基于优化遗传神经网络算法的实验室评估  

Laboratory Security Evaluation Based on Optimized Genetic Neural Network Algorithm

作　　者：陈雪芳[1] CHEN Xuefang(Experimental and Practical Management Center,Minxi Vocational and Technical College,Longyan 364021,China)

机构地区：[1]闽西职业技术学院实验实训管理中心,福建龙岩364021

出　　处：《新乡学院学报》2020年第12期47-51,共5页Journal of Xinxiang University

摘　　要：设计了一种基于优化遗传神经网络算法的评估方案。先从制度流程制定、设备管理、危险品管理、废弃物管理、人员防护及应急管理等方面出发,建立实验室安全评估指标体系;再建立神经网络模型,动态评估体系内各指标重要性权重比例,并利用遗传算法改善神经网络模型收敛速度慢、易陷入局部最优等问题;最后通过对原始指标体系的动态调整,使安全评估的结果更接近理论值。仿真结果表明:评估体系的各项功能模块运行良好,且动态寻优的收敛速度更快,训练误差与测试误差值趋近于理论值。A evaluation scheme based on the optimized genetic neural network algorithm is proposed.Firstly,the index system of laboratory safety assessment is established from the aspects of system process formulation,equipment management,hazardous material management,waste management,personnel protection and emergency management.Then,the neural network model is constructed to evaluate the importance and weight proportion of each index in the system dynamically,and genetic algorithm is used to solve the problem of slow convergence speed of neural network model,which is easy to fall into local optimization.Finally,through the dynamic adjustment to the original index system,the result of safety assessment is closer to the theoretical value.The simulation results show that the function modules of the evaluation system run well,and the convergence speed of dynamic optimization is faster,and the training error and test error value approach to the theoretical value.

关 键 词：遗传神经网络 实验室安全评估 中间隐层 激活函数 

分 类 号：G482[文化科学—教育学]