基于Volterra级数的功放行为建模与简化

Pruning the Volterra series for behavioral modeling of power amplifiers

Volterra级数行为模型因其高复杂度而只适用于弱非线性系统,对于具有强非线性和长记忆效应的功率放大器,Volterra级数模型将包含大量的待辨识参数.对射频功率放大器的行为建模进行了研究,提出了功率放大器Volterra级数行为模型的简化方法,通过引入邻对角核控制因子和动态偏离阶数控制因子,合理移除了对信号输出影响较少甚至无关的辨识参数,从而在保证模型准确度的前提下有效地降低了模型的复杂度和参数提取时间.采用了一个10 W的射频功率放大器实例用于模型提取和验证,模型与仿真实验的数据结果对比表明,简化的Volterra行为模型能够很好地描述功放的非线性特性及记忆效应,验证了简化行为模型的有效性和灵活性.

： General Volterra series based models are limited to weakly nonlinear systems due to their high complexity. It involves a great number of coefficients for behavioral modeling of power amplifier with strong nonlinearities and long memory effects. The behavioral modeling of radio frequency （RF） power amplifiers was studied. A pruned Volterra series based model was proposed by introducing a ＇ near-diagonality＇ controlling factor and a ＇ dynamic deviation reduction＇ controlling factor simultaneously to remove the coefficients which are not sensitive to the output error. The pruning approach could effectively reduce the model complexity and save time for parameters extraction under the condition of the model fidelity required. A 10 W RF power am- plifier was employed for providing input and output data which were used in model extraction and validation. The agreement obtained between the output response of simulated circuit and model validates the pruned model is flexible and effective.