论文部分内容阅读
采用模拟退火遗传算法,分析了非等间距低通滤波器不同抽头下全正对称系数及含负系数的微波光子滤波器的频率响应特性,并与窗函数法进行了对比分析。仿真结果表明:全正对称系数条件下,带抽头数量越大,滤波器的优化性能越好。正负系数滤波器的对比表明,含负系数的滤波器优化性能明显改善,特别是对低通滤波器而言,系数的全正性极大地限制了光学传递函数的取值范围,影响了滤波器的性能。在同一截止频率下,模拟退火遗传算法优化性能指标略好于窗函数法。如果限制通带和阻带的截止频率,即放宽对过渡带的要求,模拟退火遗传算法性能优势更加明显。
The simulated annealing genetic algorithm is used to analyze the frequency response characteristics of all-symmetric and non-uniform microwave photon filters under different taps of non-equally spaced low-pass filters, and compared with the window function method. The simulation results show that under the condition of full symmetry coefficient, the larger the number of tap is, the better the performance of the filter is. The comparison of the positive and negative filter shows that the performance of the filter with negative coefficient is obviously improved. Especially for the low-pass filter, the positive factor of the coefficient greatly limits the range of the optical transfer function and affects the filtering Performance of the device. Under the same cut-off frequency, the simulated annealing genetic algorithm has better performance than the window function method. If we limit the cut-off frequency of passband and stopband, that is to relax the requirements of transitionband, the performance advantage of simulated annealing genetic algorithm is more obvious.