通过反复训练模板能有较高的识别率，正确率可以达到98％，包含CV、CA、Single、当前、恒转弯速率、转弯模型，阐述了负荷预测的应用研究,包括主成分分析、因子分析、贝叶斯分析，添加噪声处理，pwm整流器的建模仿真。- Through repeated training koFEEeClate have higher recognition rate, Accuracy can reach 98 , It contains CV, CA, Single, current, constant turn rate, turning model, It describes the application of load forecasting, Including principal component analysis, factor analysis, Bayesian analysis, Add noise processing, Modeling and simulation pwm rectifie.