load direction has a more impact on the four -legged structure. With regard to the health monitoring of the support structure[7]Present [8]Wang et al. [9] proposed a digital Twin method to identify the location and the severity of the offshore jacket platform. By determining dynamic model [10]Intelligent algorithms [11] Like neural network [12] are used to carry out gangway movement compensation [13]Present [14].
In order to meet the performance requirements of the offshore structure design, structural optimization methods in offshore structures were also taken over. With regard to structural size optimization[15]Present [16]Present [17]Some scientists start intelligent algorithms [18]Present [19] In addition to conventional methods. Several studies [20]Present [21] have also examined the reduction in the mass of the jacket and at the same time maintained structural security. The size optimization design methods are restricted by a narrow construction room, which leads to limited performance improvement. Therefore, approaches that are able to extend constructions are required.
Topology optimization methods [22]Present [23] Get the purpose of the structural optimization design by controlling the material distribution in the structure, offering a significantly expanded design space beyond conventional methods beyond the conventional methods and seeing widespread use across technical disciplines. Various methods of topology optimization methods [24]Present [25]Present [26]Present [27]Present [28] were proposed and applied to structural dynamic topology optimization. The use of density methods such as Simp usually leads to very low rigidity problems in areas with low density, which leads to “local modes”, which leads to “local modes” [29] Problems with the vague topology optimization results and unclear structure in the structural dynamic topology optimization. Some scientists suggested an improved interpolation models of the density in relation to this problem, such as: B. rational material in punishment (ramp) [30]Present [31]Present [32]Present [33]. When solving structural dynamic topology optimization problems with more degrees of freedom, numerous pairs of own are usually generated, which means that solving the topology optimization problem of frequency domain reactions is very complicated. Yoon[34] Detailed three mode reduction methods, including the mode postponement method (MDM). [35]Present [36]Present [37]And developed a new method for reducing the mode. The mode overlay method was generally used to solve the reactions of the frequency domains. Younger[38] developed an efficient parallel topology optimization method, and MDM was used to reduce the order of structures of two standards for efficient frequency response analysis. Kang et al. [39] used the complex mode overlay method to cope with the non-proportional damping in optimizing the distribution of damping materials, which enables the reactions of the inpatient state to be calculated. Yan et al. [40] MDM adopted to reduce the model order and the computing costs when optimizing the impact response optimization of Shell structures. In addition to the dynamic analysis methods mentioned above, Kang et al. [41]Present [42] proposed a consecutive iterysis and design method (SIAD) that integrates the iteration of the mode solution and the topology literation and thus achieves improved arithmetic efficiency.
To improve the dynamic reaction properties of offshore engineering structures under complex stress[43]Present [44]Present [45]Present [46]Present [47]. For example, Lee et al. [48] improved the fatigue performance of wind turbine transition parts through regular load optimization. Long et al. [49] Used static load adjustment to optimize the basic layout for the light design. Tian used an equivalent static stress or special methods to reduce the mass of the jacket [50] and platform [51]Although the dynamic performance improves considerably. Random suggestion is difficult to solve, and the structural dynamic topology optimization is time-consuming, which makes the topology optimization more difficult. With the increasing complexity of the environment in which the technical structure is located, it will be exposed to more complex loads and several working conditions, so that topology optimization with random vibration becomes increasingly critical.
For the topological optimization of structural reactions under random stimulation, the right to compensation and efficiency of the analysis of the power spectra density (PSD) are of crucial importance for practical technical applications. Various approaches were proposed to tackle the dynamic structural optimization under complex random suggestions. Yang et al.[52] examined the topology optimization of Continua under a filtered white noise, while Golecki et al.[53] Examined bridge structures under random moving loads. Su et al.[54] Considered the reliable optimization of subdued structures that are not exposed to inpatient seismic inputs. Although the complete square combination (CQC) is often used in the random analysis of the frequency domain, the high computing costs limits efficiency. Lin's pseudo excitation (PEM) was proposed as a means [55]. By converting the random vibration into pseudo -harmonic reply analysis, the calculation efficiency of this method is significantly better than that of complete square combination methods and was successfully applied to piezoelectric application [56]Present [57]acoustically[58]Damping design [59]And other fields. The following studies have further expanded the application of PEM under complex working conditions such as basic acceleration stimulation[60]Dynamic stress restriction[61]and not inpatient vibration[62]. In large -scale structural optimization, scientists have effectively integrated the PEM with advanced computing strategies[63]Present [64] To compensate for the compensation and efficiency. However, the conventional iterative scheme for nested loops in PEM for the analysis of the shift response of pseudo requires repeated iterations of structural intrinsic properties (self -frequencies and self -modes), which leads to excessive computing costs and low efficiency. This is a critical challenge that requires efficient solution methods for the topology optimization of offshore engineering structures that work under complex stress.
In this study, we propose a new efficient topology optimization framework for offshore structures that are subjected to complex random stimulation. In the optimization model, the rack random response at the key positions are assumed as an objective function. A PEM-Siaad method is suggested to overcome the challenges through complex random analysis and nested optimization loops. In this context, the strategy is based on a successive iteration of analysis and design[41]Present [65] is used to calculate the structural dynamic properties and the pseudo excitation (PEM) is used to significantly improve the arithmetic efficiency of random analysis. With the proposed method, the initially inaccurate modal estimates during the iterative process are increasingly refined, which enables a more precise analysis of the dynamic reaction under random suggestion. This eliminates repeated independent modes in every nest cycle, which increases the overall efficiency of the optimization process. The random sensitivities of the structure were derived using the adjoint variable method and the method for moving asymptots [66] was used as optimization algorithm. Numerical examples show that the proposed method achieves a maximum improvement of the arithmetic efficiency of over 50 % compared to the conventional nested loop set with random suggestion via various optimization scenarios and at the same time maintain the results that exactly meet the conventional results. In addition, two representative offshore structures were successfully optimized, with their dynamic performance being significantly improved.