Papers by Author: Cong Li

Abstract: The garbage crusher is a new kind of crusher for garbage crushing when processing Municipal Solid Waste (MSW). With the development of automatic equipment and the complication of structure and properties of the garbage crusher, the fault diagnosis of garbage crusher is very important. In this paper, according to the fault symptoms and parameters, Radial Basis Function Neural Network (RBF NN) is used for fault diagnosis of the garbage crusher. The structure and inference of RBF NN are discussed in detail. The garbage crusher fault diagnosis model is established based on RBF network. At last, the fault of mechanical system is taken as an example of garbage crusher fault diagnosis. Training simulation results of the neural network are given base on MATLAB software. The result shows the RBF NN is suitable for fault diagnosis of garbage crusher.

Abstract: Shape Distribution (3DSD) and Radius Angle Histogram (RAH) are useful methods for retrieving 3D model in mechanical engineering. Through these methods have advantages such as fast speeds and simple operations, the retrieval precision are not very high enough. To improve the retrieval precision, a new method named combined histograms which integrates the advantages of 3DSD and RAH is proposed. This method makes use of the information both of shape and surface of the models to be retrieved. In the retrieval process, the shape histogram and the radius angle histogram of the retrieved model are first extracted. Then, the combined histograms of the model are established by integrating the shape histogram and the radius angle histogram. To validate the proposed method, an experiment is given. The experiment results show that the proposed method has higher retrieval precision than that of 3DSD and RAH and is suitable for mechanical model design.