【摘要】 在分析发动机结构参数和运转参数对发动机性能影响的基础上,提出了一种基于组合式神经网络的柴油机性能状态评估预测模型.该模型首先运用动态聚类法将大样本分成若干小组,然后分别用于子网络训练.性能评估时,运用模糊识别法选择相关的子网络进行评估分析.实例验证表明,这种模型能有效解决大样本下神经网络训练速度慢和难以收敛的问题,提高柴油机性能评估预测精度.更多还原

【Abstract】 By analyzing the relationship between technical states and performance parameters of diesel engine, an appraisal model based on modular neural networks is proposed. Firstly, the original samples are divided into several groups by applying the dynamic clustering method. Then each group is used to train a separate seed neural networks. During the appraisal phase, pattern recognition is employed to activate the corresponding seed neural networks for appraisal analysis. Experiment indicates that the model can solve the problem of low learning speed and difficult convergence and improve the accuracy of prediction.更多还原