基于蚁群神经网络的泵车主泵轴承性能评估

针对BP神经网络、遗传神经网络等智能算法在机械设备关键部件的性能评估过程中训练收敛速度慢，且会遇到局部极小的问题，提出一种运用蚁群算法训练神经网络的权值和阈值的混合智能算法--蚁群神经网络.将蚁群神经网络应用于混凝土泵车主泵系统中主泵轴承的模式识别和性能评估.结果表明，蚁群神经网络能很好地解决收敛速度慢、局部极小的问题，提高了分类精度，展现了良好的应用前景.

State Performance Evaluation for the Main Pump Bearing of Pump Truck Based on Ant Colony Optimization of Neural Network

During the process of the state performance evaluation for the key components of mechanical equipment, the convergence speed of BP neural network, genetic neural networks and other hybrid intelligence algorithm is slow and may inevitably meet local minimal problems. According to these problems, a kind of hybrid intelligence algorithm was proposed which combines the global optimization characteristics of ant colony optimization (ACO) and the innings optimization ability of BP neural network. And it is applied in the state performance evaluation for the main pump bearing of pump truck. According to the application results, the ant colony neural network can solve the slowly convergence speed, local minimal problems very well and the accuracy of classification is improved which also reflects the good application prospect.