基于电力物联网的一种自适应能量流控制策略

An Adaptive Energy Flow Control Strategy for Power Internet of Things

文章提出一种基于动态随机最优化的自适应高效能量流控制策略(AEOEFCS)。它采用自适应高效能量流控制架构,运用贝尔曼动态方程的成本函数,以循序渐进的方式动态编程;行为网络是3层前向反馈神经网络,权值更新规则是基于反向传播算法;评测网络包括输入层、隐层、输出层;评测网络的权值更新规则是基于自适应梯度规则。自适应高效能量流控制器协调有功和无功能量流控制。仿真结果表明,相对于传统的AGC方法,AEOEFCS不仅具有较低的燃料成本和较低的系统线损,而且系统能量流更加均匀,对系统负载具有较强的适应性。

This paper proposes an adaptive efficient optimal energy flow control strategy based on dynamic random optimization theory.It adopts the control framework of adaptive efficient optimal energy flow,uses the cost function of the bellman dynamic equation,and programs dynamically in a gradual way.Its behavior network is a forward feedback neural network of three layers and the update rule of weight is based on the back propagation algorithm.Evaluating network includes input layer,hidden layer and output layer and the update rule of weight is based on the adaptive gradient rules.Controller of the adaptive efficient optimal energy flow provides the coordinate active and no functional flow control.The experimental results showed that,compared to the traditional method of AGC,AEOEFCS achieved lower fuel costs,lower system line loss,but more uniform system energy flow.It also has strong adaptability against the unpredictable power grid workloads.