【摘要】 为减少高校实验室安全事故的发生和提高实验室安全风险评估的准确率，使用层次分析法(AHP)建立实验室安全风险评价体系，然后利用主成分分析法(PCA)对评价指标的综合权重进行降维，再将降维后的特征信息作为GA-BP神经网络的输入层，建立一种基于主成分分析(PCA)、遗传算法(GA)和人工神经网络(BP)相结合的实验室安全风险评价模型。实验结果表明，与BP神经网络、PCA-BP神经网络模型和GA-BP网络模型相比，PCA-GA-BP神经网络的评价精度和准确率更高、收敛速度更快、学习能力更强，可用于实验室安全风险评价。更多还原

【Abstract】 To reduce the occurrence of laboratory safety accidents in colleges and universities and improve the accuracy of laboratory safety risk assessment, this paper uses the analytic hierarchy process(AHP) to establish a laboratory safety risk assessment system. Then we use principal component analysis(PCA) to reduce the dimensionality of the comprehensive weight of the evaluation index, and finally, use the reduced dimensionality information as the input layer of the GA-BP neural network, and establish a combination of principal component analysis(PCA), genetic algorithm(GA), and artificial neural network(BP). The laboratory safety risk assessment model is completed. The verification results show that, compared with the BP neural network, the PCA-BP neural network model, and the GA-BP network model, the PCA-GA-BP neural network has a simpler structure, higher evaluation accuracy and faster convergence speed and stronger learning ability. There is no doubt that this model can be used for the safety risk assessment of university laboratories.更多还原