CN-121980900-A - Method and system for quantifying adjustable margin of building by coupling data prediction and physical simulation

CN121980900ACN 121980900 ACN121980900 ACN 121980900ACN-121980900-A

Abstract

The invention relates to a method and a system for quantifying adjustable margin of a building, wherein the method comprises the following steps of obtaining load data and air image data of a building to be processed, and preprocessing; according to the preprocessed load data and the preprocessed meteorological data, a load prediction model is adopted to obtain a load prediction result, a predicted load curve is constructed, a building energy consumption simulation model and a similar day screening model are built to obtain a basic scene curve, and aiming at the predicted load curve and the basic scene curve, an adjustable margin calculation model and a response duration calculation model are adopted to solve and obtain an adjustable margin and a response duration. Compared with the prior art, the method can fill the blank of the existing research on volume-time two-dimensional evaluation, improve the accuracy of building participation demand response, and realize win-win of peak regulation and valley filling of the power grid and reduction of building energy cost.

Inventors

WANG YINCHAO
GU WEN
JIANG BENJIAN
SHEN HUI
ZHANG LITING

Assignees

国网上海市电力公司

Dates

Publication Date: 20260505
Application Date: 20251205

Claims (10)

1. The method for quantifying the adjustable margin of the building by coupling data prediction and physical simulation is characterized by comprising the following steps of: Load data and meteorological data of a building to be processed are obtained, and pretreatment is carried out; according to the preprocessed load data and the preprocessed meteorological data, a load prediction model is adopted to obtain a load prediction result, a predicted load curve is constructed, a building energy consumption simulation model and a similar day screening model are built, and a basic scene curve is obtained; and solving and obtaining the adjustable margin and the response duration by an adjustable margin calculation model and a response duration calculation model aiming at the predicted load curve and the basic scene curve.
2. The method for quantifying a building tunable margin coupled with data prediction and physical simulation according to claim 1, wherein the preprocessing comprises missing value repair, outlier processing, data smoothing and normalization.
3. The method for quantifying the building adjustable margin coupled with data prediction and physical simulation according to claim 1, wherein the load prediction model is a PSO-LSTM-RF combined load prediction model, the PSO-LSTM-RF combined load prediction model captures a time sequence dependency relationship through the LSTM model to obtain a time sequence characteristic, the time sequence characteristic and the received static characteristic are fused through the RF model to conduct integrated regression prediction, and the optimal super-parameters are screened through a PSO algorithm to output a load prediction result.
4. The method for quantifying an architectural tunable margin coupled with data prediction and physical simulation according to claim 1, wherein the expression of the tunable margin calculation model is as follows: Wherein, the Load adjustable margin of the ith class of response scene in the kth time step is provided, wherein k is a time step index; Load values for the kth time step under the base scene graph; and predicting the load value of the ith class of response scene in the load curve at the kth time step.
5. The method for quantifying a building tunable margin coupled with data prediction and physical simulation of claim 1, wherein the expression of the response duration calculation model is as follows: Wherein, the Response duration for the class i response scenario; the method comprises the steps of predicting preset closing time of an air conditioner in an ith type response scene in a load curve; and predicting the actual restarting moment of the air conditioner in the i-th type response scene in the load curve.
6. The method for quantifying a building tunable margin coupled with data prediction and physical simulation of claim 1, wherein the response scenario comprises: In a basic scene, the air conditioner operates at 8:00-18:00, and the target room temperature is 24 ℃; The first response scene is that the air conditioner operates at 8:00-10:00, the target room temperature is 24 ℃, and the restarting condition is that the room temperature rises to 28 ℃; The second response scene is that the air conditioner operates at 8:00-12:00, the target room temperature is 24 ℃, and the restarting condition is that the room temperature rises to 28 ℃; in a third response scene, the air conditioner operates at 8:00-14:00, the target room temperature is 24 ℃, and the restarting condition is that the room temperature rises to 28 ℃; in a fourth response scenario, the air conditioner operates at 8:00-16:00, the target room temperature is 24 ℃, and the restarting condition is that the room temperature rises to 28 ℃.
7. The method for quantifying a building adjustable margin coupled with data prediction and physical simulation according to claim 1, wherein the similarity day screening model quantitatively calculates similarities between a prediction day and a base day through a plurality of indexes, thereby obtaining a base scene curve corresponding to a predicted load curve; the plurality of metrics includes RMSE, MAE, pearson correlation coefficients and dynamic time warping distances.
8. The method for quantifying adjustable margins of buildings coupled with data prediction and physical simulation according to claim 1, further comprising a simulation verification process, wherein the corresponding adjustable margins and response durations are solved in a simulation manner by collecting load data and weather data of offices, shops and hotels.
9. A system for quantifying a building tunable margin according to a method for quantifying a building tunable margin coupled with a data prediction and a physical simulation according to any one of claims 1 to 8, comprising: The data standard layer is used for acquiring load data and meteorological data of the building to be processed and preprocessing the load data and the meteorological data; the model cooperative layer is used for acquiring a load prediction result by adopting a load prediction model according to the preprocessed load data and the preprocessed meteorological data, constructing a predicted load curve, and establishing a building energy consumption simulation model and a similar day screening model to acquire a basic scene curve; The margin evaluation layer is used for solving and obtaining the adjustable margin and the response duration through the adjustable margin calculation model and the response duration calculation model aiming at the predicted load curve and the basic scene curve.
10. A system for quantifying building tunable margin coupled with data prediction and physical simulation, characterized in that it comprises a memory and a processor, said memory storing a computer program, said processor invoking said computer program to perform the steps of the method according to any of claims 1 to 9.

Description

Method and system for quantifying adjustable margin of building by coupling data prediction and physical simulation Technical Field The invention relates to the technical field of building adjustable margin quantization, in particular to a method and a system for quantizing a building adjustable margin by coupling data prediction and physical simulation. Background Along with the aggravation of global energy crisis and climate change problems, building energy conservation becomes a key link of sustainable development, the global building industry energy consumption accounts for 32% of the total energy consumption, and the Chinese building energy consumption ratio is still continuously rising. Meanwhile, the construction of a novel power system promotes the grid connection of high-proportion renewable energy sources, the intermittent performance and the fluctuation of the novel power system provide serious challenges for the supply and demand balance of a power grid, a building serves as an important adjustable load resource, and the energy adjustable margin of the novel power system becomes a core tie of the power grid-building collaborative optimization. The existing research is focused on potential quantification of macroscopic layers such as flexible regulation potential, demand response potential and the like, and has two major core limitations, namely firstly, visual angle deviation is evaluated, an adjustable margin (energy dimension) and response duration (time dimension) are not taken as coupling indexes to be evaluated in a special mode, the power grid quantity-time two-dimensional dispatching requirement is difficult to meet, and secondly, a technical chain is broken, or an imitation straight tool is relied on to lack accurate prediction support, or a emphasis prediction model and margin quantification do not form a closed loop. Therefore, a systematic method for integrating data prediction and physical simulation is needed to realize accurate quantification of building adjustable margin. Disclosure of Invention The invention aims to overcome the defects of single evaluation dimension and incomplete technical chain in the prior art and provide a method and a system for quantifying the adjustable margin of a building by coupling data prediction and physical simulation. The aim of the invention can be achieved by the following technical scheme: a method for quantifying the adjustable margin of a building by coupling data prediction and physical simulation comprises the following steps: Load data and meteorological data of a building to be processed are obtained, and pretreatment is carried out; according to the preprocessed load data and the preprocessed meteorological data, a load prediction model is adopted to obtain a load prediction result, a predicted load curve is constructed, a building energy consumption simulation model and a similar day screening model are built, and a basic scene curve is obtained; and solving and obtaining the adjustable margin and the response duration by an adjustable margin calculation model and a response duration calculation model aiming at the predicted load curve and the basic scene curve. Further, the preprocessing comprises missing value repair, outlier processing, data smoothing and standardization. Further, the load prediction model is a PSO-LSTM-RF combined load prediction model, the PSO-LSTM-RF combined load prediction model captures a time sequence dependency relationship through the LSTM model to obtain a time sequence characteristic, integrated regression prediction is carried out through fusion of the time sequence characteristic and the received static characteristic of the RF model, and the optimal super-parameters are screened through a PSO algorithm to output a load prediction result. Further, the expression of the adjustable margin calculation model is as follows: Wherein, the Load adjustable margin of the ith class of response scene in the kth time step is provided, wherein k is a time step index; Load values for the kth time step under the base scene graph; and predicting the load value of the ith class of response scene in the load curve at the kth time step. Further, the expression of the response duration calculation model is as follows: Wherein, the Response duration for the class i response scenario; the method comprises the steps of predicting preset closing time of an air conditioner in an ith type response scene in a load curve; and predicting the actual restarting moment of the air conditioner in the i-th type response scene in the load curve. Further, the response scenario includes: In a basic scene, the air conditioner operates at 8:00-18:00, and the target room temperature is 24 ℃; The first response scene is that the air conditioner operates at 8:00-10:00, the target room temperature is 24 ℃, and the restarting condition is that the room temperature rises to 28 ℃; The second response scene is that the air conditioner operates at 8