A Safety Assessment Method for Gantry Cranes Based on PLS Probabilistic Neural Network
A probabilistic neural network(PNN)based partial least squares(PLS)was proposed to address the problems of com-plex network structure,numerous parameters,and time-consuming computations in traditional probabilistic neural networks for gantry crane safety assessment.The partial least squares method was used for principal component extraction,partial principal components were used to replace the original sample input,thereby reducing the dimensionality of the input.The typical sample neurons was replaced by the limited pattern combination neurons to reduce the parameters of network such as nodes and connections,simplify the network,and accelerate its convergence speed.The output results were discretized to improve the training efficiency and inference speed of the model.The results of simulation tests show that the proposed method significantly reduces training and testing time compared to the traditional PNN evaluation method,thereby verifying the effectiveness of the method and achieving efficient and intelligent safety assessment of gantry cranes.
gantry craneprobabilistic neural network(PNN)partial least squares(PLS)residual matrix
NETL
NSTL
万方数据
