With the continuous expansion of network scale and the increasing complexity of attack methods, traditional network security protection equipment has been unable to cope with large-scale network security detection and protection. However, most current operation and maintenance systems cannot reasonably evaluate and predict network security. In order to be able to evaluate and prevent the network security of the bridge BIM intelligent operation and maintenance system under the background of big data, this paper uses the KDD Cup99 data set and the network attack data in the bridge BIM network environment to simulate the method proposed in this paper. The comparison results verify that the network security risk perception method proposed in this paper can realize network security risk perception more accurately and efficiently. This paper proposes a data mining method based on Bayesian network algorithm to evaluate the risk value of the bridge BIM intelligent operation and maintenance system. During the period from 0 to 110 min, the network risk value increased from 0.003 to 0.91. It can be seen that with the deepening of the attack phase, the degree of network risk will also increase. This paper uses the detection rate, false negative rate, false positive rate, AUC and other indicators to conduct simulation experiments on the data prediction-based network security risk prediction algorithm and comparison algorithm proposed in this paper. Simulation experiments show that under the four simulation experiment environments (pl = pe = 0.01/0.15, n = 20/50), the AUC of this scheme is increased by 0.018, 0.053, 0.008 and 0.11, respectively. The proposed algorithms are better than the comparison algorithms.

随着网络规模的不断扩大和攻击手段的日益复杂，传统的网络安全防护设备已无法应对大规模的网络安全检测与防护。但是，大多数当前的运维系统无法合理地评估和预测网络安全性。为了能够在大数据背景下评估和防止网桥BIM智能运维系统的网络安全，本文采用KDD Cup99数据集和网桥BIM网络环境中的网络攻击数据对网络进行仿真。本文提出的方法。比较结果验证了本文提出的网络安全风险感知方法能够更准确，有效地实现网络安全风险感知。提出了一种基于贝叶斯网络算法的数据挖掘方法，以评估桥梁BIM智能运维系统的风险价值。在0到110分钟之间，网络风险值从0.003增加到0.91。可以看出，随着攻击阶段的加深，网络风险的程度也会增加。本文利用检测率，误报率，误报率，AUC等指标对本文提出的基于数据预测的网络安全风险预测算法和比较算法进行了仿真实验。仿真实验表明，在四种仿真实验环境（pl = pe = 0.01 / 0.15，n = 20/50）下，该方案的AUC分别增加了0.018、0.053、0.008和0.11。