Situation Assessment for Soccer Robots using Deep Neural Network

Abstract

Situation assessment system describes the relationship among entities, events, entities and events, and provides a complete high-level description of the current situation based on the scene object domain model acquired by data fusion system. Whereas, problems of poor objectivity, low accuracy, etc. always existing in evaluation methods of the complex systems attract interests of many researchers. This paper puts forward a situation assessment for soccer robots using deep neural network. The key situation factors are designed firstly to build the data set and the model input. A deep neural network model are proposed in this paper to train the map relationship from the input layer to the output layer. The results of experiments in the soccer robot platform demonstrate that the proposed method has better properties in efficiency than the competing methods.