Neural Network and Synergetic Approaches to Solving the Problem of Situation Identification

Examples of intelligent technologies of the most relevant direction of situation identification are considered. Intellectual approaches the property of memory addressable by content (associativity) were selected. For a comparative analysis we chose: the Hopfield neural network (representative of neural technologies) and Haken’s synergetic model (an alternative approach to identification). The theoretical aspects of the two intelligent technologies are considered, the common and divergent attributes of the approaches are highlighted. A practical study was conducted on the example of the chosen task. To carry out the study both methods were implemented in the programming language Python. The chosen task was the determination of the threat category when controlling a helicopter, which is a problem of recovery and recognition of a noisy pattern. Based on the results of solving the problem via each method, data on the additional advantages and disadvantages of both approaches were collected experimentally. As a result of the analysis, the synergetic model showed a more accurate classification result. The Hopfield neural network has a number of disadvantages, which are critical for situations in which an immediate and accurate solution is required. In addition, the synergetic model has a number of advantages over the Hopfield neural network: the absence of false attractors, the ability to rank information, the possibility of additional learning and the universality of the algorithm in relation to different types of problems.