Airspace complexity evaluation method based on deep unsupervised learning
An unsupervised learning and complexity technology, applied in the direction of neural learning methods, kernel methods, biological neural network models, etc., can solve the problems of inaccurate data classification, inaccurate evaluation results, and evaluation difficulties, so as to reduce inaccurate model learning , accurate evaluation results, and the effect of reducing the cost of manpower and material resources
Active Publication Date: 2020-04-21
BEIHANG UNIV
View PDF8 Cites 3 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0006] In view of the current airspace complexity assessment of traffic control, there are problems such as the inaccurate classification of data leading to inaccurate assessment results, or the need for a large number of high-quality manual calibration samples, which makes the assessment difficult, which brings wrong conclusions to controllers and misjudgments, etc. situation, the present invention provides a spatial complexity evaluation method based on deep unsupervised learning
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0022] In order to facilitate those of ordinary skill in the art to understand and implement the present invention, the present invention will be further described in detail and in-depth below in conjunction with the accompanying drawings.
[0023] With the rise of deep learning, unsupervised learning based on deep neural network is widely favored again. Using deep neural network, it is theoretically possible to realize arbitrary complex nonlinear mapping. Therefore, it is possible to learn the nonlinear characteristics of data space based on data samples. It solves the blindness of traditional unsupervised learning methods. The autoencoder model in deep unsupervised learning can give the low-dimensional embedding of the original data with a black box model. Although this low-dimensional embedding lacks interpretability, through a large number of scientific research and experimental verification, this low-dimensional embedding The representation of often accurately describes t...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention provides an airspace complexity evaluation method based on deep unsupervised learning. The airspace complexity evaluation method is used for air traffic control. The method comprises thefollowing steps: establishing an airspace complexity evaluation model by adopting a deep learning network of a stacked auto-encoder model, carrying out low-dimensional embedding representation on aninput airspace complexity index factor, and obtaining a clustering centroid of airspace complexity by utilizing an obtained low-dimensional embedding point; according to the method, soft assignment distribution and real assignment distribution of low-dimensional embedded points are utilized to construct a cost function, a gradient descent method is adopted to train the established model, and a trained spatial domain complexity evaluation model and clustering centroids of three spatial domain complexities are obtained and used for evaluating the current spatial domain complexity. The method does not depend on an air traffic controller to calibrate labels, uses an unsupervised method to classify the complexity of the airspace sector data, greatly reduces the cost of manpower and material resources, achieves the accurate mining of the airspace complexity data, and enables an evaluation result to be more accurate.
Description
technical field [0001] The invention belongs to the technical field of airspace situation assessment of air traffic control, and in particular relates to an airspace complexity assessment method based on deep unsupervised learning, so as to perform traffic control based on it. Background technique [0002] With the development of the air transport industry, the air traffic flow has increased sharply, and the airspace environment has become more and more complex, which has brought an increasing workload to air traffic controllers. The calculation of airspace complexity refers to the complexity of the airspace commanded by the controller, including visible aircraft operations, invisible program operations, and so on. Airspace complexity calculation is one of the widely studied problems in the field of air traffic control. The main research content of this problem is: how to use the operating characteristics of the airspace situation to give the complexity level of the airspac...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06Q10/06G06Q50/30G06N3/08
CPCG06Q10/0639G06N3/08G06Q50/40G06Q10/04G08G5/0043G06N3/088G06N3/048G06N3/045G06N20/10G08G5/0082G08G5/0095
Inventor 杜文博曹先彬李碧月朱熙李宇萌
Owner BEIHANG UNIV