Abstract

Currently, integrated navigation systems with the inertial navigation system (INS)/geomagnetic navigation system (GNS) have been widely used in underwater navigation of autonomous underwater vehicle (AUV). Restricting AUV to navigate in the geomagnetic suitability areas (GSA) as far as possible can effectively improve the accuracy of integrated navigation systems. In order to improve the classification accuracy of GSA, a new optimal classification method based on principal component analysis (PCA) and improved back propagation (BP) neural network is proposed. PCA is used to extract the independent characteristic parameters containing the main components. Then, considering similarity coefficient, the initial weights and thresholds of BP neural network is optimized by improved adaptive genetic algorithm (IAGA). Finally, the correspondence between the geomagnetic characteristic parameters and matching performance is established based on PCA and improved adaptive genetic algorithm and back propagation (IAGA-BP) neural network for the automatic recognition of GSA. Simulated experiments based on PCA and IAGA-BP neural network shows high classification accuracy and reliability in the GSA selection. The method could provide important support for AUV path planning, which is an effective guarantee for AUV high precision and long voyage autonomous navigation.

摘要

目前，惯性导航系统（INS）/地磁导航系统（GNS）组合导航系统已广泛应用于自主水下航行器（AUV）的水下导航。限制AUV尽可能在地磁适宜区（GSA）内导航，可以有效提高组合导航系统的精度。为了提高GSA的分类精度，提出了一种基于主成分分析（PCA）和改进的反向传播（BP）神经网络的新的最优分类方法。PCA用于提取包含主要成分的独立特征参数。然后，考虑相似系数，采用改进的自适应遗传算法（IAGA）优化BP神经网络的初始权值和阈值。最后，基于PCA和改进的自适应遗传算法和反向传播（IAGA-BP）神经网络建立地磁特征参数与匹配性能的对应关系，用于GSA的自动识别。基于PCA和IAGA-BP神经网络的模拟实验表明GSA选择具有较高的分类准确率和可靠性。该方法可为AUV路径规划提供重要支持，是AUV高精度长航程自主导航的有效保障。基于PCA和IAGA-BP神经网络的模拟实验表明GSA选择具有较高的分类准确率和可靠性。该方法可为AUV路径规划提供重要支持，是AUV高精度长航程自主导航的有效保障。基于PCA和IAGA-BP神经网络的模拟实验表明GSA选择具有较高的分类准确率和可靠性。该方法可为AUV路径规划提供重要支持，是AUV高精度长航程自主导航的有效保障。