改进NSGA-Ⅱ算法及监测天线部署优化研究

Improved NSGA-Ⅱ algorithm and research on monitoring antenna optimization deployment

为求解地面电磁辐射源监测天线的最佳部署位置,以达到准确监测、高效部署的目的,构建了以最大化覆盖率、最小化冗余率和未被覆盖障碍物数量为目标,以天线间保持连通性为约束条件的多目标优化模型。通过判断每两个障碍物之间的距离与天线工作覆盖范围最大半径的大小关系预设参考点作为若干天线部署位置;通过仿真比较了不同障碍物数量和参考点数量对覆盖率、冗余率等部署效率的影响,验证了障碍物数量对冗余率的影响更大,以增加冗余率为代价可减少天线覆盖区域内的障碍物数量。利用内存存储前几代非支配解的方法改进NSGA-Ⅱ算法,在改变环境后可以更快地求得新的最优解,提高收敛速度。对模型和改进后的算法进行了仿真分析,结果表明迭代速度平均提高了30%以上,平均迭代一次的时间减少了25%以上,验证了改进算法响应动态变化的有效性;通过算法评价指标mIGDB验证了改进算法的收敛性和多样性。

In order to solve the optimum deployment location of the ground radiation monitoring antenna to achieve accurate monitoring and efficient deployment,a multi-objective optimization model is constructed to maximize the coverage rate,minimize the redundancy rate and the number of uncovered obstacles.And the model is constrained by communication between antennas.By comparing the relationship between the distance between two obstacles and the maximum radius of the antenna’s working coverge range,the reference points are preset as the antenna deployment location.The effects of the number of obstacles and the number of reference points on the deployment efficiency such as coverage rate and redundancy rate are compared by simulation.It is verified that the number of obstacles has a greater effect on the redundancy rate,and that the number of obstacles in the antenna coverage area is reduced at the expense of the increased redundancy rate.The second generation of the non-dominant sorting genetic(NSGA-Ⅱ)algorithm is improved by using the memory to store the previous generations of non-dominant solutions.After changing the environment,the new optimal solution can be obtained faster and the convergence speed can be improved.The simulation results of the model and the improved algorithm show that the iteration speed is increased by more than 30% on average,and that the average iteration time is reduced by more than 25%,which verifies the effectiveness of the improved algorithm in response to dynamic changes.The convergence and diversity of the improved algorithm are verified by the algorithm evaluation index mIGDB.