信息物理融合的智能空调响应不确定性尖峰折扣电价的动态优化模型及在线控制

A Dynamic Optimized Model and the On-line Control Strategy Response to Uncertainty PTR for the CPS of Smart Air Conditioning

降低尖峰负荷可获得巨大的经济效益,但尖峰出现时间及持续时长具有很大的不确定性。文中依托信息网络建立了一个电网运行信息、电网价格激励信息、用户室内温度喜好设置信息、外部气象信息融合的智能空调响应不确定性尖峰折扣电价的优化融合模型,采用动态优化区间的模型预测控制实现融合模型控制目标,削减其在电网尖峰期用电量,达到保证在用户定义的舒适性范围内,使用户电费支出最小,也即在电网尖峰期用电量最小的目标。采用6层120区的公共楼宇,72小时432时段,5个随机尖峰激励信息的算例表明,该融合模型及控制策略实现了平价及分时电价机制下对随机尖峰折扣电价的完全响应,在保证舒适性条件下,平价机制下电力峰值最大降幅32.7%,平均降幅25.3%;分时电价下最大降幅34.4%,平均降幅26.6%;用户的电费支出在平价机制下可节约19.9%,在分时电价机制下可节约27.6%。验证了该融合模型及控制策略的强响应能力及灵活适应性。

By existing measures, it is difficult to reduce effectively the peak of grid due to the uncertainty of occurred and lasting time. In this paper, therefore, a dynamic optimized model in response to the peak time rebates（PTR） and an on-line control strategy for the smart air conditioning were proposed. The model integrated the relevant cyber information such as the uncertain PTR in accordance with the peak of grid, weather forecast, load forecast, and setting of users＇ preference on indoor temperature for minimizing users＇ electric bill（or power consumption） under comfortable conditions. A model predictive control（MPC） improved by a dynamic interval optimization measure is used to deal with the uncertain PTR information to realize online optimal control of the air conditioning. Finally, the case with 120 zones of a 6-floor public building, 432 time frames, 72 hours, and 5 random PTR is employed to analyze the validity and flexibility. The result shows that, under comfortable conditions, the response to PTR under the time-of-use（TOU） and the flat price is done fully. The maximum range of the peak power is decreased up to 32.7% and the average is 25.3% under flat price, the maximum range of energy is decreased up to 34.4% and the average is 26.6% under TOU, the electricity bill is saved by 19.9%, 27.6% under flat price and TOU, respectively.