论文部分内容阅读
时间序列空间重构中，时间延迟参数的选择具有重要意义。本文在分析求时间延迟的自相关法和平均位移法基础上，推导出较好的求时间延迟的方法，即复自相关法。复自相关法具有强的理论依据，其计算复杂度不大，对数据长度的依赖性不强，具有优秀的抗噪能力。应用于语音信号相空间重构的实验表明，其度量可得到合适的时间延迟。
Time series space reconstruction, the choice of time delay parameters is of great significance. Based on the analysis of the autocorrelation method and the average displacement method of time delay, this paper deduces a better method to find the time delay, namely the autocorrelation method. Complex auto-correlation method has a strong theoretical basis, its computational complexity is not large, the dependence on the data length is not strong, with excellent noise immunity. Experiments applied to the phase space reconstruction of speech signals show that the appropriate time delay can be obtained for the measurement.