分形维是混沌信号的一个重要参数 ,在确定分形维时 ,通常需要确定线性尺度区 ,然而线性尺度区的确定仍然是分形维提取算法中的一个富有挑战性的问题 ,线性尺度区的准确性 ,将直接影响分形维的准确性 .本文对此问题进行了比较深入的研究 ,较好地解决了线性尺度区的选择问题 ,并基于GP算法给出了一套完整的相关维提取算法 .该算法不仅运算量小、精度高、抗噪声干扰的能力较强 ,而且完全实现了对相关维的自动提取 Fractal dimension is an important parameter of chaotic signal. In determining the fractal dimension, it is usually necessary to determine the linear scale area. However, the determination of the linear scale area is still a challenging problem in the fractal dimension extraction algorithm. The accuracy of the linear scale area , Will directly affect the accuracy of fractal dimension.In this paper, a more in-depth study of this issue, a better solution to the problem of linear scale area selection, and based on the GP algorithm gives a complete set of related dimension extraction algorithm. The algorithm is not only a small amount of computation, high precision, strong ability to resist noise interference, but also fully realized the automatic extraction of related dimensions