基于多元线性回归法的配电网导线剥皮控制方法

Distribution Network Conductor Stripping Control Method Based on Multiple Linear Regression Method

针对配电网运维检修中搭接引流线的导线剥皮问题,分析了传统剥皮控制方式的局限性,提出了一种基于多元线性回归的电机力矩控制方法。首先,分析了导线剥皮受力情况,建立了剥皮力学模型;然后,为了精准判别剥皮目标区域,引入与力矩相关的权重因子,构建多元线性回归方程并采用梯度下降法求解表征剥皮过程特征的量化因子;最后,通过对量化后的状态阈值进行判定即可实现精准剥皮控制。采用JKLYJ-70型电缆进行了剥皮实验,结果表明,所提力矩控制方法的剥皮准确率为99.2%,线损率为0.4%,验证了该方法的可行性。

A motor torque control method based on multiple linear regression is proposed to address the issue of wire stripping in the operation and maintenance of distribution networks,analyzing the limitations of traditional stripping control methods.Firstly,the force situation of wire stripping is analyzed,and a stripping mechanical model is established.Then,in order to accurately identify the target area for peeling,weight factors related to torque are introduced,and a multiple linear regression equation is constructed.The gradient descent method is used to solve the quantification factor that characterizes the peeling process characteristics.Finally,by determin⁃ing the quantified state threshold,precise peeling control can be achieved.The peeling experiment was conduct⁃ed using JKLYJ-70 cable,and the results showed that the peeling accuracy of the proposed torque control meth⁃od was 99.2%,and the line loss rate was 0.4%,verifying the feasibility of the method.