At present, China’s offshore photovoltaic (PV) industry is developing rapidly. Risk assessment on offshore PV power generation (OPVPG) projects is important. The existing risk assessment methods in OPVPG projects in China are insufficient in dealing with different types of uncertainties. In this paper, we propose a computational model based on D Numbers and analytic network process (ANP) method to handle risk assessment on OPVPG projects. Firstly, factors that influencing offshore photovoltaic power generation projects in China are identified and the weights of the factors are determined by domain analyst using ANP. Secondly, risk evaluation judgment on each factor is given by expert referring to predefined anchors and linguistic labels under specific situations. Thirdly, the judgments are represented as D numbers and are integrated into a fused D number. Finally, the overall risk level of OPVPG in China is obtained based on the fused D number. The main contribution of the proposed method based on D numbers is to deal with uncertainties and the degree of confidence in experts judgments. Uncertainty includes situations like: do not know how to assign accurate beliefs to risk levels and incomplete frame of discernment. The degree of confidence shows the cognition of the experts. The situation that the confidence of expert is equal to 1 shows the certainty of experts. When the confidence of expert is equal to 0, it means that experts cannot decide the risk levels. It can be considered as a new strategy for risk assessment and enriches the framework of decision-making. Several case studies are illustrated to show the effectiveness and flexibility of the proposed method.

The development of surrogate models based on limited data is crucial in enhancing the speed of structural analysis and design optimization. Surrogate models are highly effective in alleviating the challenges between design variables and performance evaluation. Bidirectional Long Short-Term Memory (BiLSTM) is an advanced recurrent neural network that exhibits significant advantages in processing sequential data. However, the training of BiLSTM involves the adjustment of multiple hyperparameters (such as the number of layers, the number of hidden units, and the learning rate), which complicates the training process of the model. To enhance the efficiency and accuracy of neural network model development, this study proposes an Improved Whale Optimization Algorithm-assisted BiLSTM establishment strategy (IWOA-BiLSTM). The new algorithm enhances the initial population design and population position update process of the original Whale Optimization Algorithm (WOA), thereby improving both the global search capability and local exploitation ability of the algorithm. The IWOA is employed during the training process of BiLSTM to search for optimal hyperparameters, which reduces model training time and enhances the robustness and accuracy of the model. Finally, the effectiveness of the model is tested through a parameter optimization problem of a specific analog circuit. Experimental results indicate that, compared to traditional neural network models, IWOA-BiLSTM demonstrates higher accuracy and effectiveness in the optimal parameter design of analog circuit engineering problems.

On basis of fault categories detection, the diagnosis of rotor fault causes is proposed, which has great contributions to the field of intelligent operation and maintenance. To improve the diagnostic accuracy and practical efficiency, a hybrid model based on the particle swarm optimization-extreme gradient boosting algorithm, namely, PSO-XGBoost is designed. XGBoost is used as a classifier to diagnose rotor fault causes, having good performance due to the second-order Taylor expansion and the explicit regularization term. PSO is used to automatically optimize the process of adjusting the XGBoost’s parameters, which overcomes the shortcomings when using the empirical method or the trial-and-error method to adjust parameters of the XGBoost model. The hybrid model combines the advantages of the two algorithms and can diagnose nine rotor fault causes accurately. Following diagnostic results, maintenance measures referring to the corresponding knowledge base are provided intelligently. Finally, the proposed PSO-XGBoost model is compared with five state-of-the-art intelligent classification methods. The experimental results demonstrate that the proposed method has higher diagnostic accuracy and practical efficiency in diagnosing rotor fault causes.

The Markov method is a common reliability assessment method. It is often used to describe the dynamic characteristics of a system, such as its repairability, fault sequence and multiple degradation states. However, the “curse of dimensionality”, which refers to the exponential growth of the system state space with the increase in system complexity, presents a challenge to reliability assessments for complex systems based on the Markov method. In response to this challenge, a novel reliability assessment method for complex systems based on non-homogeneous Markov processes is proposed. This method entails the decomposition of a complex system into multilevel subsystems, each with a relatively small state space, in accordance with the system function. The homogeneous Markov model or the non-homogeneous Markov model is established for each subsystem/system from bottom to top. In order to utilize the outcomes of the lower-level subsystem models as inputs to the upper-level subsystem model, an algorithm is proposed for converting the unavailability curve of a subsystem into its corresponding $2\times 2$ dynamic state transition probability matrix (STPM). The STPM is then employed as an input to the upper-level system’s non-homogeneous Markov model. A case study is presented using the reliability assessment of the Reactor Protection System (RPS) based on the proposed method, which is then compared with the models based on the other two contrast methods. This comparison verifies the effectiveness and accuracy of the proposed method.

In order to meet the higher accuracy and system reliability requirements, the information fusion for multi-sensor systems is an increasing concern. Dempster–Shafer evidence theory (D–S theory) has been investigated for many applications in multi-sensor information fusion due to its flexibility in uncertainty modeling. However, classical evidence theory assumes that the evidence is independent of each other, which is often unrealistic. Ignoring the relationship between the evidence may lead to unreasonable fusion results, and even lead to wrong decisions. This assumption severely prevents D–S evidence theory from practical application and further development. In this paper, an innovative evidence fusion model to deal with dependent evidence based on rank correlation coefficient is proposed. The model first uses rank correlation coefficient to measure the dependence degree between different evidence. Then, total discount coefficient is obtained based on the dependence degree, which also considers the impact of the reliability of evidence. Finally, the discount evidence fusion model is presented. An example is illustrated to show the use and effectiveness of the proposed method.