【摘要】 利用电地热对居民区进行供暖时,为实现对用户室内下一时刻温度的精确预测,该文提出一种改进的自适应遗传算法(IAGA)。该算法对自适应遗传算法的交叉概率和变异概率进行改进,通过函数测试证明所提算法比传统的遗传算法稳定性好、收敛速度快,并将改进后的算法对BP网络进行优化,从而克服BP网络算法易陷入局部极值、学习效率低和收敛速度慢的缺点,最终建立基于IAGA-BP网络的电地热室内温度预测模型。将其与粒子群算法(PSO)优化的BP神经网络模型进行仿真对比,实验表明:IAGA-BP网络相对于PSO-BP网络具有更好的预测准确度,其平均绝对误差、均方差分别为0.132 8℃、0.079 2,均优于PSO-BP网络预测,该模型建立可为后期的电地热温度控制提供依据。更多还原

【Abstract】 When heating electricity to residential areas, to achieve the accurate prediction of temperature in the user’s room at the next moment, an improved adaptive genetic algorithm(IAGA) is proposed. This method improved the crossover probability and mutation probability of the adaptive genetic algorithm. The function test proves that the improved adaptive genetic algorithm has better stability and faster convergence speed than the traditional genetic algorithm. And combining the improved algorithm with the BP network, the disadvantage that the BP algorithm is easy to get into local minima and slow convergence speed can be overcome. Finally, the indoor temperature prediction model based on IAGA-BP network is established.Compared with the BP model optimized by PSO, experiments show that the IAGA-BP neural network has better prediction accuracy than PSO-BP neural network, and its average absolute error and mean square error are 0.132 8 ℃ and 0.079 2, respectively, which are better than PSO-BP network. The model can be used to provide a basis for the later thermal control of geothermal temperature.更多还原