混沌粒子群优化神经网络的井下人员无线定位方法研究

针对井下人员定位系统定位精度较低,不能满足智慧煤矿的需求,提出一种基于混沌粒子群算法优化Elman神经网络的井下人员无线定位方法。该定位方法首先在井下巷道无线网络环境中,利用无线终端采集一定数量的样本点指纹数据库。其次初始化Elman神经网络,利用混沌粒子群优化算法对神经网络权值和自连接反馈增益因子寻优。再次用指纹数据库对优化过的Elman神经网络进行训练和测试,建立神经网络定位算法模型。最后通过无线终端采集定位点的指纹数据,由神经网络定位算法模型进行实时定位。经试验表明,该井下人员无线定位方法平均定位误差为1.35 m;而混沌粒子群算法优化Elman神经网络定位算法,其算法全局搜索能力更强,更适合井下时变环境中应用。

Wireless Position Method for Underground Personnel Based on Neural Network by Chaos Particle Swarm Optimization

The positioning accuracy of underground personnel positioning system is low,which can not meet the needs of intelligent coal mine.Therefore,a wireless positioning method for underground personnel based on Elman neural network by means of chaos particle swarm optimization algorithm was proposed.Firstly,a certain number of sample fingerprint databases were collected by wireless terminals in the wireless network on underground roadway.Secondly,Elman neural network was initialized and chaotic particle swarm optimization algorithm was used to optimize the weights and self-connected feedback gain factor of the neural network.Thirdly,the fingerprint database was used to train and test the optimized Elman neural network,and the position algorithm model of the neural network was established.Finally,the real-time position was achieved with the neural network location algorithm model through the fingerprint data of the position point collected by wireless terminal.The test shows that the average position error of this underground wireless positioning method is 1.35 m;Elman neural network position algorithm by the chaos particle swarm optimization has stronger global search ability and is more suitable for the application in time-varying underground environment.