7 results about "Thermosphere" patented technology

The invention belongs to the technical field of atmospheric density prediction, and discloses a cross-source data fusion thermal layer atmospheric density hierarchical progressive full-scale prediction system and method, which utilize respective advantages of different gradient data sources to construct a reference density field-corrected density field-instantaneous refined density field three-level hierarchical fusion refinement architecture. Cross-source data calibration and physical constraint modeling are combined to realize month-year scale long-term prediction, day-week scale medium-short-term prediction and short-term prediction of the thermal layer atmospheric density. According to the method, through a three-level layered refinement architecture, the long-term coverage advantage of TLE data, the mesoscale variable rate description advantage of precise orbit data and the high-frequency and high-precision advantages of accelerometer data are fully mined, the defect of a single data source is avoided, full-scale and full-area density precise description is achieved, the thermal layer atmospheric density refinement precision and prediction suitability can be improved, and the method is suitable for large-scale and large-scale measurement. And data cost is reduced.

The invention belongs to the technical field of ionosphere total electron content prediction, and particularly discloses an ionosphere total electron content prediction method based on physical parameter enhancement. The data set comprises ionosphere TEC observation data, space environment data and two-dimensional physical parameter data output by a thermal layer-ionosphere electrodynamic circulation model, and the thermal layer-ionosphere electrodynamic circulation model is based on a first principle; the evolution of a thermal layer and an ionized layer is calculated by adopting thermodynamics, momentum conservation, electromagnetic field coupling and plasma kinetic equations; the data set is preprocessed, so that the TEC observation data and the physical parameter data are matched on the spatial scale; inputting the preprocessed data set into a pre-established ConvLSTM deep learning model for training; and using the trained ConvLSTM model to predict the future ionosphere TEC distribution. According to the invention, the prediction precision of the ionized layer TEC in an extreme space weather state can be effectively improved.

This invention discloses a method for reconstructing the peak electron density of the ionosphere, comprising: S1, constructing a peak electron density reconstruction neural network SRON2NN model, which adopts a fully connected feedforward neural network and includes an input layer, several hidden layers, and an output layer; S2, constructing and introducing key physical features, including effective solar radiation EUV×cosχ and thermosphere ΣO / N2; S3, training the model using ionospheric F2 layer peak electron density NmF2 data, time information, geomagnetic index Dst, solar activity index F10.7, effective solar radiation EUV×cosχ, and ΣO / N2 obtained from COSMIC-1 radio occultation observations; S4, reconstructing and verifying the ionospheric F2 layer peak electron density NmF2 based on the trained SRON2NN model. This invention achieves high-precision and interpretable reconstruction of the ionospheric F2 layer peak electron density NmF2 globally by explicitly introducing effective solar radiation and thermosphere ΣO / N2.

The invention discloses a thermal layer atmospheric density calibration method and system, and solves the technical problem of poor precision of a thermal layer atmospheric density calibration result caused by the fact that an existing thermal layer atmospheric density calibration method depends on a single technical path to carry out calibration work. The method comprises the following steps: acquiring actually measured real density, space environment data and an empirical model, and fitting to obtain daily factor data of the empirical model at the current moment; calculating a residual error by combining the actually measured real density and the original calculation density at the current moment, calculating a priori model predicted value by relying on the residual error and daily factor data, and comparing to obtain a calibrated residual error of the priori model; then training a residual fusion recurrent neural network to be trained by using the residual and the spatial environment data based on an adaptive moment estimation optimizer, and determining a trained network; and finally, outputting a residual prediction value by the network, superposing the residual prediction value with a priori model prediction value, and outputting the final calibration density of the thermal atmosphere.

The application belongs to the technical field of atmospheric density prediction, and discloses a thermosphere atmospheric density layered progressive full-scale prediction system and method based on cross-source data fusion, which utilizes respective advantages of different gradient data sources, constructs a three-level layered fusion refinement architecture of a benchmark density field, a correction density field and an instantaneous refined density field, and realizes month-year scale long-term prediction, day-week scale medium and short-term prediction and short-term prediction of thermosphere atmospheric density by combining cross-source data calibration and physical constraint modeling. Through the three-level layered refinement architecture, the application fully excavates the long-term coverage advantage of TLE data, the mesoscale variability description advantage of precise orbit data and the high-frequency high-precision advantage of accelerometer data, avoids defects of a single data source, realizes accurate description of full-scale and full-area density, and can improve refinement accuracy and prediction adaptability of thermosphere atmospheric density and reduce data cost.

A computer-implemented method for determining an orbital effect on a low-Earth orbiting object includes, forming an ensemble of thermosphere-ionosphere state models to represent thermosphere-ionosphere state model uncertainty, assimilating measured thermosphere-ionosphere data into a select thermosphere-ionosphere state model of the ensemble in accordance with a Kalman gain derived from statistics of the ensemble, computing drag on the low-Earth orbiting object from elements of the assimilated select model of thermosphere-ionosphere state models and randomizing elements of the assimilated select thermosphere-ionosphere model in accordance with statistics of the ensemble to generate a new ensemble of thermosphere-ionosphere state models. A system is arranged to perform the method. A computer program product is arranged to cause one or more computers to perform the method.

The present application belongs to the technical field of thermosphere mass density prediction method and space exploration and information processing, and specifically discloses a thermosphere mass density short-term prediction method and system based on ensemble learning and an electronic device, step 1: pre-processing and standard normalization of feature data; step 2: constructing an ensemble learning model MBiLE, integrating a multivariate perceptron model and a bidirectional long short-term memory neural network model, and training; step 3: selecting the optimal model and model parameters according to the loss value size of the proposed adaptive loss function RMSE-Huber; step 4: combining the feature data, and predicting the thermosphere mass density in a wider height range by using the optimal model. The proposed MBiLE model has strong stability and high reliability, can short-term predict the thermosphere mass density at a lower height, and is conducive to further realizing the prediction of the thermosphere mass density at any time in the future.