| 摘要:针对传统RBF算法收敛速度慢，易于陷入局部极值的问题，提出了一种经优化的粒子群算法PSO，对RBF神经网络粒子群的改进参数、权值线性递减参数和标准参数进行训练寻优，构建出最优PSO-RBF神经网络，并将其用于柴油机的故障诊断预报。对MAN B&W 6L23/30H柴油机三种不同工况下第一缸试验参数的训练表明:改进的PSO-RBF神经网络在柴油机故障诊断中判别率更高，故障诊断的准确性与可靠性得到提高。 |
| 关键词: 柴油机 故障诊断 神经网络 粒子群算法 |
|
| A Diesel Engine Fault Diagnosis Method Based on Optimized PSO-RBF |
| Qin Lixin,Zhang Kai,Wang Yubao,Chen Ning |
| Eleventh Naval Representative Office in Shanghai, Shanghai 200129;Jiangsu University of Science ofTechnology,Jiangsu Zhenjiang 212000;Hudong Heavy Machinery Co., Ltd., Shanghai 200129 |
| Abstract:Aiming at the problem that the traditional RBF algorithm converges slowly and is easy to fall into local extremum, an optimized particle swarm algorithm PSO was proposed. It was used to train and optimize the improved parameters, linear weight decreasing parameters and standard parameters of RBF neural network particle swarm. The optimal PSO-RBF neural network was constructed and applied to the fault diagnosis and prediction of diesel engines. The training of the first cylinder test parameters of MAN B&W 6L23/30H diesel engine under three different working conditions was carried out. The simulation results show that the improved PSO-RBF neural network has higher discrimination rate, and improved accuracy and reliability of fault diagnosis. |
| Key words: diesel engine fault diagnosis neural network particle swarm optimization |
