Bayesian Optimization Algorithm Research Articles

Fault diagnosis for spent fuel shearing machines based on Bayesian optimization and CBAM-ResNet

Spent fuel shearing machines in nuclear power plants are important equipment for the head end of spent fuel reprocessing in power reactors. Condition monitoring and fault diagnosis play important roles in ensuring the safe operation of spent fuel shearing machines, avoiding serious accidents, and reducing their maintenance time and cost. Existing research on fault diagnosis of spent fuel shearing machines has some shortcomings: (a) the current research on fault diagnosis of shearing machines is small and diagnostic accuracy is not high. The research methodology of shearing machines needs to be updated; (b) the high difficulty in obtaining fault data and the often limited and highly informative fault data for shearing machines lead to low diagnostic performance. To solve these problems, this study constructs a residual network (ResNet) model based on Bayesian optimization (BO) and convolutional block attention module (CBAM). First, dual-channel difference method is introduced into the preprocessing of noise signals, and two data enhancements were applied to the Mel spectrograms used as inputs to the model. Second, the attention mechanism CBAM is introduced to improve the ResNet to enhance the deep feature extraction ability of the network, and the BO algorithm is used to train the hyperparameters, such as the optimizer, and retrain the network model after obtaining the optimal hyperparameters. Finally, the feasibility and effectiveness of the proposed model are verified through experiments on the noise signals of spent fuel shearing machines. The experimental results show that the diagnostic accuracy of the constructed model is 93.67%, which is a significant improvement over the other methods.

Gray code based gradient-free optimization algorithm for parameterized quantum circuit

A Gray code based gradient-free optimization (GCO) algorithm is proposed to update the parameters of parameterized quantum circuits (PQCs) in this work. Each parameter of PQCs is encoded as a binary string, named as a gene, and a genetic-based method is adopted to select the offsprings. The individuals in the offspring are decoded in Gray code way to keep Hamming distance, and then are evaluated to obtain the best one with the lowest cost value in each iteration. The algorithm is performed iteratively for all parameters one by one until the cost value satisfies the stop condition or the number of iterations is reached. The GCO algorithm is demonstrated for classification tasks in Iris and MNIST datasets, and their performance are compared by those with the Bayesian optimization algorithm and binary code based optimization algorithm. The simulation results show that the GCO algorithm can reach high accuracies steadily for quantum classification tasks. Importantly, the GCO algorithm has a robust performance in the noise environment.

A random forest regression with Bayesian optimization-based method for fatigue strength prediction of ferrous alloys

Fatigue is a prevalent failure mode in mechanical structures, and the significance of fatigue strength as a process parameter for predicting the life of a material or structure cannot be emphasized enough. Given the high cost and time consumption associated with traditional experimental approaches for determining fatigue strength, which leads to prolonged analysis and design cycles, it is crucial to investigate efficient and precise methods for predicting fatigue strength. In this paper, we present a fatigue strength prediction method that utilizes Bayesian optimization for the random forest regression (RFR) model. During the training of the RFR model, a Bayesian optimization algorithm is employed to optimize the five main parameters of the model, mitigating the issue of overfitting or underfitting. The optimized RFR model demonstrates superior performance and accuracy in comparison to linear regression (LR), artificial neural network (ANN), decision tree regression (DTR), and support vector regression (SVR). This is determined by evaluating the mean absolute error (MAE), root mean square error (RMSE), R-squared (R2), and mean absolute percentage error (MAPE) scores of the feature parameter-fatigue strength dataset. The fatigue strength of unknown ferrous alloys can be predicted with the proposed model, which largely reduces the need to obtain fatigue strength data through traditional fatigue tests and saves fatigue analysis time and test costs.

Optimization of a Weyl-semimetal-based near-field heat transfer system

Near-field radiative heat transfer (NFRHT) is inherently limited by Lorentz reciprocity, which can be partially mitigated through the application of an external magnetic field. However, recent studies have discovered that Weyl semimetals (WSMs) possess the ability to achieve nonreciprocal surface plasmon polariton modes without relying on an external magnetic field. Building upon the nonreciprocal nature of WSMs, this study investigates NFRHT between two WSM slabs and compares different WSM materials. The primary focus of the comparison lies in analyzing the heat flux under varying emitter and receiver temperatures. Three of the selected WSM materials exhibit similar performance in NFRHT, showing sensitivity to a decrease in the receiver temperature due to their low dielectric tensor components. In contrast, Eu2IrO7 displays distinctive characteristics, being sensitive to changes in the emitter temperature as a result of its higher dielectric tensor components. Additionally, a reassessment of the nonreciprocity in Co2MnGa, Eu2IrO7 and Co3Sn2S2 is conducted. Furthermore, Bayesian optimization algorithms are employed to optimize parameters and address the challenge of low thermal radiation observed in existing WSM NFRHT systems. The results highlight the achievement of an optimal WSM material that achieves an impressive heat flux of 5.29 × 108 (W/m2) at an emitter temperature of 310 K and a receiver temperature of 290 K. Notably, this heat flux surpasses that of Co2MnGa by 2.5 times in the context of NFRHT. The discoveries presented in this study promises to offer practical implications for the development of nonreciprocal thermal energy management systems.

Weakly-supervised structural surface crack detection algorithm based on class activation map and superpixel segmentation

This paper proposes a weakly-supervised structural surface crack detection algorithm that can detect the crack area in an image with low data labeling cost. The algorithm consists of a convolutional neural networks Vgg16-Crack for classification, an improved and optimized class activation map (CAM) algorithm for accurately reflecting the position and distribution of cracks in the image, and a method that combines superpixel segmentation algorithm simple linear iterative clustering (SLIC) with CAM for more accurate semantic segmentation of cracks. In addition, this paper uses Bayesian optimization algorithm to obtain the optimal parameter combination that maximizes the performance of the model. The test results show that the algorithm only requires image-level labeling, which can effectively reduce the labor and material consumption brought by pixel-level labeling while ensuring accuracy.

State-of-health and remaining-useful-life estimations of lithium-ion battery based on temporal convolutional network-long short-term memory

Accurate estimations in state of health (SOH) and remaining useful life (RUL) are significant for safe and efficient operation of batteries. With the development of big data and deep learning technology, the neural network method has been widely used for SOH and RUL estimations because of its excellent nonlinear mapping performance, adaptive performance and self-learning performance. In this paper, a novel hybrid model based on temporal convolutional network-long short-term memory (TCN-LSTM) for SOH and RUL estimations is proposed. The hyperparameters of each layer in the model are optimized using Bayesian optimization algorithm. Three different models, including convolutional neural network-long short-term memory (CNN-LSTM) model, temporal convolutional network (TCN) model and long short-term memory (LSTM) model, are adopted as comparisons to evaluate the performance of the proposed model. All the models are tested using two public battery datasets from National Aeronautics and Space Administration (NASA dataset) and Oxford University (OX dataset). In SOH task, the TCN-LSTM model achieves an accuracy improvement of >16 % and 14 % in NASA and OX datasets, respectively. In RUL task, the accuracies of the TCN-LSTM model and the CNN-LSTM model are superior to other models in NASA dataset; while the LSTM model and the CNN-LSTM model have better performance in OX dataset.

A robust error correction method for numerical weather prediction wind speed based on Bayesian optimization, variational mode decomposition, principal component analysis, and random forest: VMD-PCA-RF (version 1.0.0)

Abstract. Accurate wind speed prediction is crucial for the safe and efficient utilization of wind resources. However, current single-value deterministic numerical weather prediction methods employed by wind farms do not adequately meet the actual needs of power grid dispatching. In this study, we propose a new hybrid forecasting method for correcting 10 m wind speed predictions made by the Weather Research and Forecasting (WRF) model. Our approach incorporates variational mode decomposition (VMD), principal component analysis (PCA), and five artificial intelligence algorithms: deep belief network (DBN), multilayer perceptron (MLP), random forest (RF), eXtreme gradient boosting (XGBoost), light gradient boosting machine (lightGBM), and the Bayesian optimization algorithm (BOA). We first predict wind speeds using the WRF model, with initial and lateral boundary conditions from the Global Forecast System (GFS). We then perform two sets of experiments with different input factors and apply BOA optimization to tune the four artificial intelligence models, ultimately building the final models. Furthermore, we compare the aforementioned five optimal artificial intelligence models suitable for five provinces in southern China in the wintertime: VMD-PCA-RF in December 2021 and VMD-PCA-lightGBM in January 2022. We find that the VMD-PCA-RF evaluation indices exhibit relative stability over nearly a year: the correlation coefficient (R) is above 0.6, forecasting accuracy (FA) is above 85 %, mean absolute error (MAE) is below 0.6 m s−1, root mean square error (RMSE) is below 0.8 m s−1, relative mean absolute error (rMAE) is below 60 %, and relative root mean square error (rRMSE) is below 75 %. Thus, for its promising performance and excellent year-round robustness, we recommend adopting the proposed VMD-PCA-RF method for improved wind speed prediction in models.

Uncertainty quantification and identification of SST turbulence model parameters based on Bayesian optimization algorithm in supersonic flow

AbstractThe Reynolds‐Averaged Navier–Stokes (RANS) model is the main model in engineering applications today. However, the normal value of the closure coefficient of the RANS turbulence model is determined based on some simple basic flows and may no longer be applicable for complex flows. In this paper, the closure coefficient of shear stress transport (SST) turbulence model is recalibrated by combining Bayesian method and particle swarm optimization algorithm, so as to improve the numerical simulation accuracy of wall pressure in supersonic flow. First, the obtained prior samples were numerically calculated, and the Sobol index of the closure coefficient was calculated by sensitivity analysis method to characterize the sensitivity of the wall pressure to the model parameters. Second, combined with the uncertainty of propagation parameters by non‐intrusive polynomial chaos (NIPC). Finally, Bayesian optimization is used to quantify the uncertainty and obtain the maximum likelihood function estimation and optimal parameters. The results show that the maximum relative error of wall pressure predicted by the SST turbulence model decreases from 29.71% to 9.00%, and the average relative error decreases from 9.86% to 3.67% through the parameter calibration of Bayesian optimization method. In addition, the system evaluated the calibration effect of three criteria, and the calibration results parameters under the three criteria were all better than the calculated results of the nominal values. Meanwhile, the velocity profile and density profile of the flow field were also analyzed. Finally, the same calibration method was applied to the supersonic hollow cylinder and BSL (Baseline) turbulence model, and the same calibration results were obtained, which verified the universality of the calibration method.

Research on fault diagnosis of electric submersible pump based on improved convolutional neural network with Bayesian optimization

A conventional intelligent fault diagnosis approach for electric submersible pumps (ESPs) is heavily dependent on manual expertise for feature extraction. Meanwhile, a conventional convolutional neural network (CNN) exhibits an excess of parameters and requires a substantial volume of training samples. In this paper, a fault diagnosis algorithm of ESPs based on a Bayesian optimization-one-dimensional convolutional neural network-support vector machine (BO-1DCNN-SVM) is proposed by combining a 1DCNN with a SVM and using the algorithm BO to tune the improved 1DCNN. First, the method uses the self-extracting feature capability of the 1DCNN to solve the problem that traditional diagnosis methods over-rely on manual experience extraction. Meanwhile, the last layer of the convolutional neural network Softmax layer is replaced by the SVM to effectively process a few sample data. The accuracy and generalization ability of the fault classification of the electric submersible pump are improved. Then, the Bayesian optimization algorithm is used to find the optimal combination of hyperparameters for the improved 1DCNN-SVM model to further improve the prediction performance of the fault diagnosis model. Finally, the experimental results achieved a classification accuracy of 96.64%, which is 5% higher than existing CNN approaches with data samples of similar scale in this paper. The proposed method also proved to be highly accurate and robust in fault diagnosis.

Optimizing Perovskite Thin‐Film Parameter Spaces with Machine Learning‐Guided Robotic Platform for High‐Performance Perovskite Solar Cells

AbstractSimultaneously optimizing the processing parameters of functional thin films remains a challenge. The design and utilization of a fully automated platform called SPINBOT is presented for the engineering of solution‐processed functional thin films. The SPINBOT is capable of performing experiments with high sampling variability through the unsupervised processing of hundreds of substrates with exceptional experimental control. Through the iterative optimization process enabled by the Bayesian optimization (BO) algorithm, the SPINBOT explores an intricate parameter space, continuously improving the quality and reproducibility of the produced thin films. This machine learning (ML)‐guided reliable SPINBOT platform enables the acceleration of the optimization process of perovskite solar cells via a simple photoluminescence characterization of films. As a result, this study arrives at an optimal film that, when processed into a solar cell in an ambient atmosphere, immediately yields a champion power conversion efficiency (PCE) of 21.6% with satisfactory performance reproducibility. The unsealed devices retain 90% of their initial efficiency after 1100 h of continuous operation at 60–65 °C under metal‐halide lamps. It is anticipated that the integration of robotic platforms with the intelligent algorithm will facilitate the widespread adoption of effective autonomous experimentation to address the evolving needs and constraints within the materials science research community.

NPENAS: Neural Predictor Guided Evolution for Neural Architecture Search.

Neural architecture search (NAS) adopts a search strategy to explore the predefined search space to find superior architecture with the minimum searching costs. Bayesian optimization (BO) and evolutionary algorithms (EA) are two commonly used search strategies, but they suffer from being computationally expensive, challenging to implement, and exhibiting inefficient exploration ability. In this article, we propose a neural predictor guided EA to enhance the exploration ability of EA for NAS (NPENAS) and design two kinds of neural predictors. The first predictor is a BO acquisition function for which we design a graph-based uncertainty estimation network as the surrogate model. The second predictor is a graph-based neural network that directly predicts the performance of the input neural architecture. The NPENAS using the two neural predictors are denoted as NPENAS-BO and NPENAS-NP, respectively. In addition, we introduce a new random architecture sampling method to overcome the drawbacks of the existing sampling method. Experimental results on five NAS search spaces indicate that NPENAS-BO and NPENAS-NP outperform most existing NAS algorithms, with NPENAS-NP achieving state-of-the-art performance on four of the five search spaces.

Book-ahead ride-hailing trip and its determinants: Findings from large-scale trip records in China

The cruising of ride-hailing vehicles generates negative externalities such as traffic congestion and vehicular emissions. These externalities can be mitigated by reducing cruising driving via operating book-ahead ride-hailing services, where the platform dispatches and routes drivers based on precise information on travelers’ departure time and origin–destination (OD). However, the effects of factors influencing book-ahead ride-hailing trips have rarely been empirically examined with real data. Using six-month trip data from China, this study employs a gradient boosting decision tree (GBDT) method with hyperparameters optimized by the Bayesian optimization algorithm to examine the factors associated with book-ahead ride-hailing trips across OD pairs (hexagon cells-to-hexagon cells) at various spatial scales. The relative importance rankings generated from this study indicate that trip features, weather conditions, and accessibility to transportation hubs are significant determinants correlated with the usages of book-ahead ride-hailing. The partial dependence plots demonstrate the nonlinear threshold effects of these determinants on the hourly number of book-ahead ride-hailing per OD pair. Moreover, this study compares the differences in associations between peak and non-peak hours as well as weekdays and weekends. The disparity in the nonlinear threshold effects between weekdays and weekends is only observable during the evening peak period and not at other times. These findings provide valuable insights into developing practical strategies for promoting book-ahead ride-hailing services.

Data-driven rapid lifetime prediction method for lithium-ion batteries under diverse fast charging protocols

Accurate lithium-ion battery lifetime prediction is essential for equipment maintenance and safety assurance in practical applications. The influence of different charge protocols on the lifetime varies greatly, which raises tremendous challenges for rapid lifetime prediction. In this paper, a comprehensive data-driven rapid lifetime prediction method for batteries under diverse fast charging protocols is proposed. It straightforwardly establishes the relationship between the charging conditions and the impacts on battery lifetime. A deep neural network model with the Bayesian optimization algorithm for hyperparameter search (BOA-DNN) and three traditional shallow machine learning algorithms consisting of support vector machine (SVM), Gaussian process regression (GPR), and extreme gradient boosting (XGBoost) are employed. The verification is performed on an authoritative and popular dataset with 69 types of distinct fast charging conditions. The eventual results indicate that the proposed approach has attained reliable performance of battery lifetime degradation trajectory and battery end-of-life (EOL) prediction, in which the BOA-DNN model outperforms SVM, GPR, and XGBoost. Especially in battery EOL prediction, the prediction mean absolute percentage error of the BOA-DNN model is below 3.58%.

Water level prediction of Liuxihe Reservoir based on improved long short-term memory neural network

Abstract To meet the demand of accurate water level prediction of the reservoir in Liuxihe River Basin, this paper proposes an improved long short-term memory (LSTM) neural network based on the Bayesian optimization algorithm and wavelet decomposition coupling. Based on the improved model, the water levels of Liuxihe Reservoir and Huanglongdai Reservoir are simulated and predicted by the 1 h prediction length, and the prediction accuracy of the improved model is verified separately by the 3, 6 and 12 h prediction lengths. The results show that: first, Bayesian optimization coupling can significantly reduce the average absolute error and root mean square error of the model and improve the overall prediction accuracy, but this algorithm is insufficient in the optimization of model extremum; Wavelet decomposition coupling can significantly reduce the outliers in model prediction and improve the accuracy of extremum, but it plays relatively weaker role in the overall optimization of the model. Second, by the prediction lengths of 1, 3, 6 and 12 h, the improved model based on the LSTM neural network and coupled with Bayesian optimization and wavelet decomposition is superior to Bayesian optimization and wavelet decomposition coupling model in overall prediction accuracy and prediction accuracy of extremum.

Performance degradation prediction method of PEM fuel cells using bidirectional long short-term memory neural network based on Bayesian optimization

Proton exchange membrane (PEM) fuel cell is the core equipment that can directly convert hydrogen energy into electricity. In the process of long-term operation, due to the aging of membrane electrode assembly and other components, the fuel cell performance gradually deteriorates. The voltage prediction of fuel cells is very important for performance and lifetime optimization. Long short-term memory neural network is one of the widely used prediction methods. Based on the prediction method of bidirectional long short-term memory neural network, the hyperparameters of the neural network model by Bayesian optimization algorithm is optimized to improve the accuracy of fuel cell performance degradation prediction. When the sampling time interval is 25 min and the training set is 45 %, the root mean square error and the average absolute percentage error of the prediction results is reduced to 6.3 mV and 0.1245 %, respectively. Moreover, by analyzing the influence of different sampling time intervals and training set proportion on the prediction results, a data set that takes into accounts both efficiency and accuracy is obtained. The proposed method based on Bayesian optimization can achieve more accurate performance degradation prediction.

Machine learning-based design for additive manufacturing in biomedical engineering

While ceramic additive manufacturing (AM) technologies have shown great promise to create functional scaffolds with tailored biomechanical properties, the true potential of these advanced techniques has not been fully exploited yet due to lack of practical design optimisation approaches. To address this challenge, a machine learning (ML)-based design approach is proposed herein where ceramic 3D printing techniques are combined to fabricate functionally graded tissue scaffolds composed of Triply Periodic Minimal Surfaces (TPMS), aiming to fulfil the anticipated biomechanical requirements for the target bone regeneration outcomes. The proposed ML based design strategy couples a Bayesian optimisation (BO) algorithm to enable time-dependent mechano-biological optimisation of the 3D printed ceramic scaffolds at a reasonably low computational cost. For a representative example relating to bone scaffolding in a segmental defect of sheep tibia, the simulated results demonstrate that the optimised functionally graded scaffolds significantly enhance bone ingrowth outcomes. Furthermore, a Lithography-based Ceramic Manufacturing (LCM) technique is employed to fabricate the optimised scaffolds based on the proposed ML-based design framework, followed by micro-CT analyses of the additively manufactured ceramic scaffolds to assess their geometric qualities. This study is expected to gain new insights into mechanical sciences on design for varying material conditions and provide an effective design tool for ceramic additive manufacturing.

Development of hybrid models based on deep learning and optimized machine learning algorithms for brain tumor Multi-Classification

Accurate classification of magnetic resonance imaging (MRI) images of brain tumors is crucial for early diagnosis and effective treatment in clinical studies. In these studies, many models supported by artificial intelligence (AI) have been proposed as assistant systems for experts. In particular, state-of-the-art deep learning (DL) models that have proven themselves in different fields have been effectively used in the classification of brain MRI images. However, the low accuracy of multiple classification of these images still leads researchers to conduct different studies in this field. Especially there is a need to develop models that achieve high accuracy on original images, and it is believed that this need can be met not only by DL models but also by classical machine learning (ML) algorithms. However, it is critical to choose the hyperparameters correctly for the hybrid use of ML algorithms with DL models. This study proposes a powerful new hybrid method to perform multiple classifications of brain tumors with high accuracy. This method also uses a novel convolutional neural network (CNN) model for feature extraction, and ML algorithms are used for feature classification. In addition, nine state-of-the-art CNN models are used for CNN performance comparison. The Bayesian optimization algorithm is used to obtain the optimal hyperparameter values of ML algorithms. The results obtained from the experimental studies show that the proposed hybrid model achieved 97.15% mean classification accuracy and 97% recall, precision, and F1-score values. Other hybrid models, including DarkNet19-SVM, DarkNet53-SVM, DenseNet201-SVM, EfficientNetB0-SVM, InceptionV3-SVM, NasNetMobile-SVM, ResNet50-SVM, ResNet101-SVM, and Xception-SVM, achieved mean classification accuracies of 95.01%, 95.58%, 96.87%, 97.01%, 95.3%, 95.01%, 96.3%, 95.87%, and 96.23%, respectively. Additionally, the proposed hybrid model exhibited remarkable time efficiency, accomplishing the classification process in a mere 67 min. Conversely, the model that exhibited the lowest time efficiency was the InceptionV3, with a processing time of 370 min. In terms of computational complexity, the EfficientNetB0 model is the most efficient. Despite the higher computational complexity of the proposed CNN model compared to some other models, it achieves the second-best classification accuracy. These results show that the proposed method performs better than previous studies on the same dataset. Especially in the classification problem, the optimized ML algorithms were superior to CNN classifiers. Finally, except for one, the proposed CNN model achieved better classification accuracies than the state-of-the-art CNN models.

A Multimodel Edge Computing Offloading Framework for Deep-Learning Application Based on Bayesian Optimization

With the rapid development of Internet of Things (IoT), data generated by IoT devices are also increasing exponentially. The edge computing has alleviated the problems of limited network and transmission delay when processing tasks of IoT devices in traditional cloud computing. And with the popularity of deep learning, more and more terminal devices are embedded with AI (Artificial Intelligence) processors for higher processing capability at the edge. However, the problems of Deep Learning task offloading in heterogeneous edge computing environment have not been fully investigated. In this paper, a multi-model edge computing offloading framework is proposed, using NVIDIA Jetson edge devices (Jetson TX2, Jetson Xavier NX, and Jetson Nano) and GeForce RTX GPU servers (RTX3080 and RTX2080) to simulate the edge computing environment, and make binary computational offloading decisions for face detection tasks. We also introduce a Bayesian Optimization algorithm, namely MTPE (Modified Tree-structured Parzen Estimator), to reduce the total cost of edge computation within a time slot including response time and energy consumption, and ensure the accuracy requirements of face detection. In addition, we employ the Lyapunov model to obtain the harvesting energy between time slots to keep the energy queue stable. Experiments reveal that MTPE algorithm can achieve the globally optimal solution in fewer iterations. The total cost of multi-model edge computing framework is reduced by an average of 17.94% compared to a single-model framework. In contrast to the Double Deep Q-Network (DDQN), our proposed algorithm can decrease the computational consumption by 23.01% for obtaining the offloading decision.

Automatic SWMM Parameter Calibration Method Based on the Differential Evolution and Bayesian Optimization Algorithm

In response to the low accuracy exhibited by the Storm Water Management Model (SWMM), we propose an enhanced Differential Evolution and Bayesian Optimization Algorithm (DE-BOA). This algorithm integrates the global search capability of the differential evolution algorithm with the local search capability of the Bayesian optimization algorithm, which enables a more comprehensive exploration of the vector solution space. A comparative analysis of various types of rainfall events is conducted. For model calibration and validation, a drainage subzone in Jinshazhou, Guangzhou City, is selected as the research subject. In total, 20 specific rainfall events are selected, and the DE-BOA algorithm outperforms the manual calibration, the differential evolution algorithm, and the Bayesian optimization algorithm regarding model calibration accuracy. Furthermore, the DE-BOA algorithm exhibits robust adaptability to rainfall events characterized by multiple peaks and higher precipitation levels, with the Nash–Sutcliffe efficiency coefficient values surpassing 0.90. This study’s findings could hold significant reference value for dynamically updating model parameters, thereby enhancing the model simulation performance and improving the accuracy of the urban intelligent water management platform.

Dispatch of highly renewable energy power system considering its utilization via a data-driven Bayesian assisted optimization algorithm

In recent years, renewable energy (RE) has been widely deployed, and the power system with high penetration RE is gradually formed. However, the high proportion of RE may threaten transmission security of power systems, which in turn limits its utilization. What is more, the interaction between RE penetration and power system transmission security has not been comprehensively investigated so far. To this end, we develop a bi-objective stochastic dispatch model to investigate the relationship between RE utilization and transmission security. It aims to solve the optimal power system dispatch (OPSD) problem with high penetration RE, in which the RE curtailment and the capacity margin of transmission lines are considered as two objectives of the dispatch problem and formulated in the probabilistic forms. With this, the proposed model is a complicated bi-objective stochastic optimization problem, which is difficult to be solved for traditional optimization algorithms. Therefore, we propose a data-driven Bayesian assisted optimization (DBAO) algorithm, based on Bayesian evolutionary optimization and estimation of distribution algorithm to improve the searching efficiency for the proposed model. Case studies on a modified Midwestern US power system verify the effectiveness of our proposed dispatch model and the optimization algorithm of DBAO.