论文部分内容阅读
针对航空电机发热及散热受电机功率、结构,以及由于海拔高度改变带来的大气温度、粘度、压力变化等众多因素影响,温升模型难以准确建立的问题,通过已有试验数据,建立起遗传算法-支持向量机表面温升模型,解决了遗传算法局部搜索能力差的问题,有效地利用支持向量机学习速度快的特点,在小样本情况下具有良好的非线性建模和泛化能力。基于LS-SVM的预测控制算法具有很好的控制性能。试验表明,该模型实现了对航空电机表面温升的智能预测。由于支持向量机具有自学习功能,可在应用中不断提高预测精度,因而这种方法在电机设计中具有广阔的应用前景。
In view of the fact that the heating and cooling of aero-electric motor are affected by many factors such as the power and structure of the motor and the atmospheric temperature, viscosity and pressure caused by the altitude change, it is difficult to accurately establish the temperature rise model. Based on the existing experimental data, This algorithm solves the problem of poor local search ability of genetic algorithm, and makes full use of the characteristics of support vector machine (SVM) learning speed. It has good nonlinear modeling and generalization ability in the case of small samples. Predictive control algorithm based on LS-SVM has good control performance. Experiments show that the model achieves intelligent prediction of the surface temperature of aeronautic motor. Because SVM has the function of self-learning, it can improve the prediction precision constantly in the application, so this method has broad application prospect in the motor design.