中文摘要：

脊柱牵引系统是一种典型的非线性、时变的控制系统,为了实现牵引力的精确控制,提出了一种PID神经元网络控制算法.介绍了基于PID神经元网络算法的脊柱牵引系统的原理,分析了PID神经元网络的算法结构、权重学习规律和初值的选择.进行实际系统试验,在不同的控制条件下比较了常规PID控制和PID神经元网络控制对牵引力的控制效果,实验结果表明：采用PID神经元网络控制,牵引力的控制精度、抗干扰能力和适应参数变化的能力都优于常规PID控制,达到了较好的控制效果,证明了脊柱牵引系统和控制算法的正确性和可行性.

英文摘要：

Spinal traction system is a non-linear and time-varying control system.In order to control the traction force accurately,the PID neural network algorithm is put forward in this paper.Firstly,the principle of the spinal traction control system based on PID neural network algorithm is introduced.Then the structure and the initial values as well as the learning rule of the weights of the PID neural network are analyzed.Finally,the PID neural network was compared with conventional PID control under the different control conditions in the experiments.The experimental results showed that the PID neural network had the property of high force tracking accuracy and anti-interference ability,as well as the better adaptability to parameter change,compared with conventional PID control.The feasibility and correctness of the spinal traction system and the control algorithm were verified.