利索-全球专利检索

  1. 登录
  2. 注册
发明公开
CN106980133A 利用神经网络算法补偿和修正的GPS INS组合导航方法及系统 无效 - 驳回

利用神经网络算法补偿和修正的GPS INS组合导航方法及系统
摘要：
本发明公开一种利用神经网络算法补偿和修正的GPSINS组合导航方法及系统：构建基于GPS与惯性导航模块的组合导航模型，设计Kalman滤波器对组合导航模型的组合导航数据进行滤波，输出惯性导航模块的导航误差数据；在GPS信号正常状态下，惯性导航数据作为神经网络模型的训练样本的输入数据，Kalman滤波器对导航数据滤波后的输出数据作为神经网络模块的训练样本的输出数据，对神经网络模型进行训练；在GPS信号丢失状态下，持续利用训练好的神经网络模型来预测惯性导航模块的输出误差，并用该预测的输出误差对惯性导航进行补偿和修正。本发明利用神经网络算法补偿和修正GPSINS组合导航模型，实现GPS信号丢失情况下，惯性导航系统在神经网络算法的辅助下输出精确的导航数据。
摘要（英）：

The invention discloses a GPSINS combination navigation method and system utilizing a neural network algorithm to realize compensation and correction. The method comprises steps that a combination navigation model based on a GPS and an inertial navigation module is constructed, a Kalman filter is designed to filter combination navigation data of the combination navigation model, and navigation error data of the inertial navigation module is outputted; in a normal GPS signal state, inertial navigation data is taken as input data of a training sample of a neural network model, output data of the navigation data after filtered by the Kalman filter is taken as output data of a training sample of a neural network module, and the neural network model is trained; in a GPS signal loss state, the trained neural network model is utilized to predict an output error of the inertial navigation module, and the predicted output error is further utilized to compensate and correct inertial navigation. The method is advantaged in that the GPSINS combination navigation model is compensated and corrected through utilizing the neural network algorithm, and the inertial navigation system is assisted by the neural network algorithm to output the accurate navigation data in the GPS signal loss state.

0/0