Aiming at the problem of nonlinear observation model mismatch and insufficient anti-interference ability of SINS/GNSS integrated navigation system in complex dynamic environment, this paper proposes an adaptive robust...
详细信息
Aiming at the problem of nonlinear observation model mismatch and insufficient anti-interference ability of SINS/GNSS integrated navigation system in complex dynamic environment, this paper proposes an adaptive robust filtering algorithm with improved fading factor. Aiming at the problem that the traditional Kalman filter is easy to diverge in severe heave motion and abnormal observation, a multi-source information fusion framework integrating satellite positioning geometric accuracy factor (PDOP), solution quality factor (Q value), effective satellite observation number (Satnum), and residual vector is constructed. The dynamic weight adjustment mechanism is designed to realize the real-time optimization of the fading factor. Through the collaborative optimization of robust estimation theory and adaptive filtering, a dual robust mechanism is constructed by combining the sequential update strategy. In the measurement update stage, the observation weight is dynamically adjusted according to the innovation covariance, and the fading memory factor is introduced in the time update stage to suppress the error accumulation of the model. The experimental results show that compared with EKF, Sage-Husa adaptive filtering and robust filtering algorithms, the three-dimensional positioning accuracy is improved by 47.12%, 35.26%, and 9.58%, respectively, in the vehicle strong maneuvering scene. In the scene of ship-borne heave motion, the corresponding increase is 19.44%, 10.47%, and 8.28%. The research results provide an effective anti-interference solution for navigation systems in high dynamic and complex environments.
Molten salts in phase change materials offer significant advantages, including high thermal storage density, a wide operational temperature range, and low cost. However, the development of novel high-latent-heat molte...
详细信息
Molten salts in phase change materials offer significant advantages, including high thermal storage density, a wide operational temperature range, and low cost. However, the development of novel high-latent-heat molten salts remains largely empirical. Machine learning offers the potential to expedite theoretical advancements and enable precise, cost-efficient performance predictions. Nonetheless, the diversity of molten salt s complicates the accuracy and generalizability of machine learning models. This study proposes a novel latent heat prediction methodology that integrates data analysis and machine learning. A comprehensive dataset encompassing various inorganic salts was systematically analyzed to extract key features influencing latent heat. Subsequently, a predictive model was constructed by combining a backpropagation neural network (BPNN) with particle swarm optimization (PSO). The PSO-BPNN model demonstrated high predictive accuracy, achieving R2 values of 0.9389 and 0.9413 for binary and ternary molten salts, respectively, with experimental validation indicating prediction errors within 10 %. This approach establishes a high-precision, scalable framework for predicting the latent heat of multicomponent molten salts, thereby advancing the design of salts with tailored thermal properties and offering a valuable reference for predicting other thermophysical characteristics.
The use of optimization algorithms is essential to train neural networks effectively. The usage of a combination of two different optimizers is proposed in this method that, used together, can perform single optimizer...
详细信息
We test a simple bi-fidelity strategy for accelerating trajectory optimization under uncertainty in the presence of robust constraints. Our approach combines the accuracy of high-fidelity models with the computational...
详细信息
ISBN:
(数字)9781624107115
ISBN:
(纸本)9781624107115
We test a simple bi-fidelity strategy for accelerating trajectory optimization under uncertainty in the presence of robust constraints. Our approach combines the accuracy of high-fidelity models with the computational efficiency of low-fidelity counterparts to increase the accuracy of constraint computation for a given cost. Specifically, we accelerate sample average approaches that use high-fidelity models, with additional samples of low-fidelity models for assessing whether or not a risk constraint is violated. The low-fidelity model's purpose is thus to reduce the uncertainty in the estimate of constraint violation, at the expense of introduction of bias. Our empirical test on a simulation of glider dynamics tasked with avoiding a region indicates an order-of-magnitude less failures using the bi-fidelity strategy.
This paper studies rigid body trajectory optimization for spacecraft landing. Formulating the rigid-body dynamics on the special Euclidean group (SE(3)) mitigates singularity and nonuniquness issues of attitude parame...
详细信息
ISBN:
(数字)9781624107115
ISBN:
(纸本)9781624107115
This paper studies rigid body trajectory optimization for spacecraft landing. Formulating the rigid-body dynamics on the special Euclidean group (SE(3)) mitigates singularity and nonuniquness issues of attitude parameterization sets. To perform the optimization problem using these dynamics, classical methods for optimization are generalized to Riemannian manifolds and trajectory optimization problems. Riemannian on-manifold optimization methods are used to provide an algorithm for optimizing the spacecraft landing problem on SE(3). In addition, inequality constraints and control input limits are applied, and the problem is solved using a Riemannian augmented Lagrangian method with trapezoidal collocation.
This article delves into an innovative radar working pattern recognition algorithm based on multi-layer perceptron (MLP). Through carefully designed optimization algorithms, we systematically searched and determined t...
详细信息
ISBN:
(纸本)9798400716959
This article delves into an innovative radar working pattern recognition algorithm based on multi-layer perceptron (MLP). Through carefully designed optimization algorithms, we systematically searched and determined the optimal MLP network structure to solve the radar operating pattern recognition problem. In a detailed simulation experiment, we carefully analyzed the effects of various network parameters, including the number of network layers, number of neurons, learning rate, and batch rate. The experimental results show that the MLP network can exhibit optimal performance when the number of layers is 5, the number of neurons is 512, the learning rate is 0.006, and the batch rate is 10. This discovery provides us with a highly promising solution to the problem of radar working pattern recognition.
Rock masses are naturally affected by discontinuities, joints and fractures that affect their exploitation. After block extraction, different cutting pattern can produce different recovery ratio of the block. The opti...
详细信息
Rock masses are naturally affected by discontinuities, joints and fractures that affect their exploitation. After block extraction, different cutting pattern can produce different recovery ratio of the block. The optimization of the cutting pattern can be computed if the discontinuities are mapped by the use of non-destructive methods. We propose a software code able to compute the number of intersected slabs by different cuttings scenarios. The algorithm adopts a brute force computation of several scenarios as specified by the user. The software uses the Open MP library in order to reduce computation time.
This paper presents a novel bio-inspired algorithm inspired by starlings' behaviors during their stunning murmuration named starling murmuration optimizer (SMO) to solve complex and engineering optimization proble...
详细信息
This paper presents a novel bio-inspired algorithm inspired by starlings' behaviors during their stunning murmuration named starling murmuration optimizer (SMO) to solve complex and engineering optimization problems as the most appropriate application of metaheuristic algorithms. The SMO introduces a dynamic multi-flock construction and three new search strategies, separating, diving, and whirling. The separating search strategy aims to enhance the population diversity and local optima avoidance using a new separating operator based on the quantum harmonic oscillator. The diving search strategy aims to explore the search space sufficiently by a new quantum random dive operator, whereas the whirling search strategy exploits the vicinity of promising regions using a new operator called cohesion force. The SMO strikes a balance between exploration and exploitation by selecting either a diving strategy or a whirling strategy based on the flocks' quality. The SMO was tested using various benchmark functions with dimensions 30, 50, 100. The experimental results prove that the SMO is more competitive than other state-of-the-art algorithms regarding solution quality and convergence rate. Then, the SMO is applied to solve several mechanical engineering problems in which results demonstrate that it can provide more accurate solutions. A statistical analysis shows that SMO is superior to the other contenders. (c) 2022 Elsevier B.V. All rights reserved.
To improve the model's efficiency, people use many different methods, including the Transfer Learning algorithm, to improve the efficiency of recognition and classification of image data. The study was carried out...
详细信息
To improve the model's efficiency, people use many different methods, including the Transfer Learning algorithm, to improve the efficiency of recognition and classification of image data. The study was carried out to combine optimization algorithms with the Transfer Learning model with MobileNet, MobileNetV2, InceptionV3, Xception, ResNet50V2, DenseNet201 models. Then, testing on rice disease data set with 13.186 images, background removed. The result obtained with high accuracy is the RMSprop algorithm, with an accuracy of 88% when combined with the Xception model, similar to the F1, Xception model, and ResNet50V2 score of 87% when combined with the Adam algorithm. This shows the effect of gradients on the Transition learning model. Research, evaluate and draw the optimal model to build a website to identify diseases on rice leaves, with the main functions including images and recording of disease identification points for better management purposes on diseased areas of rice.
Numerical simulation of the Boltzmann equation, most commonly by the DSMC method, can provide highly accurate descriptions of supersonic, rarefied flows. However, it is computationally expensive in the transitional fl...
详细信息
ISBN:
(数字)9781624107115
ISBN:
(纸本)9781624107115
Numerical simulation of the Boltzmann equation, most commonly by the DSMC method, can provide highly accurate descriptions of supersonic, rarefied flows. However, it is computationally expensive in the transitional flow regime, with Knudsen numbers 0.01 less than or similar to Kn less than or similar to 0.1, motivating the use of simpler, but less accurate models such as the BGK equation. We develop and implement optimization methods for calibrating machine learning-based BGK models to high-fidelity DSMC data. We derive and validate an adjoint equation for optimization over the BGK equation with a general form of the collision frequency. We train a neural network model for the collision frequency using automatic differentiation to implement the adjoint equations, and apply this model to the formation of a 1D shock.
暂无评论