Abstract: The current cybersecurity situation is getting worse. In order to improve the accuracy of network security situation prediction, a network security situation prediction method based on SOA_BP neural network is proposed. The algorithm uses the four behavioral characteristics of SOA algorithm: the self-interest, altruism, pre-action and uncertainty reasoning to determine the search strategy, find the best fitness individual, obtain the optimal weight and threshold, and then assign them to the random initial weights and thresholds. After training the neural network, the predicted values are obtained. Finally, it is compared with the predicted values obtained by othertwo optimization algorithms. The experimental results show that this prediction algorithm has higher accuracy, smaller error and better stability