基于粒子群优化算法的PSO-BP海底声学底质分类方法  

Back propagation neural network classification of sediment seabed acoustic sonar images based on particle swarm optimization algorithms

作　　者：陈佳兵[1,2] 吴自银[1,2] 赵荻能 周洁琼[1,3] 李守军[1,2] 尚继宏[1,2] 

机构地区：[1]国家海洋局第二海洋研究所,浙江杭州310012 [2]国家海洋局海底科学重点实验室,浙江杭州310012 [3]浙江大学地球科学学院,浙江杭州310027

出　　处：《海洋学报》2017年第9期51-57,共7页

基　　金：国家自然科学基金(41476049);科技基础性工作专项(2013FY112900);海洋公益项目(201105001)

摘　　要：利用粒子群优化算法(PSO)较强的鲁棒性和全局搜索能力等优点,将PSO算法与BP神经网络相结合,优化了BP神经网络分类时的初始权值和阈值。基于珠江河口三角洲的侧扫声呐图像数据,提取了海底声呐图像中砂、礁石、泥3类典型底质的6种主要特征向量,利用PSO-BP方法对海底底质进行分类识别。实验表明,3类底质分类精度均大于90%,高于BP神经网络70%左右的分类精度,表明PSO-BP方法可有效应用于海底底质的分类识别。By combining Particle Swarm Optimization （PSO） with BP neural network, the initial weights and thresholds of BP neural network classification are optimized by utilizing PSO with strong robustness and global searching ability. Extracting six main feature vectors of sandy, rocks and mud in the seabed sonar images based on the data of side scan sonar in the Zhujiang Estuary Delta, using the PSO-BP method to classify seabed sediment. The experiment shows that the accuracy of the sediments classification is more than 90%, higher than the accuracy about 70% which using BP neural network only. It proves that the PSO-BP method can be effectively applied to the identification and classification of sediment seabed.

关 键 词：基于粒子群优化算法的BP神经网络 特征向量 粒子群算法 底质分类 

分 类 号：P733.23[天文地球—物理海洋学]