The traditional method uses terminal network monitoring method to estimate the security probability of multi-track segmentation network, but the detection performance is unsatisfactory. A security probability estimation and intrusion detection algorithm for multi-track segmentation networks in network attack environment based on fuzzy reasoning is proposed. The security probability estimation model of multi-track segmentation network in network attack environment is constructed. Fuzzy reasoning and probability density feature detection method are combined to evaluate the security data of multi-track segmentation network. The infection membership characteristics of multi-track segmentation network intrusion data are extracted, and the security probability calculation and virus attack detection of multi-track segmentation network are realized. The results show that the proposed algorithm has higher accuracy in calculating the security probability of multi rail segmented network, realizes the security probability calculation and data detection of multi rail segmented network, and enhances the security defense of multi rail segmented network.