基于改进SPSO-BP神经网络的温度传感器湿度补偿

摘要

In order to solve the problem that the humidity affects temperature sensor accuracy,this paper derives the simplified particle swarm algorithm by analyzing the basic particle swarm algorithm.The simplified particle swarm al-gorithmis not affected by the velocity vectorand the linear decreasing inertia weight is adopted.The humidity com-pensation method which using improved simplified particle swarm optimization algorithm(SPSO)of BP neural net-work is proposed.The weight threshold of the BP neural network is optimized by the continuous iteration of the im-proved simplified particle swarm optimization algorithm,until the optimal threshold is obtained and assigned to the BP neural network.According to the data measured in the experiment of humidity influence,the humidity compensa-tion model is established by this method and compared with the BP neural network. The results show that the im-proved SPSO-BP neural network model has simple structure,high compensation precision and fast convergence speed,and effectively compensates the humidity of the temperature sensor.%针对在实际使用中湿度影响温度传感器准确性的问题,通过对基本粒子群算法的分析,得出不受速度向量影响的简化粒子群算法,同时采用线性递减惯性权重,提出了一种改进SPSO-BP神经网络温度传感器的湿度补偿方法.通过改进的简化粒子群算法的不断迭代,优化BP神经网络的权阈值,直到得到最优权阈值,并赋给BP神经网络.根据湿度影响实验中测得的数据,运用此方法建立湿度补偿模型,与BP神经网络方法对比分析.结果表明,改进SPSO-BP神经网络的模型结构简单、补偿精度高,收敛速度快,有效地对温度传感器进行了湿度补偿.