基于增量式极限学习机的风电场杂波抑制算法研究

A Study on Wind Farm Clutter Suppression Using Incremental Extreme Learning Machine

由于时变风力涡轮机杂波(WTC)严重影响气象雷达的探测性能,文中将极限学习机(ELM)思想引入到气象雷达风电场杂波抑制中,并针对ELM算法隐藏层节点数难以确定的问题,研究了基于增量式极限学习机(I-ELM)算法以抑制风力涡轮机杂波。首先构建训练样本选取策略,其次对模型参数进行搜索与优化,最后采用I-ELM算法自适应抑制WTC。仿真实验结果表明:该算法有效提高了WTC的抑制性能,实现了气象信号的高精度恢复。

As the time-varying wind turbine clutter(WTC)seriously affects the detection performance of weather radar,.Extreme learning machine(ELM)is introduced into wind farm clutter suppression of weather radar in this paper,and the wind farm clutter suppression algorithm based on incremental limit learning machine(I-ELM)is further proposed.The algorithm first constructs training samples,secondly searches and optimizes model parameters,and finally uses the optimized ELM algorithm to establish a prediction model to recover the meteorological signal of the contaminated distance unit.Aiming at the problem that the number of nodes in the hidden layer of the ELM algorithm is difficult to determine,the I-ELM algorithm is studied,and the network structure is modified by gradually adding hidden layer nodes,and the WTC adaptive suppression model is established.The simulation exper-iment results show that the algorithm effectively improves the suppression performance of WTC and realizes the high-precision re-covery of meteorological signals.