摘要
飞参数据压缩是减少飞参数据的存储空间和传输通信流量的关键。针对飞参数据的特点,提出了一种基于粒子群优化的小波神经网络近无损压缩算法。该算法将小波网络参数作为原始数据的重构信息,在小波神经网络BP算法的基础上,引入粒子群优化算法,克服了粒子群优化算法的早熟收敛,增强了小波神经网络学习算法的全局搜索能力,提高了网络收敛速度;同时将重构误差作为启发信息,在保证较小失真度的情况下,通过粒子的迭代寻求最优的小波神经网络结构。飞参数据压缩仿真实验结果表明了算法的可行性和有效性,可以获得较高的压缩比和较小的重构误差。
The compression of flight data is the key technology for reducing storage space and communication flow in transmission. Considering the characteristics of flight data, a novel near-lossless data compression algorithm was proposed by using the wavelet neural network based on particle swarm optimization. In this algorithm, the parameters of wavelet network were regarded as the reconstruction information of original data. On the basis of BP algorithm in wavelet network, the particle swarm optimization was introduced to improve global search ability and convergence speed of wavelet network learning algorithm, and overcome the premature problem. At the same time, reconstruction error was taken as the heuristic information for searching the best framework of wavelet network through particle iteration on the premise of keeping low degree of distortion. The experiment results of flight data compression showed that the algorithm is feasible and effective, which can obtain better compression performance.
出处
《电光与控制》
北大核心
2010年第3期64-67,共4页
Electronics Optics & Control
关键词
飞行参数
近无损压缩
粒子群优化
小波网络
flight data
near-lossless compression
particle swarm optimization
wavelet network