An improved-self organizing mapping neural network (SOM) track correlation algorithm is proposed to overcome the disadvantages of the traditional track correlation algorithm in practical application, which has low correlation accuracy and complex calculation process. The algorithm uses all the feature data after Kalman filtering as the input vector to improve the learning rules of SOM neural network, so that the same neuron in the competition layer can only win once at the same time when the data of the same sensor is input. The feature data of multi-sensor at the same time are input into SOM neural network for self-organizing clustering. Experiments show that the SOM neural network indirectly transforms the track correlation problem into the track clustering problem. Thus, the problem of increasing computation caused by the increase of the number of targets and sensors is well solved. At the same time, improved SOM neural network can use the multidimensional characteristic information of the targets, and can effectively solve the problems of low correlation accuracy and slow convergence rate when the dense targets track crosses and bifurcates.
Track Correlation Algorithm Based on Improved SOM Neural Network
Lect. Notes Electrical Eng.
International Conference on Autonomous Unmanned Systems ; 2021 ; Changsha, China September 24, 2021 - September 26, 2021
Proceedings of 2021 International Conference on Autonomous Unmanned Systems (ICAUS 2021) ; Kapitel : 146 ; 1473-1483
2022-03-18
11 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Intelligent networked automobile real-time track prediction method based on improved neural network
Europäisches Patentamt | 2023
|Safety automatic driving lane changing track planning method based on improved LSTM neural network
Europäisches Patentamt | 2021
|Track Predition Based on CNN-GRU Neural Network
IEEE | 2023
|