CAJ | 학술논문

针对我国部分地区冬季采暖用电负荷较大的问题，介绍了一种既经济又可缓解电网峰谷差的蓄能供热采暖技术。考虑到BP神经网络模型在蓄能供热采暖系统负荷预测方面容易陷入局部极小点、收敛速度慢等缺点，结合BP神经网络和遗传算法( GA)的优势，提出了一种GA-BP神经网络算法。仿真试验验证了GA-BP神经网络在蓄能供热采暖系统负荷预测方面的实用性与优势。
침대아국부분지구동계채난용전부하교대적문제，개소료일충기경제우가완해전망봉곡차적축능공열채난기술。고필도BP신경망락모형재축능공열채난계통부하예측방면용역함입국부겁소점、수렴속도만등결점，결합BP신경망락화유전산법( GA)적우세，제출료일충GA-BP신경망락산법。방진시험험증료GA-BP신경망락재축능공열채난계통부하예측방면적실용성여우세。
Against the problem of large electricity load in winter for heating usage in part of the regions of our country, the energy storage heating technology that is economic and relieves peak and valley difference of the grid is introduced. Considering the BP neural network model is easily falling into local minimum point and with slow convergence speed in load forecast for energy storage heating system, so the GA-BP neural network algorithm is proposed, this algorithm combines the superiors of BP neural network and genetic algorithm (GA). Through contrast the simulation tests, the practicability and advantages of GA-BP neural network in load forecast for energy storage heating system are verified.