基于WSN的回油温度神经网络PID控制方法

PID Neural Network Control Method for Return Oil Temperature Based on WSN Technique

为了解决油田计量间回油温度控制中多热水管路流量耦合、控制响应高时滞的问题,引入了一种PID神经网络控制方法,该方法基于无线传感器网络(WSN)技术实现,所设计的PID神经网络控制算法能够在低成本微控制单元(MCU)上实时运行。该PID神经网络由3个并联的子网络组成,每个子网络隐含层中包含3个神经元,用以模拟PID控制中的比例、积分、微分作用。仿真分析结果标明,在3路回油热水管路强耦合情况下,针对3路不同给定控制量,均能起到较好控制效果。

In order to solve the problem of multi-hot water pipeline flow coupling and high time-delay control response in oil return temperature control of metering room in oil field, a PID neural network control method is introduced. The method is based on wireless sensor network （WSN） technology. The designed PID neural network control algorithm can run in real time in low-cost micro-control unit （MCU）. The PID neural network consists of three parallel sub-networks, each of which contains three neurons in the hidden layer to simulate the proportional, integral and differential functions of PID control. The simulation results show that under the strong coupling condition of the three return oil hot water pipelines, a better control effect can be played for the three different given control variables.