Airport terminal dynamic cooling load prediction method and device and program product

Abstract

The invention discloses an airport terminal dynamic cooling load prediction method and device and a program product, and the method comprises the following steps: S1, constructing a dynamic cooling load prediction neural network model, learning the corresponding relationship between the cold load of the airport terminal at the t moment and the cold load of the t-4-t-1 time period, and the meteorological parameter sequence and the arrival and departure flight shift sequence of the t-4-t time period from the historical sample data; and S2, collecting a meteorological parameter sequence of a t-1-t time period, an arrival and departure flight shift sequence and a cooling load of a t-4-t-1 time period, inputting the collected data into the dynamic cooling load prediction neural network model, and outputting an airport terminal cooling load prediction value of the t time period. According to the invention, dynamic cooling load prediction of the airport terminal is realized.

Description

Technical Field

[0001] The invention relates to a method, a device and a computer program product for predicting dynamic cooling load of an airport terminal. Background Art

[0002] Due to complex factors such as flight restrictions, severe weather, and epidemic prevention and control, the actual demand for cooling load in airport terminals fluctuates frequently and changes in a complex manner. How to accurately predict the dynamic changes in terminal cooling load, obtain the optimal control strategy for the air-conditioning system, respond to cooling load demand in a timely manner, create a healthy and comfortable terminal thermal environment, and ensure the low-carbon and economical operation of the air-conditioning system has become a challenge. Summary of the invention

[0003] The object of the present invention is to provide a method, device and program product for dynamic cooling load prediction of an airport terminal building, so as to timely respond to the cooling load demand of the airport terminal building.

[0004] To this end, the present invention provides a method for dynamic cooling load prediction of an airport terminal, comprising the following steps: S1, constructing a dynamic cooling load prediction neural network model, and learning from historical sample data the correspondence between the cooling load of the airport terminal at time t and the cooling load in the period t-4-t-1, as well as the meteorological parameter sequence and the inbound and outbound flight sequence in the period t-4~t; S2, collecting the meteorological parameter sequence, inbound and outbound flight sequence, and the cooling load in the period t-4-t-1 in the period t-1~t, and inputting them into the dynamic cooling load prediction neural network model, and outputting the predicted value of the cooling load of the airport terminal at time t.

[0005] Furthermore, the above-mentioned meteorological parameter sequence collected during the period t-4 to t includes real data of outdoor air temperature, relative humidity, solar radiation, and wet-bulb temperature collected during the period t-4 to t-1 and predicted data of outdoor air temperature, relative humidity, solar radiation, and wet-bulb temperature at time t from the weather forecast.

[0006] Furthermore, the above-mentioned dynamic cooling load prediction modeling adopts a TCN-Transformer deep learning model, including an input layer, a TCN network, a Transformer network, a fully connected layer and an output layer, wherein the TCN network model includes a void causal convolution, an activation function, and a normalization layer, and the Transformer network includes a masked self-attention mechanism layer and a feedforward layer.

[0007] Furthermore, the above-mentioned various collected data are normalized before being input into the dynamic cooling load prediction modeling.

[0008] The present invention also provides a computer program product, including a computer program, which is used to implement the steps of the above-mentioned airport terminal dynamic cooling load prediction method when executed.

[0009] The present invention also provides a device for predicting dynamic cooling load of an airport terminal, which stores a computer program, and when the program is executed, it is used to implement each step of the above-mentioned method for predicting dynamic cooling load of an airport terminal.

[0010] Compared with the prior art, the present invention has the following technical effects: The present invention proposes for the first time a dynamic cooling load prediction method for airport terminals based on a multi-time series input and single-variable output deep learning regression model. Short-term outdoor weather forecasts and planned inbound and outbound flight time series are selected as inputs. Based on a deep learning algorithm, a multi-time series input and single-variable output regression prediction model is trained and established for dynamic cooling load prediction of airport terminals.

[0011] In addition to the above-described purposes, features and advantages, the present invention has other purposes, features and advantages. The present invention will be further described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The drawings constituting a part of the present application are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings: Figure 1 It is a structural diagram of the prediction model in the airport terminal building dynamic cooling load prediction method of the present invention; Figure 2 yes Figure 1 Modeling flowchart of the prediction model shown; Figure 3 It is an architecture diagram of the TCN-Transformer deep learning model of the present invention; Figure 4 This is a schematic diagram of hourly data of the terminal building’s cooling load; Figure 5 It is a schematic diagram of outdoor temperature hourly data; Figure 6 It is a schematic diagram of outdoor temperature hourly data; Figure 7 It is a schematic diagram of hourly data of outdoor relative humidity; Figure 8 It is a schematic diagram of outdoor wet-bulb temperature hourly data; Fig. 9 It is a schematic diagram of the hourly data of inbound and outbound flights; Fig.10 This is a comparison chart of the experimental results of the prediction method of the present invention and other methods in May 2023; Fig.11 This is a comparison chart of the experimental results of the prediction method of the present invention and other methods in June 2023; Fig.12This is a comparison chart of the experimental results of the prediction method of the present invention and other methods in July 2023. DETAILED DESCRIPTION

[0013] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0014] The airport terminal building dynamic cooling load prediction method of the present invention is a method for predicting the airport terminal building dynamic cooling load based on a multi-time series input single variable output deep learning regression model. Figures 1 to 12 , the prediction method includes the following steps S1-S2.

[0015] S1. Construct a dynamic cooling load prediction neural network model, and learn the correspondence between the cooling load of the airport terminal at time t and the cooling load in the period t-4-t-1, as well as the meteorological parameter sequence and the inbound and outbound flight sequence in the period t-4~t from historical sample data; S2. Collect the meteorological parameter sequence, inbound and outbound flight sequence, and cooling load in the period t-4-t-1 from t-1~t, input them into the dynamic cooling load prediction neural network model, and output the predicted value of the airport terminal cooling load at time t.

[0016] In the present invention, short-term outdoor weather forecasts and planned inbound and outbound flight time series are selected as inputs, and based on a deep learning algorithm, a multi-time series input single variable output regression prediction model is trained and established for dynamic cooling load prediction of airport terminals.

[0017] Dynamic cooling load prediction model for airport terminal, such as Figure 1 shown.

[0018] Through Pearson coefficient correlation analysis, five parameters related to dynamic cooling load prediction are preferably selected: average daily outdoor temperature, average daily relative humidity, average daily solar radiation, average daily wet-bulb temperature, and the total number of daily inbound and outbound flights.

[0019] Specifically, the model selects the time series of outdoor air temperature (T), relative humidity (RH), solar radiation (SR), wet bulb temperature (WT), number of inbound and outbound flights (FN) from t-4 to t, and the time series of terminal cooling load (q) from t-4 to t-1 as input, and the dynamic cooling load prediction value of the airport terminal in period t as output.

[0020] The dynamic cooling load prediction model of the airport terminal uses the TCN-Transformer deep learning model. Figure 3As shown in the figure, it includes input layer, TCN network, Transformer network, fully connected layer and output layer. The TCN network model includes hole causal convolution, activation function and normalization layer. The Transformer network model includes masked self-attention mechanism layer and feedforward layer. Example

[0021] Taking a large airport terminal in a cold region of my country as an example, the building area of ​​the airport terminal is about 70,000 square meters. The daily cooling load, outdoor weather and historical data of inbound and outbound flights of the terminal during the summer operation of air conditioning in 2023 (the horizontal axis unit is half an hour) are collected, and the TCN-Transformer deep learning algorithm is used to train and establish a dynamic cooling load prediction model for the terminal. Among them, the outdoor meteorological data includes temperature, relative humidity, solar radiation and wet bulb temperature. The collected data are as follows Figures 5 to 9 shown.

[0022] The prediction method proposed in this invention is used to carry out the dynamic cooling load prediction of the airport terminal. The prediction experimental results are as follows: Figure 10-12 As shown, the coefficient of determination (R2) of the prediction results is 0.9368.

[0023] The above description is only an embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for predicting dynamic cooling load of an airport terminal, characterized in that: The following steps are involved: S1. Build a dynamic cooling load prediction neural network model to learn the correspondence between the cooling load of the airport terminal at time t and the cooling load during the period t-4-t-1, as well as the meteorological parameter sequence during the period t-4~t and the inbound and outbound flight sequence from historical sample data; S2, collect the meteorological parameter sequence, inbound and outbound flight sequence, and cooling load of the period t-4-t-1 from t-1 to the dynamic cooling load prediction neural network model, and output the cooling load prediction value of the airport terminal at time t.

2. The method for predicting dynamic cooling load of an airport terminal according to claim 1, characterized in that: The meteorological parameter sequence collected during the period t-4 to t includes the actual data of outdoor air temperature, relative humidity, solar radiation, and wet-bulb temperature collected during the period t-4 to t-1 and the predicted data of outdoor air temperature, relative humidity, solar radiation, and wet-bulb temperature at time t from the meteorological forecast.

3. The method for predicting dynamic cooling load of an airport terminal according to claim 1, characterized in that: The dynamic cooling load prediction modeling adopts a TCN-Transformer deep learning model, including an input layer, a TCN network, a Transformer network, a fully connected layer and an output layer, wherein the TCN network model includes a void causal convolution, an activation function, and a normalization layer, and the Transformer network includes a masked self-attention mechanism layer and a feedforward layer.

4. The method for predicting dynamic cooling load of an airport terminal according to claim 1, characterized in that: The collected data are normalized before being input into the dynamic cooling load prediction modeling.

5. A computer program product, comprising a computer program, characterized in that When executed, the program is used to implement each step of the airport terminal dynamic cooling load prediction method according to any one of claims 1 to 4.

6. A dynamic cooling load prediction device for an airport terminal, which stores a computer program, characterized in that: When executed, the program is used to implement each step of the airport terminal dynamic cooling load prediction method according to any one of claims 1 to 4.

Images