滑动电接触中最小温升的计算研究

摘要

The temperature rise will change the contact condition in sliding electric contact system. In order to reduce the influ-ence of the temperature rise on contact performance, the temperature rise should be ensured at the least value. Some experiments have been done to get the relationship of data changing about the temperature rise using the sliding electrical contact testing machine. In these data, the temperature rise is output while the speed, contact current and load are inputs. The neural network is used to fit the model of the temperature rise. Since there is no exact function between the input and output, this model can be called black box model. Based on this model, after analyzing the least temperature rise, an optimization has been gotten, which can be solved by the Particle Swarm Optimization algorithm to fix the best contact load. The minimal temperature rises are given under some running conditions, and the result shows the validity of the method.%　　在滑动电接触中，接触面温升会改变接触环境，为了降低温升对接触性能的影响，应保证接触温升最小。采用滑动电接触实验机，对其温升系统进行了实验研究，得到了以温升为输出，运行速度、接触载流以及载荷为输入的数据变化关系。采用神经网络拟合出数据所表现出的模型关系，得到从输入映射到输出变量的黑箱模型。在此基础上，通过对最小温升问题的分析，将其转化为最优问题的求解，进而采用粒子群优化算法实现最优载荷的确定，给出了部分运行工况下最小温升，结果表明了设计方法的有效性。