The Fault Diagnosis Techniques For The Concrete Pump Truck

Concrete pump truck is one of the complex equipment which is high technical content、huge difficult maintenance and more expensive in engineering machinery. It has important significance to diagnose the pump fault with real-time、remote and online through using the advanced things technology and artificial intelligence fault diagnostic technology in ensuring the technical security of the key industries large equipment operatoring, achieving the energy saving and environmental protection, and improving the manufacturing ability of the sustainable development.In this paper, in consideration of the current pump hydraulic system fault diagnostic problem, the pump hydraulic fault diagnostic system based on Internet of Things is proposed; In the case of considering to improve the real-time and reliability of the system and reduce cost of the system comprehensively, the paper completes the establishing of the security vehicle terminal system which is facing to the Internet of Things structures, what’s more,the paper studies the intelligent diagnostic algorithm which are suitable for the paper background and uses the intelligent diagnostic technology as the theoretical basis, the intelligent information processing method include artificial neural networks, particle swarm optimization algorithm, D-S evidence theory, multi-sensor information fusion theory and fuzzy vector machines are introduced into the pump hydraulic system fault diagnosis. The control subsystem and power subsystem key parts of the hydraulic system of the pump are respectively diagnosed. The main innovations of this paper are as follows:(1) The paper proposes a fault diagnostic system solutions of hydraulic system of the pump based on Internet of things, taking the terminal platform security into considered, the TPM chip is added to the vehicle terminal in order to make the vehicle terminal have higher security.(2) This paper presents a fault diagnosis method based on PSO-Elman neural network.Through improving the inertia weight and learning factors of PSO algorithm and applying it to the training learning of Elman neural network, the Elman neural network gets better in terms of training time, the convergence rate and diagnostic accuracy; this paper puts forward a fault diagnosis method based on PSO-H-BP neural network, the method combinate the method of PSO algorithm、 Hopfield neural network and BP neural network. First, using the PSO algorithm to optimize the Hopfield network to obtain the stable network structure, and then using the Hopfield network to preprocess the input data of BP network, finally, using the BP neural network to diagnose the fault, this method has improved the speed of convergence and diagnostic accuracy of the network. Finally, the effectiveness of the algorithm is verified by experiments.(3) The paper presents a fault diagnosis method based on double-layer FSVM model structure, and the method is applied to the hydraulic system solenoid valve fault diagnosis, it achieves good results; Additionally,in this method, the training algorithms and parameters selection method of fuzzy support vector machine are optimized, the experimental results show that learning performance of the support vector machine can significantly improve by using the optimized parameters.(4) The paper puts forward a fault diagnosis method based on the information fusion model with three fusion levels,it adopts multiple PSO-BP and MPSO-RBF neural network to compose the vibration subnet and temperature subnet which is make used of local diagnosis; According to the study background the multi-sensor information fusion method based on the modified D-S evidence theory is proposed and the paper puts forward a decisions method based the pl&bl; finally, the three levels multi-source information fusion fault diagnostic method is used in the key parts of the hydraulic system power subsystem, this study has been verified.