Intelligent simulation method for flow of air-conditioning water system of rail transit station

Abstract

The invention provides an intelligent simulation method for the flow of an air conditioning water system of a rail transit station. The intelligent simulation method comprises the following steps that A1, a 3D model, recognized by EnergyPlus, of the rail transit station is built through SketchUp software; a2, an air conditioning system is built through EnergyPlus software; a3, modifying key parameters influencing the thermal load of the air conditioning system to generate data; a4, constructing an indoor air conditioning system load LSTM (Long Short Term Memory) prediction model of the traffic station; and A5, simulating a predicted load value of the indoor air-conditioning system through the data set and the LSTM prediction model, converting the predicted load value into the total control flow of the air-conditioning water system of the rail transit station, and adjusting the air-conditioning water system of the rail transit station. Multiple times of simulation are carried out through EnergyPlus simulation software, a large amount of data generated during operation are summarized to form a data set and preprocessed, a real-time indoor air conditioning system load value is simulated based on the preprocessed data set and an LSTM prediction model and converted into the total air conditioning water system control flow, and intelligent simulation of the air conditioning water system is achieved.

Description

Technical Field

[0001] The invention relates to the technical field of intelligent control of indoor environment of transportation buildings, and in particular to an intelligent simulation method for flow of air conditioning water system of rail transportation station. Background Art

[0002] In today's rail transit station systems, air conditioning water systems are usually used to transmit cold and heat to meet indoor cooling or heating needs. With the continuous development of social economy, building energy consumption has gradually increased, especially the energy consumption of air conditioning systems, which accounts for 30%-50% of the total building energy consumption; among them, 35%-45% of the energy consumption of air conditioning operation comes from the consumption of air conditioning water systems. Therefore, it is necessary to achieve precise intelligent control of air conditioning water systems; and ensuring that the total water system control flow required under the current air conditioning load can be accurately obtained has become a key prerequisite for achieving precise intelligent control; only through accurate load data can the system be adjusted according to actual needs to avoid energy waste and unnecessary operating costs.

[0003] Therefore, the use of advanced monitoring technology and intelligent control systems to obtain and analyze load data in real time has become an indispensable means to improve the efficiency of air conditioning water systems, reduce energy consumption, and optimize operating costs.

[0004] Currently, a large number of studies have been devoted to reducing unnecessary indoor air-conditioning system loads by controlling the water flow of air-conditioning systems; for example, patent number CN112254320B discloses an AI-based adaptive variable pressure difference control method for air-conditioning variable flow water system, which predicts the load of each terminal device through an artificial intelligence algorithm, making the prediction result more accurate and reliable, and better improving the control accuracy; for example, patent number CN116878145A discloses an air-conditioning system control method based on variable water temperature setting and variable water flow, which adjusts the set temperature of the air-conditioning water system according to changes in ambient temperature, and adjusts the water pump operating frequency and flow state according to the temperature difference of the water system, thereby reducing excess energy consumption under low-load conditions of air conditioning, making substantial contributions to achieving carbon peak and carbon neutrality, and achieving the goal of energy conservation and emission reduction; however, these methods only rely too much on real buildings to control the flow of air-conditioning water systems based on monitoring data, and are restricted by the actual operation of rail transit stations, and often have problems of missing data and lack of data sets. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide an intelligent simulation method for the flow of the air-conditioning water system of a rail transit station. The indoor air-conditioning system load is simulated by a deep learning algorithm and the EnergyPlus simulation tool, and converted into the total flow of the water system control currently required to achieve precise intelligent control of the air-conditioning water system, thereby optimizing the energy utilization of the air-conditioning water system of the rail transit station, ensuring that parameters such as temperature and humidity in the station are within an appropriate range, and providing a comfortable and pleasant indoor environment.

[0006] The technical solution adopted by the present invention to solve its technical problem is:

[0007] A method for intelligently simulating flow of an air conditioning water system in a rail transit station comprises the following steps:

[0008] A1. Collect architectural parameter information of rail transit stations and construct a 3D model of the rail transit station identified by EnergyPlus using SketchUp software;

[0009] A2. Based on the 3D model of the rail transit station in step A1, build an air conditioning system using EnergyPlus software;

[0010] A3. Modify the key parameters that affect the heat load of the air conditioning system to generate data, which is used as the data set for LSTM prediction model training;

[0011] A4. Construct an LSTM prediction model for the air conditioning system load in the transportation station.

[0012] Set the structure of the LSTM prediction model, including the network layer and the number of neurons, divide the data set in step A3 into a training set and a test set according to the ratio, train the LSTM prediction model using the training set data, and verify and evaluate the effect of the LSTM prediction model using the test set;

[0013] A5. Use the data set in step A3 and the LSTM prediction model to simulate the predicted load value of the indoor air-conditioning system, convert the predicted load value into the total control flow of the air-conditioning water system of the rail transit station, adjust the air-conditioning water system of the rail transit station, continuously optimize the LSTM prediction model, and realize intelligent simulation of the air-conditioning water system;

[0014] The LSTM prediction model is trained by using the deep learning algorithm and the training set in step A4, and the parameters of the LSTM prediction model are iteratively updated until a preset number of iterations is reached. After the training is completed, the model is verified and evaluated by calculating the root mean square error and the mean absolute error using the test set data;

[0015] After the model is able to run, the real-time data of the rail transit station is collected through the indoor and outdoor temperature sensor monitoring equipment, and it is input into the trained LSTM prediction model for simulation prediction to obtain the predicted load value of the indoor air-conditioning system. Then, the temperature difference between the target temperature and the current indoor temperature is obtained, and the predicted load value of the air-conditioning system is converted into the total control flow of the air-conditioning water system according to the conversion formula;

[0016] The total load data of the air-conditioning system formed by summarizing the real-time data and the predicted load value, and recording and regularly updating the preprocessed data set for the LSTM prediction model to realize the load prediction and real-time intelligent simulation of the total flow parameters of the air-conditioning water system control.

[0017] In step A1, the rail transit station is a rail transit building, and the building parameter information of the rail transit station includes geometric information of the building, passenger flow and station area.

[0018] In step A2, the method for constructing the air conditioning system using the EnergyPlus software is as follows:

[0019] B1. Import the 3D model of the rail transit station and the meteorological data file into the EnergyPlus software, and set the basic structural parameters of the rail transit station;

[0020] B2. Build an air-conditioning system according to the actual operation of the rail transit station, and input the building parameter information and basic structural parameters of the 3D model of the rail transit station into the air-conditioning system to ensure that the building parameter information and basic structural parameters are consistent with the equipment parameters of the transportation building energy system.

[0021] The meteorological data includes meteorological parameters and weather phenomena for each hour of a year; the basic structural parameters include the building's envelope, room orientation, geographic information, population density, and power-consuming equipment information.

[0022] In step A3, the key parameters include outdoor temperature and passenger flow of rail transit stations;

[0023] The key parameters are simulated in EnergyPlus and the corresponding output results are recorded to obtain the data output by EnergyPlus operation and summarize them to form a data set;

[0024] The data set is converted into a format for LSTM prediction model training by preprocessing to form a data set used for LSTM prediction model training;

[0025] The LSTM prediction model is an LSTM prediction model for the indoor air conditioning system load in a rail transit station; the preprocessing includes: data cleaning, data standardization, and data conversion into a format for model training.

[0026] In step A5, the real-time data is the current indoor environment monitoring data, including indoor and outdoor temperature, humidity, human flow and CO2 concentration.

[0027] In step A5, the conversion formula is: total control flow of air conditioning water system = coefficient * predicted load value of air conditioning system / temperature difference, where the temperature difference is the difference between the current indoor temperature and the target temperature.

[0028] The beneficial effects of the present invention are:

[0029] 1. By using EnergyPlus software to establish a 3D model of the rail transit station, and building an air-conditioning system based on the 3D model, the architectural parameter information, basic structural parameters and indoor target temperature of the rail transit station are set, and the key parameters of the heat load of the air-conditioning system are modified by using EnergyPlus simulation software for multiple simulations. The large amount of data generated by the operation is aggregated into a data set and preprocessed. Based on the preprocessed data set and LSTM prediction model, the real-time indoor air-conditioning system load value is simulated and converted into the total control flow of the air-conditioning water system to realize the intelligent simulation of the air-conditioning water system.

[0030] 2. It solves the problems of unreasonable utilization of air-conditioning water system and energy waste in the existing rail transit station system, optimizes the energy utilization of air-conditioning water system in rail transit stations, ensures that parameters such as temperature and humidity in the station are within an appropriate range, and provides a comfortable and pleasant indoor environment.

[0031] 3. By summarizing the monitoring data and the short-term predicted load values, the total load data of the air-conditioning system is recorded and regularly updated. The preprocessed data set is used for the LSTM prediction model to improve the adaptability of the model under different environmental changes, complete the continuous optimization of the LSTM prediction model, and realize the prediction of the load of the air-conditioning system of rail transit buildings and the real-time intelligent simulation of the total flow parameters of the air-conditioning water system control.

[0032] 4. By synchronously recording relevant monitoring data and regularly updating the data set when the LSTM prediction model is used, the adaptability of the model can be enhanced while enriching the data set, and an LSTM prediction model for the indoor air-conditioning system load in transportation stations with high accuracy and continuous updating and optimization can be trained. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 is a flow chart of the method of the present invention;

[0034] Figure 2 Schematic diagram of the 3D model of a rail transit station built using SketchUp software. DETAILED DESCRIPTION

[0035] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0036] Example 1

[0037] like Figures 1 to 2 As shown, a method for intelligently simulating the flow of air conditioning water system in a rail transit station comprises the following steps:

[0038] A1. Collect architectural parameter information of rail transit stations and use SketchUp software to build a 3D model of rail transit stations that can be recognized by EnergyPlus;

[0039] A2. Based on the 3D model of the rail transit station in step A1, build an air conditioning system using EnergyPlus software;

[0040] A3. Modify the key parameters that affect the heat load of the air conditioning system to generate a large amount of data, which is used as the data set for LSTM prediction model training;

[0041] A4. Construct an LSTM prediction model for the air conditioning system load in the transportation station.

[0042] Set the structure of the LSTM prediction model, including the network layer and the number of neurons. Select the appropriate loss function and optimizer according to the actual situation. Divide the data set in step A3 into a training set and a test set according to the ratio (8:2). Use the training set data to train the LSTM prediction model, and use the test set to verify and evaluate the effect of the LSTM prediction model.

[0043] A5. Use the data set in step A3 and the LSTM prediction model to simulate the predicted load value of the indoor air-conditioning system, convert the predicted load value into the total control flow of the air-conditioning water system of the rail transit station, adjust the air-conditioning water system of the rail transit station, continuously optimize the LSTM prediction model, and realize intelligent simulation of the air-conditioning water system;

[0044] The LSTM prediction model is trained using the deep learning algorithm and the training set in step A4, and the parameters of the LSTM prediction model are iteratively updated until a preset number of iterations is reached. After the training is completed, the performance of the model is verified and evaluated by calculating the root mean square error and mean absolute error using the test set data;

[0045] After the model is able to run, the real-time data of the rail transit station is collected through the indoor and outdoor temperature sensor monitoring equipment, and it is input into the trained LSTM prediction model for simulation prediction to obtain the short-term predicted load value of the indoor air-conditioning system. Then, the temperature difference between the target temperature and the current indoor temperature is obtained, and the short-term predicted load value of the air-conditioning system is converted into the total control flow of the air-conditioning water system according to the conversion formula;

[0046] The total load data of the air-conditioning system formed by summarizing the real-time data and the predicted load value, recording and regularly updating the preprocessed data set for the LSTM prediction model, completing the continuous optimization of the LSTM prediction model, and realizing the load prediction and real-time intelligent simulation of the total flow parameters of the air-conditioning water system control.

[0047] In step A1, the rail transit station is a rail transit building, and the architectural parameter information of the rail transit station includes geometric information of the building, passenger flow and station area, and the Legacy OpenStudio plug-in is used to save the rail transit station 3D model in the .idf format recognizable by EnergyPlus.

[0048] In step A2, the method for constructing the air conditioning system using the EnergyPlus software is as follows:

[0049] B1. Import the 3D model of the rail transit station and the meteorological data file of the area into the EnergyPlus software, and set the basic structural parameters of the rail transit station;

[0050] B2. Build an air-conditioning system according to the actual operation of the rail transit station, and input the building parameter information and basic structural parameters of the 3D model of the rail transit station into the air-conditioning system to ensure that the building parameter information and basic structural parameters are consistent with the equipment parameters of the energy system of the actual transportation building.

[0051] The meteorological data file refers to the meteorological data provided by the EnergyPlus official website for simulating the actual building operation conditions, including meteorological parameters and weather phenomena for each hour of the year; the basic structural parameters include the building's envelope structure, room orientation, geographic information, population density and power-consuming equipment information.

[0052] In step A3, set the indoor target temperature parameter, select the key parameters that affect the heat load of the air-conditioning system, including the outdoor temperature and the passenger flow of the rail transit station, determine the range and value step of each key parameter, and independently define a schedule that better conforms to the passenger flow changes in Schedule: Compact;

[0053] The number of simulations is set according to the accuracy requirements of the actual situation;

[0054] The key parameters are simulated in EnergyPlus and the corresponding output results are recorded, so as to obtain a large amount of data output by EnergyPlus and summarize them into a data set;

[0055] The simulation was performed by obtaining EP software to simulate the environment of the target station at a frequency of 10 minutes for a total of one year, obtaining a total of 8760*6 data

[0056] The data set is converted into a format suitable for LSTM prediction model training by preprocessing to form a data set used for LSTM prediction model training;

[0057] The LSTM prediction model is an LSTM prediction model for the indoor air conditioning system load in a rail transit station; the preprocessing includes: data cleaning, data standardization, and data conversion into a format suitable for model training.

[0058] In step A5, the real-time data is the current indoor environment monitoring data, including indoor and outdoor temperature, humidity, human flow and CO2 concentration.

[0059] In step A5, the conversion formula is: total control flow of air conditioning water system = coefficient * short-term predicted load value of air conditioning system / temperature difference, where the temperature difference is the difference between the current indoor temperature and the target temperature, and the coefficient is set according to the actual geographical location.

[0060] The calculation error evaluation index includes root mean square error and mean absolute error; the data set division ratio is set according to actual needs;

[0061] The root mean square error is a statistic used to measure the difference between the model prediction value and the actual observation value. The calculation formula of the root mean square error RMSE is as follows:

[0062]

[0063] In formula (1), n ​​is the total number of observations; Y i is the i-th actual observation value; is the i-th predicted value;

[0064] The mean absolute error is a statistic used to measure the average error between the predicted value and the actual observed value. The calculation formula of the mean absolute error MAE is as follows:

[0065]

[0066] In formula (2), n is the total number of observations; Y i is the i-th actual observation value; is the i-th predicted value; |·| is the absolute value.

[0067] By calculating the root mean square error (RMSE), the squares of the prediction errors are averaged and then squared to obtain a non-negative error measurement value. The smaller the RMSE, the smaller the difference between the model's predicted value and the actual observed value, and the better the model's prediction effect.

[0068] The 3D model of the rail transit station is established by using EnergyPlus software, and the air-conditioning system is built based on the 3D model. The architectural parameter information, basic structural parameters and indoor target temperature of the rail transit station are set. The key parameters of the heat load of the air-conditioning system are modified by using EnergyPlus simulation software for multiple simulations. The large amount of data generated by the operation is aggregated into a data set and preprocessed. Based on the preprocessed data set and the LSTM prediction model, the real-time indoor air-conditioning system load value is simulated and converted into the total control flow of the air-conditioning water system, realizing the intelligent simulation of the air-conditioning water system.

[0069] It solves the problems of unreasonable utilization of air-conditioning water system and energy waste in the existing rail transit station system, optimizes the energy utilization of air-conditioning water system in rail transit stations, ensures that parameters such as temperature and humidity in the station are within an appropriate range, and provides a comfortable and pleasant indoor environment.

[0070] By aggregating monitoring data and short-term predicted load values, the total load data of the air-conditioning system is recorded and regularly updated. The preprocessed data set is used for the LSTM prediction model to improve the adaptability of the model under different environmental changes, complete the continuous optimization of the LSTM prediction model, and realize the prediction of the air-conditioning system load of rail transit buildings and the real-time intelligent simulation of the total flow parameters of the air-conditioning water system control.

[0071] By synchronously recording relevant monitoring data and regularly updating the data set when the LSTM prediction model is used, the adaptability of the model can be enhanced while enriching the data set. This enables the training of an LSTM prediction model for the indoor air-conditioning system load in transportation stations with high accuracy and continuous updating and optimization.

[0072] The above description is only a preferred 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 changes and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for intelligent simulation of flow rate of air conditioning water system in rail transit station, characterized by: The following steps are involved: A1. Collect architectural parameter information of rail transit stations and construct a 3D model of the rail transit station identified by EnergyPlus using SketchUp software; A2. Based on the 3D model of the rail transit station in step A1, build an air conditioning system using EnergyPlus software; A3. Modify the key parameters that affect the heat load of the air conditioning system to generate data, which is used as the data set for LSTM prediction model training; A4. Construct an LSTM prediction model for the air conditioning system load in the transportation station. Set the structure of the LSTM prediction model, including the network layer and the number of neurons, divide the data set in step A3 into a training set and a test set according to the ratio, train the LSTM prediction model using the training set data, and verify and evaluate the effect of the LSTM prediction model using the test set; A5. Use the data set in step A3 and the LSTM prediction model to simulate the predicted load value of the indoor air-conditioning system, convert the predicted load value into the total control flow of the air-conditioning water system of the rail transit station, adjust the air-conditioning water system of the rail transit station, continuously optimize the LSTM prediction model, and realize intelligent simulation of the air-conditioning water system; The LSTM prediction model is trained by using the deep learning algorithm and the training set in step A4, and the parameters of the LSTM prediction model are iteratively updated until a preset number of iterations is reached. After the training is completed, the model is verified and evaluated by calculating the root mean square error and the mean absolute error using the test set data; After the model is able to run, the real-time data of the rail transit station is collected through the indoor and outdoor temperature sensor monitoring equipment, and it is input into the trained LSTM prediction model for simulation prediction to obtain the predicted load value of the indoor air-conditioning system. Then, the temperature difference between the target temperature and the current indoor temperature is obtained, and the predicted load value of the air-conditioning system is converted into the total control flow of the air-conditioning water system according to the conversion formula; The total load data of the air-conditioning system formed by summarizing the real-time data and the predicted load value, and recording and regularly updating the preprocessed data set for the LSTM prediction model to realize the load prediction and real-time intelligent simulation of the total flow parameters of the air-conditioning water system control.

2. The method for intelligent simulation of flow rate of air conditioning water system in rail transit station according to claim 1, characterized in that: In step A1, the rail transit station is a rail transit building, and the building parameter information of the rail transit station includes geometric information of the building, passenger flow and station area.

3. The method for intelligent simulation of flow rate of air conditioning water system in rail transit station according to claim 1, characterized in that: In step A2, the method for constructing the air conditioning system using the EnergyPlus software is as follows: B1. Import the 3D model of the rail transit station and the meteorological data file into the EnergyPlus software, and set the basic structural parameters of the rail transit station; B2. Build an air-conditioning system according to the actual operation of the rail transit station, and input the building parameter information and basic structural parameters of the 3D model of the rail transit station into the air-conditioning system to ensure that the building parameter information and basic structural parameters are consistent with the equipment parameters of the transportation building energy system.

4. The method for intelligent simulation of flow rate of air conditioning water system in rail transit station according to claim 3, characterized in that: The meteorological data includes meteorological parameters and weather phenomena for each hour of a year; the basic structural parameters include the building's envelope, room orientation, geographic information, population density, and power-consuming equipment information.

5. The method for intelligent simulation of flow rate of air conditioning water system in rail transit station according to claim 1, characterized in that: In step A3, the key parameters include outdoor temperature and passenger flow of rail transit stations; The key parameters are simulated in EnergyPlus and the corresponding output results are recorded to obtain the data output by EnergyPlus operation and summarize them to form a data set; The data set is converted into a format for LSTM prediction model training by preprocessing to form a data set used for LSTM prediction model training; The LSTM prediction model is an LSTM prediction model for the indoor air conditioning system load in a rail transit station; the preprocessing includes: data cleaning, data standardization, and data conversion into a format for model training.

6. The method for intelligent simulation of flow rate of air conditioning water system in rail transit station according to claim 1, characterized in that: In step A5, the real-time data is the current indoor environment monitoring data, including indoor and outdoor temperature, humidity, human flow and CO2 concentration.

7. The method for intelligent simulation of flow rate of air conditioning water system in rail transit station according to claim 1, characterized in that: In step A5, the conversion formula is: total control flow of air conditioning water system = coefficient * predicted load value of air conditioning system / temperature difference, where the temperature difference is the difference between the current indoor temperature and the target temperature.

Images