中文摘要：

针对家庭器具优化调度中电价预测误差高和用户舒适度衡量偏差大的问题,提出了一种基于小波、微粒子群和自适应神经网络模糊推理系统（Wavelet-PSO-ANFIS,WPA）电价预测的家庭器具优化调度算法.并在该算法中针对器具运行时长不同的问题,提出了一种新颖的相对量化舒适度的方法.仿真结果表明,此电价预测方法在不牺牲计算复杂度的基础上提高了预测精度,不仅可以权衡用户的用户支付和不满意度,还可以降低系统的峰均比.

英文摘要：

Aiming at the problem of the high error of price forecasting and the low accuracy of user satisfaction in optimal scheduling of appliances, an optimal scheduling algorithm of appliances based on a novel hybrid approach is presented, combining wavelet transform, particle swarm optimization, and adaptive-network-based fuzzy inference （ Wavelet - PSO - AN FIS, WPA） system. In addition, according to appliances＇ different length of operation time, a relative quantitative method of user comfort is proposed to measure user satisfaction of different appliances. Simulation results show that the proposed electricity price forecasting method presents better forecasting accuracy with an acceptable computation time. The proposed optimal scheduling method of household appliances can benefit both users, by balancing their electricity cost and user satisfaction, and utility companies, by reducing the peak- to-average ratio.