Abstract: By combining with genetic regulatory networks’ nonlinear characteristic，an improved differential dynamical model based on the inertia principle is presented in this paper.The derivations demonstrate that it can work like the recurrent neural network.The experimental results for a time series gene expression data of DNA Repair networks indicate that with the help of the particle-swarm-optimization based training procedure for the parameter adjustment，the proposed model is effective.

Key words: inertia principle, gene regulatory networks, particle swarm optimization, DNA repair networks, time series gene expression data

摘要： 结合基因调控网络本身的非线性特征，提出了一个改进型的基于惯性法则的微分动力学模型，并证明其具有递归神经网络特征。使用DNA修复网络的一组时序基因表达数据进行仿真实验，实验中用粒子群优化算法优化网络参数，得到了较有意义的结果。

关键词: 惯性法则, 基因调控网络, 粒子群优化算法, DNA修复网络, 时序基因表达数据