基于改进粒子群优化算法的电力线通信多径传输模型参数辨识

基于已有的电力线多径传输模型结构，以0.5~20 MHz范围内的实际低压载波通信信道测量数据为样本，将改进粒子群优化算法应用于低压载波通信信道模型的多参数辨识，通过自适应改变惯性权重提高搜索效率，同时采用模拟退火算法并自适应调整退火温度，克服了基本粒子群算法容易发生早熟收敛的缺点。对路径数为4和18的信道模型进行参数辨识的结果表明：与遗传算法相比，改进的粒子群算法可加速收敛，缩短辨识时间，同时提高了拟合精度，克服了参数的分散性，所取路径数越多，拟合效果越好。

Parameter Identification of Multi-Path Transmission Model for Power Line Communication Based on Improved Particle Swarm Optimization

Based on existing structures of multi-path transmission model for power line and taking the measured data of practical low voltage power line carrier communication within the range form 0.5MHz to 20MHz as samples, the improved particle swarm optimization (PSO) is applied to the multi-parameter identification of channel model for low voltage carrier communication. By means of adaptively changing the inertia weight the searching efficiency is improved, meanwhile, by use of simulated annealing algorithm and adjusting the annealing temperature adaptively, the shortcoming of easy to occur premature convergence in basic PSO is overcome. The parameter identification results of 4-channel and 18-channel models show that by use of the improved PSO the convergence speed is faster than by genetic algorithm (GA) and the time for the identification is saved, the fitting accuracy is improved as well as the dispersivity of parameters is overcome. The more number of paths being taken, the better the fitting result will be.