To improve the on-site management level of open-pit mines and the computing efficiency of the on-site visualization system of intelligent construction,without changing the overall logic architecture of the on-site visualization system,the communication mode of all terminal IoT probes is changed to 5G communication.A computer network security evaluation algorithm based on particle optimization neural network is studied.The neural network is used to assign values to the particle swarm initialization process,and after the particle swarm optimization convergence is completed,the neural network is used to further reduce the dimensionality of its output results,finally,4 data warning levels are built using the fuzzy decision matrix.The simulation software was used to test the system,and it was found that the algorithm had high comprehensive efficiency in detecting flood attacks,scanning attacks,injection attacks,and internal faults.The application of 5G network security technology effectively reduces the management workload for on-site infrastructure,greatly increasing the throughput of system data and prominently improving the sensitivity of visualization and network stability.