基于稀疏表示的火灾自动报警系统产品性能评价

摘要

性能良好的火灾自动报警系统可以有效地避免和减少火灾发生，降低火灾带来的损失。根据《火灾自动报警系统性能评价》（GB／Z 24978—2010）中提出的火灾自动报警系统的产品性能评价一级指标，制定了详细的二级指标，由于各二级指标的相互耦合性，先根据二级指标对一级指标的隶属度建立模糊关系矩阵，再对此模糊关系矩阵进行带稀疏性约束的非负矩阵分解，得出稀疏性较高的初始运算矩阵；通过对样本的二级指标进行评价并得出评价向量，据此评价向量和稀疏分解后的关系矩阵进行运算，可得出火灾自动报警系统的产品性能评价等级。%Automatic fire alarm system with good performance can effectively prevent and reduce the occurrence of fire,and reduce the fire loss.According to the FireAlarm System Performance Evaluation(GB/Z 24978—2010) which proposes the first level indicators of product performance evaluation for automatic fire alarm system,this pa-per designs detailed second level indicators.Due to the mutual coupling between the second level indicators,the pa-per establishes fuzzy relation matrix and obtains the initial operation matrix with high sparseness by non-negative matrix factorization with sparseness constraints.Then,the paper evaluates each of the second level indicators of the samples and gets the evaluation vector.Using the vector and the relationship matrix after sparseness factorization for computing,the paper retrieves the product performance evaluation grade of the automatic fire alarm system.