基于BP神经网络的高层建筑防火评估

Fire Safety Assessment of High-Rise Building Based on BP Neural Network

应用BP神经网络对高层建筑进行防火评估,弥补目前国内防火评估的不足。通过对高层建筑防火系统、灭火系统、安全疏散系统、管理及其它因素进行分析,运用加动量因子和变步长的优化算法建立BP神经网络。运用Matlab7.0实现对高层建筑防火评估的仿真,仿真结果误差控制在5%以内。实例表明,该系统符合要求,具有一定的实用价值。

the application of BP neural network for high rise building fire safety assessment, make up the current domestic fire inadequate assessment of the The high-rise building fire protection systems, fire extinguishing system, safety evacua- tion system,management and other factors were analyzed, using the momentum factor and variable step size algorithm based BP neural networko The use of MatlabT. 0 to achieve the protection of high - rise building fire safety assessment of simulation,the simulation results of error control in less than 5%. Examples show that the results meet the requirements, the system, has a certain practical value.