论文部分内容阅读
本文叙述了用计算机通过使某个预定的目标函数取得最小值来选择扬声器无源分频网络元件参数的设计方法。为了说明问题,所选用的目标函数为多路扬声器系统的分路系统所要求的声压—频率响应与计算值之间的差值。该目标函数的最小化是用数值最优化技术在 PDP—11上实现的。为了保证在最优化过程中所使用的扬声器系统与分频网络的总频响符合预定的精确度,在目标函数的计算中考虑到了扬声器系统的实测的声压—频率响应及其输入阻抗。在设计程序中考虑了这些因素之后,结果就使得传统的要对实际分频网络的参数进行反复试验和调整的方法无须进行了。由于目标函数是在一系列离散的频率点上进行的数值计算而得出的,因此所要求的频响可以是任意指定的。本文举了几个设计实例说明了这一方法。
This paper describes the design method of selecting parameters of speaker passive crossover network elements by making the computer obtain the minimum value of a predetermined objective function. To illustrate the problem, the objective function chosen is the difference between the required sound pressure-frequency response and the calculated value for the shunt system of a multi-speaker system. This objective function is minimized using the numerical optimization technique on the PDP-11. In order to ensure that the total frequency response of the speaker system and crossover network used in the optimization process meets the predetermined accuracy, the measured sound pressure-frequency response of the speaker system and its input impedance are taken into account in the objective function calculation. After considering these factors in the design process, the result is that the traditional method of repeatedly testing and adjusting the parameters of the actual crossover network does not need to be carried out. Since the objective function is calculated numerically at a series of discrete frequency points, the desired frequency response can be arbitrarily specified. This article cited several design examples illustrate this method.