Metro station central air conditioner full-link energy-saving system and method based on wind balance

Abstract

The invention provides a subway station central air conditioner full-link energy-saving system and method based on wind balance. Belongs to the technical field of air-conditioning energy conservation. The method comprises the steps that environmental data inside and outside a subway station are collected in real time through a sensor network, and the collected environmental data are uploaded to a central air conditioner full-link energy-saving control platform; the central air conditioner full-link energy-saving control platform dynamically adjusts the air volume of the air supply pipe and the air return pipe through fan control according to the collected data; by collecting environment data in real time and combining an advanced control algorithm, the system can dynamically adjust the rotating speed of a fan, the load of a refrigerating unit, the ratio of fresh air to return air and the like, so that the running state of the air conditioning system is optimized, energy consumption is effectively reduced, and the overall energy-saving effect of the system is improved.

Description

Technical Field

[0001] The present invention proposes a full-link energy-saving system and method for central air conditioning in a subway station based on wind balance, and belongs to the technical field of air conditioning energy saving. Background Art

[0002] As an important node of urban transportation, the central air-conditioning system of subway stations has a significant impact on passenger comfort and energy consumption. Traditional central air-conditioning systems in subway stations have problems such as mismatch between designed load and actual load, frequent problems with chilled water systems, and improper control of cooling water systems, which lead to serious energy waste and difficulty in achieving precise temperature and humidity control. Therefore, a full-link energy-saving method for central air-conditioning in subway stations is needed, which can achieve wind balance, intelligent control, and greatly improve energy efficiency. Summary of the invention

[0003] The present invention provides a full-link energy-saving system and method for central air conditioning in subway stations based on wind balance, which is used to solve the problems mentioned in the above background technology: The present invention proposes a full-link energy-saving method for central air conditioning in a subway station based on wind balance, and the method comprises: S1. Collect environmental data inside and outside the subway station in real time through the sensor network, and upload the collected environmental data to the central air-conditioning full-link energy-saving control platform; S2. The central air conditioning full-link energy-saving control platform dynamically adjusts the air volume of the supply and return air ducts through fan control based on the collected data; S3, read all technical parameters of the refrigeration unit through the Modbus network protocol, use the extreme value search algorithm to optimize the overall COP of the central air conditioner, and control the host addition and subtraction according to the cooling demand; S4. According to relevant conditions, adjust the air volume and number of cooling tower fans, adjust the supply of cold and hot sources in real time according to the load status of the fresh air unit, conduct wind and water joint adjustment, and adjust the supply air temperature according to the corresponding parameters; S5. Adjust the air valve opening according to the changes in the indoor and outdoor environment, and conduct centralized intelligent management of central air-conditioning equipment; S6. The opening angle of the electric proportional-integral regulating air valve in the corresponding area is adjusted based on the feedback signals from the high-precision temperature sensors installed in different areas of the platform layer.

[0004] The present invention proposes a subway station central air conditioning full-link energy-saving system based on wind balance, and the system includes: Data collection module: collects environmental data inside and outside the subway station in real time through the sensor network, and uploads the collected environmental data to the central air-conditioning full-link energy-saving control platform; Dynamic adjustment module: The central air-conditioning full-link energy-saving control platform dynamically adjusts the air volume of the supply and return air ducts through fan control based on the collected data; Parameter reading module: reads all technical parameters of the refrigeration unit through the Modbus network protocol, uses the extreme value search algorithm to optimize the overall COP of the central air conditioner, and controls the host addition and subtraction according to the cooling demand; Supply adjustment module: adjust the air volume and number of cooling tower fans according to the cooling water temperature difference and related conditions, adjust the supply of cold and hot sources in real time according to the load status of the fresh air unit, conduct wind and water joint adjustment, and adjust the supply air temperature according to the corresponding parameters; Opening adjustment module: adjust the air valve opening according to the changes in indoor and outdoor environment, and conduct centralized intelligent management of central air-conditioning equipment; Angle adjustment module: The opening angle of the electric proportional-integral damper in the corresponding area is adjusted based on the feedback signals from the high-precision temperature sensors installed in different areas of the platform layer.

[0005] The beneficial effects of the present invention are as follows: by collecting environmental data in real time and combining advanced control algorithms (such as PID control, fuzzy control, extreme value search algorithm, etc.), the system can dynamically adjust the fan speed, refrigeration unit load, fresh air and return air ratio, etc., so as to optimize the operation state of the air conditioning system, effectively reduce energy consumption, and improve the overall energy-saving effect of the system; the system realizes the centralized intelligent management of central air-conditioning equipment by integrating sensor networks, data acquisition platforms, machine learning algorithms and other technical means. The system can monitor the equipment status in real time, automatically identify and eliminate abnormal values, and perform adaptive learning and adjustment, further improving the stability and reliability of the operation of air-conditioning equipment; through machine learning and data analysis, the system can predict load demand based on historical data and environmental change trends, and formulate energy-saving strategies based on the prediction results. This can not only help optimize equipment operating parameters and prevent energy waste, but also warn of potential system failure risks in advance, and avoid unnecessary energy consumption caused by excessive operation of equipment; with the help of fluid mechanics models, time series analysis, cluster analysis and other technologies, the system can accurately calculate the air volume of the supply air duct and the return air duct, and adjust the control strategy in time, so as to realize the precise control of the central air-conditioning system. Especially in terms of cooling tower fans, air valve openings, cooling water temperature differences, etc., they can be adjusted in real time according to environmental changes to ensure that the air-conditioning system is always in the optimal operating state; the system uses Internet of Things technology to achieve remote monitoring of equipment. Combined with a fault warning mechanism, it can identify potential problems in advance by analyzing historical data and real-time monitoring information before a fault occurs, and perform preventive maintenance, greatly improving the service life of the equipment and the operational reliability of the system; during operation, the system continuously optimizes and iterates the machine learning model through feedback data, which not only improves the current energy-saving effect, but also provides data support for future system upgrades and performance improvements. BRIEF DESCRIPTION OF THE DRAWINGS

[0006] Figure 1 It is a step diagram of the method of the present invention; Figure 2 This is a system module diagram of the present invention. DETAILED DESCRIPTION

[0007] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0008] One embodiment of the present invention, as Figure 1 As shown, a full-link energy-saving method for central air conditioning in a subway station based on wind balance comprises: S1. Collect environmental data inside and outside the subway station in real time through the sensor network, and upload the collected environmental data to the central air-conditioning full-link energy-saving control platform; S2. The central air conditioning full-link energy-saving control platform dynamically adjusts the air volume of the supply and return air ducts through fan control based on the collected data; S3, read all technical parameters of the refrigeration unit through the Modbus network protocol, use the extreme value search algorithm to optimize the overall COP of the central air conditioner, and control the host addition and subtraction according to the cooling demand; S4. Adjust the air volume and number of cooling tower fans according to the cooling water temperature difference, real-time air conditioning load, outdoor wet-bulb temperature and other related conditions, and adjust the cold and heat source supply in real time according to the load status of the fresh air unit, perform wind and water joint adjustment, and adjust the supply air temperature according to the outdoor temperature and humidity changes and the corresponding parameters such as the on-site temperature; S5. Adjust the air valve opening according to the changes in the indoor and outdoor environment, and conduct centralized intelligent management of central air-conditioning equipment; S6. The opening angle of the electric proportional-integral regulating air valve in the corresponding area is adjusted based on the feedback signals from the high-precision temperature sensors installed in different areas of the platform layer.

[0009] The working principle of the above technical solution is as follows: a dense sensor network is arranged inside and outside the subway station, and these sensors can capture key environmental data such as temperature, humidity, and passenger flow inside and outside the subway station in real time. These data are uploaded to the central air-conditioning full-link energy-saving control platform in real time through a high-speed communication network. The control platform is a central system that integrates advanced algorithms and data processing capabilities, which can quickly analyze and process massive data; the central air-conditioning full-link energy-saving control platform dynamically adjusts the air volume of the supply and return air ducts according to the collected environmental data, including fine-tuning parameters such as fan speed and damper opening. By avoiding excessive entry of fresh air or excessive discharge of return air, it can ensure that the air circulation in the subway station not only meets the comfort requirements of passengers, but also maximizes the use of energy; through the Modbus network protocol, the central air-conditioning full-link energy-saving control platform can read all technical parameters of the refrigeration unit, including operating status, energy consumption, refrigeration efficiency, etc. Using the extreme value search algorithm, the platform can optimize the overall COP (energy efficiency ratio) of the central air-conditioning. According to the actual cooling demand of the subway station, the platform can intelligently control the operation of the host's adder and subtractor to ensure that the cooling efficiency is always in the optimal state; the cooling tower is an important part of the central air-conditioning system of the subway station, and its operating efficiency directly affects the energy consumption of the entire system. The central air-conditioning full-link energy-saving control platform intelligently adjusts the air volume and number of cooling tower fans according to the cooling water temperature difference, the real-time air-conditioning load, and the outdoor wet-bulb temperature. At the same time, the platform also adjusts the supply of cold and heat sources in real time according to the load status of the fresh air unit to achieve wind and water joint adjustment. In addition, the platform can also fine-tune the supply air temperature according to the corresponding parameters such as the changes in outdoor temperature and humidity and the on-site temperature to ensure that the air environment in the subway station is always in the best state; according to the changes in the indoor and outdoor environment, the central air-conditioning full-link energy-saving control platform can intelligently adjust the air valve opening by installing an electric proportional integral regulating air valve at the outlet position of each air valve, and setting high-precision temperature sensors in different areas such as the strong cold zone, weak cold zone, and comfort level of the platform layer. These temperature sensors can monitor the temperature changes in the area in real time and accurately, and feed the data back to the control system. The control system accurately adjusts the opening angle of the electric proportional integral regulating air valve in the corresponding area according to the feedback signals from these temperature sensors and the preset temperature values ​​in different areas, so as to achieve effective control of the temperature in different areas. When the air outlet is close to the wind cabinet, the area will be easier to reach the set temperature due to the large air volume. At this time, the system automatically reduces the opening angle of the electric regulating air valve so that the excess air volume can be transported to the farther air outlet area, thereby ensuring that the temperature of the farther air outlet area can also reach the set value. At the same time, a lower temperature value is set for the strong cooling zone. When the temperature feedback from the temperature sensor in the strong cooling zone does not meet the standard, the system appropriately increases the opening of the air valve in the area; and for the weak cooling zone and the comfort zone, the air valve opening is adjusted according to their respective preset temperatures.This not only ensures that the air circulation in the subway station meets the comfort needs of passengers, but also maximizes the use of energy. At the same time, the platform also centrally manages the central air-conditioning equipment, including equipment operation status monitoring, fault warning and remote control. Through centralized management, the overall optimization and energy saving and consumption reduction of the central air-conditioning system of the subway station can be achieved.

[0010] The effects of the above technical solutions are as follows: environmental data inside and outside the subway station are collected in real time through the sensor network and uploaded to the energy-saving control platform, so that the system can be accurately adjusted according to the actual situation; the dynamic adjustment of the fan control ensures the pressure balance of the mixing chamber, avoids excessive entry of fresh air or excessive discharge of return air, thereby reducing energy waste; the extreme value search algorithm is used to optimize the overall COP of the central air-conditioning, making the operation of the refrigeration unit more efficient and further improving the energy utilization efficiency; due to the improvement in energy utilization efficiency, the energy consumption of the central air-conditioning system of the subway station is significantly reduced, thereby reducing operating costs; centralized intelligent management makes the operation of the equipment more stable, reducing downtime and maintenance costs caused by equipment failures; by accurately adjusting the supply air temperature and air volume, the air environment in the subway station is made more comfortable, It meets the needs of passengers; Feng Shui coordination ensures that the supply of cold and heat sources matches the changes in the environment inside and outside the subway station, further improving the comfort of passengers; this technical solution contributes to energy conservation and emission reduction by improving energy utilization efficiency and reducing energy consumption; reduced energy consumption means reduced fossil fuel consumption and greenhouse gas emissions, which is beneficial to environmental protection and sustainable development; centralized intelligent management makes the operation of the subway station central air-conditioning system more intelligent and automated, reducing manual intervention and operational difficulty; real-time monitoring and early warning functions enable managers to discover and solve problems in a timely manner, improving management efficiency; dynamic adjustment of fan control, refrigeration unit optimization, Feng Shui coordination and other measures make the operation of the subway station central air-conditioning system more stable and reliable; reducing downtime caused by equipment failure and the impact on subway operations.

[0011] In one embodiment of the present invention, the S1 includes: S11. Environmental data is collected through a high-precision sensor network deployed at key locations inside and outside the subway station (such as platforms, station halls, entrances and exits, tunnels, etc.), including temperature sensors, humidity sensors, CO2 concentration sensors, PM2.5 sensors, and infrared pedestrian flow counters; S12. Preliminary processing of the collected environmental data is performed through data verification and anomaly detection mechanisms, and outliers are eliminated; S13, uploading the preliminarily processed environmental data to the central air-conditioning full-link energy-saving control platform via a wired or wireless network; S14, storing the real-time data and the historical data respectively through the real-time database and the historical database established on the control platform; S15. Perform mining and analysis on the stored data through machine learning algorithms (such as cluster analysis and time series analysis), identify environmental change trends, and predict future load demand; the load demand is predicted by the following formula;

[0012] in, represents the load demand predicted at future time t+Δt; represents the value of the i-th environmental parameter at time t (such as temperature, humidity, etc.); P(t) represents the flow of people at time t; α i represents the weight coefficient of the i-th environmental parameter; β represents the weight coefficient of the impact of human flow on load; h represents a complex nonlinear function to capture the interaction effect between environmental parameters and human flow, such as a neural network; θ Represents model parameters.

[0013] The working principle of the above technical solution is as follows: deploy a high-precision sensor network at key locations inside and outside the subway station, such as platforms, station halls, entrances and exits, and tunnels. These sensors include temperature sensors, humidity sensors, CO2 concentration sensors, PM2.5 sensors, and infrared passenger flow counters; these sensors can collect real-time environmental data inside and outside the subway station, such as temperature, humidity, CO2 concentration, PM2.5 concentration, and passenger flow; the collected environmental data first undergoes data integrity verification to ensure that each sensor provides a complete data set, and then performs data rationality verification to check whether the data is within a reasonable range, such as whether the temperature is between -20℃ and 50℃, whether the humidity is between 0% and 100%, etc.; finally, apply statistical anomaly detection algorithms, such as threshold detection, box plot detection, or machine learning-based anomaly detection models, to further analyze the data and eliminate possible outliers; the environmental data after preliminary processing is uploaded to the central air-conditioning full-link energy-saving control platform via wired or wireless networks. Wired networks may include high-speed transmission media such as optical fiber, while wireless networks may include 4G / 5G or dedicated networks. A real-time database and a historical database are established on the control platform to store real-time data and historical data respectively. The real-time database is used to store data currently being collected and processed in order to quickly respond to environmental changes. The historical database is used to store data from a period of time in the past for long-term analysis and trend forecasting. The stored data is mined and analyzed through machine learning algorithms such as cluster analysis and time series analysis. Cluster analysis can help identify the characteristics of environmental data in different time periods or regions, thereby formulating more accurate energy-saving strategies. Time series analysis can be used to predict future load demand and provide a scientific basis for the operation and scheduling of central air-conditioning systems.

[0014] The effect of the above technical solution is: by deploying a high-precision sensor network at key locations inside and outside the subway station, it is possible to collect environmental data inside and outside the subway station in real time and accurately, such as temperature, humidity, CO2 concentration, PM2.5 concentration, and pedestrian flow. These data provide important input information for the energy-saving control of the central air-conditioning system; ensure that each sensor provides a complete data set to avoid inaccurate control caused by missing data; through data rationality verification, it can be checked whether the data is within a reasonable range, thereby eliminating unreasonable data and improving data reliability; the statistical anomaly detection algorithm can further identify and eliminate outliers to ensure the accuracy and effectiveness of the data. This helps to avoid miscontrol caused by sensor failure or sudden changes in environmental factors; upload the initially processed environmental data to the central air-conditioning full-link energy-saving control platform through a wired or wireless network, and realize real-time data transmission. It helps the control platform to obtain the latest environmental data in a timely manner, so as to make accurate control decisions; the real-time database and historical database established on the control platform are used to store real-time data and historical data respectively. It helps to achieve long-term preservation and efficient retrieval of data, and provides convenience for subsequent data analysis and mining; mining and analysis of stored data through machine learning algorithms such as cluster analysis and time series analysis can identify environmental change trends and predict future load demands. This provides a scientific basis for the intelligent control and decision optimization of central air-conditioning systems; intelligent control and decision optimization based on environmental data help achieve energy saving and consumption reduction of central air-conditioning systems. For example, the operating state of the refrigeration unit is adjusted according to the predicted future load demand to avoid energy waste caused by over-cooling or insufficient cooling; by real-time monitoring of environmental data inside and outside the subway station and adjusting the operating state of the central air-conditioning system according to these data, a more comfortable and healthy riding environment can be provided for passengers; intelligent control and decision optimization help reduce the failure rate and downtime of the central air-conditioning system and improve the efficiency of subway operations. At the same time, energy saving and consumption reduction also help reduce subway operating costs and improve the economic benefits of subway operations. The above load demand formula contains multiple environmental parameters (such as temperature, humidity, etc.) and passenger flow as independent variables, which can more comprehensively reflect the changes in actual load demand. By comprehensively considering these factors, the forecast results are more accurate and closer to the changing trend of the actual load demand; the weight coefficients in the formula can be flexibly adjusted according to the actual situation to reflect the relative importance of different factors on future load demand. This flexibility enables the forecasting model to better adapt to changes in demand in different scenarios and time periods. The nonlinear function h in the formula can capture the complex interaction effects between environmental parameters and human flow. This interaction effect often exists in the actual environment and is difficult to describe with a simple linear relationship. By introducing nonlinear functions, the forecasting model can better capture these complex relationships, thereby improving the accuracy of the forecast.The formula implies the idea of ​​time series analysis, which is to use historical data to predict future load demand. Time series analysis can capture the trend and periodicity of load demand over time, thereby further improving the accuracy of the prediction. Although the specific form of the time series analysis model f is not directly given in the formula, its existence provides the prediction model with stronger time series analysis capabilities. The θ in the formula represents the model parameters, which can be optimized by machine learning algorithms to minimize the prediction error. By continuously optimizing the model parameters, the generalization ability and prediction accuracy of the prediction model can be improved. Through accurate load demand forecasting, strong data support can be provided for relevant decision-making departments. This helps to formulate a more scientific and reasonable resource scheduling plan, improve resource utilization efficiency, reduce operating costs, and improve user satisfaction and service quality.

[0015] In one embodiment of the present invention, the S15 includes: S151. Based on the characteristics of environmental data, select features that have an important impact on load demand prediction; such as temperature, humidity, CO2 concentration, PM2.5 concentration, human flow, etc., and perform feature extraction, and normalize or standardize the feature data; S152. Select machine learning algorithms based on the characteristics of historical data and forecasting requirements, such as time series analysis (ARIMA, LSTM), cluster analysis (K-means, DBSCAN), regression models (linear regression, decision tree regression, random forest regression), etc.; S153, using the processed stored data to train the selected machine learning model, and verifying the trained model through cross-validation to evaluate the generalization ability of the model on unknown data; the generalization ability is evaluated by the following formula:

[0016] Where k represents the number of cross-validation folds; N i represents the number of data points in the i-th fold; y ij represents the true value of the jth data point in the i-th fold; represents the predicted value of the jth data point in the i-th fold; S154. Use time series analysis or cluster analysis methods to identify the changing trends of environmental parameters in the stored data, such as seasonal changes, periodic changes, etc., and predict the load demand in the future based on the trained machine learning model, including predicting the air conditioning load demand in the next few hours, days or weeks; S155. Quantitatively evaluate the prediction results, such as calculating the prediction error, etc., fine-tune the model according to the evaluation results, and formulate energy-saving strategies based on the prediction results, such as adjusting the operating parameters of the air-conditioning system, optimizing the operating strategies of the fan and the refrigeration unit, etc.; the prediction error is obtained by the following formula:

[0017] Where n represents the number of test data points; y i Represents the true value of the i-th data point; Represents the predicted value of the i-th data point; S156. Feedback the prediction results to the central air-conditioning full-link energy-saving control platform in real time, and adjust the operating status of the air-conditioning system in real time according to the predicted load demand; S157. Collect feedback data based on the actual operating results of the air-conditioning system and continuously optimize and iterate the prediction model.

[0018] The working principle of the above technical solution is as follows: based on the characteristics of environmental data, select features that have an important impact on load demand forecasting, such as temperature, humidity, CO2 concentration, PM2.5 concentration, and passenger flow. These features can reflect the environmental conditions inside and outside the subway station and are closely related to the air conditioning load demand; extract these features from the original data and perform normalization or standardization to eliminate the dimensional differences between different features and improve the training efficiency and prediction accuracy of the model; select a suitable machine learning algorithm based on the characteristics of historical data and prediction needs. For example, for data with time series characteristics, you can choose a time series analysis algorithm (such as ARIMA, LSTM); for data that requires cluster analysis, you can choose a cluster analysis algorithm (such as K-means, DBSCAN); for data that requires a prediction model, you can choose a regression model (such as linear regression, decision tree regression, random forest regression), etc.; taking LSTM (Long Short-Term Memory Network) in time series analysis as an example, it is a neural network that is particularly suitable for processing and predicting important events with long intervals and delays in time series. LSTM can learn long-term dependency information and store and update information by adding special units (i.e., memory units) in the hidden layer, thereby achieving accurate prediction of time series data; the selected machine learning model is trained using the processed stored data. During the training process, the model learns the mapping relationship between feature data and target load demand; the trained model is verified by cross-validation. Cross-validation is a commonly used model evaluation method that divides the data set into multiple subsets, using one subset as the validation set each time and the remaining subsets as the training set. Through multiple training and validation processes, the generalization ability of the model on unknown data can be evaluated; time series analysis or cluster analysis methods are used to identify the changing trends of environmental parameters in the stored data, such as seasonal changes, cyclical changes, etc. These trends help to understand the changing laws of load demand and provide a basis for future predictions; based on the trained machine learning model, the load demand in the future is predicted. The prediction time range can be set according to actual needs, such as predicting the air conditioning load demand in the next few hours, days, or weeks; quantitative evaluation of the prediction results, such as calculating prediction errors, confidence intervals, etc. These evaluation indicators can reflect the prediction accuracy and reliability of the model; fine-tune the model according to the evaluation results to improve the prediction performance of the model. Fine-tuning methods include adjusting model parameters and increasing training data; based on the prediction results, energy-saving strategies are formulated, such as adjusting the operating parameters of the air-conditioning system and optimizing the operating strategies of the fan and refrigeration unit. These strategies aim to reduce the energy consumption of the air-conditioning system while meeting the comfort of passengers; and feed back the prediction results to the central air-conditioning full-link energy-saving control platform in real time.The control platform can adjust the operating status of the air-conditioning system in real time according to the predicted load demand to achieve the goal of energy saving and consumption reduction; according to the actual operation effect of the air-conditioning system, feedback data is collected to continuously optimize and iterate the prediction model. By continuously optimizing the model, the prediction accuracy and adaptability of the model can be improved, providing more reliable support for future load demand prediction.

[0019] The effect of the above technical solution is: by selecting features that have an important impact on load demand prediction based on the characteristics of environmental data, and performing feature extraction and normalization or standardization processing, it can ensure that the data input into the machine learning model is high-quality and representative. This helps to improve the accuracy of the model's load demand prediction, so that the prediction results are closer to the changes in actual load demand; the selected machine learning model is trained using the processed stored data, and the trained model is verified by cross-validation, which can evaluate the generalization ability of the model on unknown data. This verification method helps to ensure that the model can still maintain a high prediction accuracy when facing new data, thereby enhancing the practicality and reliability of the model; based on the trained machine learning model, the load demand in the future period is predicted, and the load demand of the air-conditioning system can be accurately grasped. According to the prediction results, energy-saving strategies are formulated, such as adjusting the operating parameters of the air-conditioning system, optimizing the operating strategies of the fan and the refrigeration unit, etc., to achieve intelligent energy-saving control. This helps to reduce the energy consumption of the air-conditioning system, improve energy utilization efficiency, and thus achieve the goal of energy saving and emission reduction; the prediction results are fed back to the central air-conditioning full-link energy-saving control platform in real time, and the operating status of the air-conditioning system can be adjusted in real time according to the predicted load demand. This real-time feedback mechanism helps to improve the response speed and flexibility of the system, so that the air-conditioning system can adapt to changes in load demand more quickly, thereby providing a more comfortable and stable indoor environment; collecting feedback data based on the actual operation effect of the air-conditioning system, and continuously optimizing and iterating the prediction model can continuously improve the prediction performance and adaptability of the model. This continuous optimization and iteration process helps to ensure that the model always remains in the best state, thereby providing users with more accurate and reliable load demand prediction services. The above generalization ability calculation formula comprehensively evaluates the generalization ability of the model on unknown data by calculating the average value of the mean square error (MSE) of each fold of data during the cross-validation process. This evaluation method can fully consider the diversity and distribution characteristics of the data set, thereby more accurately reflecting the actual performance of the model; cross-validation is an effective method to prevent overfitting. By dividing the data set into multiple folds and training and validating each fold separately, it can ensure that the model can maintain good performance on unseen data. By calculating the MSE of each fold and taking the average value, this formula helps to discover the possible overfitting of the model on a certain fold, so as to make timely adjustments and optimizations; during the cross-validation process, the data of each fold is randomly divided, which helps to evaluate the stability of the model on different data sets. This formula can reflect whether the model performs consistently on different data sets by calculating the average value of multi-fold MSE, thereby judging whether the model is stable. This formula is simple to calculate and the result is intuitive and easy to understand. As a commonly used error metric, MSE can clearly reflect the difference between the model's predicted value and the true value.By comparing the MSE values ​​of different models, you can intuitively determine which model has a stronger generalization ability on unknown data; during the model training process, you can adjust the parameters and structure of the model according to the MSE value calculated by this formula to optimize the performance of the model. Through continuous iteration and optimization, the MSE value can be gradually reduced, thereby improving the prediction accuracy of the model on unknown data. This formula is not only applicable to regression problems, but can also be extended to other machine learning scenarios such as classification problems. By appropriately adjusting the error metric in the formula (such as replacing MSE with classification accuracy, etc.), it can be applied to model evaluation and optimization in different fields.

[0020] In one embodiment of the present invention, the S2 includes: S21. Based on the collected environmental data, the current air volume values ​​of the air supply duct and the air return duct are calculated using a fluid mechanics model; and the fan speed and inverter output are dynamically adjusted through a PID control algorithm; the current air volume value is calculated using the following formula:

[0021] Among them, ΔP represents the pressure loss (air volume) in the pipeline, usually in Pascal (Pa); f represents the friction factor, which depends on the roughness of the pipeline and the Reynolds number (Re); L represents the length of the pipeline in meters (m); D represents the diameter of the pipeline in meters (m); represents the density of the fluid in kilograms per cubic meter (kg / m³); v represents the average velocity of the fluid in the pipe in meters per second (m / s); The friction factor is calculated using the following formula:

[0022] The Reynolds number Re is obtained by the following formula:

[0023] Wherein, μ represents the dynamic viscosity of the fluid, in Pascal seconds (Pa·s); S22, monitor the pressure of the air mixing chamber in real time, and use fuzzy control strategy or neural network control strategy to dynamically adjust the ratio of fresh air and return air according to pressure changes; S23. Based on the pressure balance simulation model, simulate the wind pressure changes under different working conditions and verify the effectiveness of the control strategy; S24. Based on the adaptive learning mechanism, the control parameters are continuously optimized according to the operating data, and regular pressure balance tests are carried out to evaluate the system performance.

[0024] The working principle of the above technical solution is: based on the collected environmental data (such as temperature, humidity, wind speed, etc.), the current air volume values ​​of the supply air duct and the return air duct are calculated using the fluid mechanics model. The fluid mechanics model takes into account factors such as the unit friction air volume of the air duct, the length of the air duct, the ratio of the local air volume to the friction air volume loss, and can accurately calculate the air volume value in the air duct; through the PID (proportional-integral-differential) control algorithm, the fan speed and the inverter output are dynamically adjusted according to the calculated air volume value. The PID control algorithm can monitor the changes in air volume in real time and make precise adjustments according to the set target value to achieve a balance between the supply and return air volumes; the pressure sensor is used to monitor the pressure changes in the mixing chamber in real time to ensure that the pressure in the mixing chamber remains within the set range; according to the changes in the pressure in the mixing chamber, the fuzzy control strategy or the neural network control strategy is used to dynamically adjust the ratio of fresh air and return air. The fuzzy control strategy can handle complex nonlinear relationships and improve the accuracy of working condition identification; while the neural network control strategy has the ability of self-learning and self-adaptation, and can continuously optimize the control strategy based on historical data; and when the wind system adjusts the air volume, the built-in water system of the system can also adjust the water volume synchronously according to the change of air volume. For example, when the opening of the air valve in a certain area increases and the air supply volume increases, the water system automatically increases the opening of the water valve of the corresponding wind cabinet to ensure that the increased air volume can carry enough cold or heat to meet the temperature requirements of the area. Through the air volume-water volume matching model, according to the load changes in different seasons and time periods, as well as the temperature feedback of each area, the operating parameters of the wind system and the water system are dynamically adjusted to keep the two in the best matching state at all times and improve energy utilization efficiency; after implementing the air volume balance energy-saving measures, the output frequency of the wind cabinet is reduced, and the water system is also adjusted accordingly. As the air volume of the wind cabinet decreases, the water system reduces the frequency of the refrigeration pump and the cooling pump, reduces the opening of the water valve, reduces the production and delivery of cold, and avoids energy waste. According to the changes in the outlet water temperature of the host, the opening of the wind system and the air valve and the operating parameters of the water system are coordinated and adjusted. When the outlet water temperature of the main unit is increased, the wind system adjusts the air supply volume appropriately, and the water system further optimizes the flow and temperature of the chilled water to ensure that the temperature control effect of each area is not affected while saving energy. Based on the pressure balance simulation model, the wind pressure changes under different working conditions are simulated. The model can consider the influence of various factors (such as duct air volume, fan speed, inverter output, etc.) on the wind pressure, so as to simulate the wind pressure changes under different working conditions; through the simulation model, the effectiveness of the proposed control strategy is verified. By comparing the simulation results with the actual operation results, the advantages and disadvantages of the control strategy can be evaluated, and necessary adjustments and optimizations can be made; based on the adaptive learning mechanism, the control parameters are continuously optimized according to the operation data. The mechanism can monitor the operating status of the system in real time, and automatically adjust the control parameters according to the changes in the operating status to improve the accuracy and response speed of the wind balance adjustment; regular pressure balance tests are carried out to evaluate the system performance.Through testing, problems in the system can be discovered and solved in a timely manner to ensure the continuous stability of the wind balance state. At the same time, the test data can also provide important reference for subsequent optimization and improvement.

[0025] The effect of the above technical solution is: by using the fluid mechanics model to calculate the current air volume values ​​of the supply air duct and the return air duct based on the collected environmental data, the air volume status of the air system can be accurately grasped. Combined with the PID control algorithm, the fan speed and the inverter output are dynamically adjusted to achieve the balance of the supply and return air volumes, effectively avoiding the problems of uneven air volume distribution and increased energy consumption caused by unbalanced air volume; real-time monitoring of the mixing chamber pressure, and the use of fuzzy control strategy or neural network control strategy to dynamically adjust the ratio of fresh air and return air according to pressure changes. This intelligent control method can quickly respond to environmental changes and load requirements, ensure the reasonable ratio of fresh air and return air, improve indoor air quality, and optimize energy utilization; based on the pressure balance simulation model, the wind pressure changes under different working conditions are simulated, and the effectiveness of the control strategy is verified. This step helps to discover and solve potential problems in the design stage, ensuring that the control strategy can achieve the expected results in practical applications. At the same time, through simulation verification, the control strategy can be optimized to improve its adaptability and robustness; based on the adaptive learning mechanism, the control parameters can be continuously optimized according to the operating data, which can further improve the accuracy and response speed of the wind balance adjustment. This continuous optimization process helps to maintain the optimal operating state of the system and improve the overall performance of the system. At the same time, regular pressure balance tests are carried out to evaluate the system performance, which can timely discover and solve potential problems and ensure the continuous stability of the wind balance state; through the comprehensive application of the above technical solutions, the wind balance state of the air conditioning system can be accurately controlled and continuously optimized, thereby improving the overall energy efficiency and reducing operating costs. This not only helps to improve the economic benefits of the enterprise, but also conforms to the current social development trend of energy conservation and emission reduction. The above formula can accurately calculate the current air volume value in the supply duct and the return duct through the given pressure loss formula. This is crucial for optimizing the air conditioning system, reducing energy consumption and improving comfort; with the help of the PID control algorithm, the fan speed and inverter output can be dynamically adjusted in real time according to the calculated air volume value. This helps to achieve a balance between the supply and return air volumes, ensuring stable and efficient operation of the system; the calculation formula for the friction factor f takes into account the different ranges of the Reynolds number Re, making the air volume calculation applicable to pipe conditions with different roughness and different flow rates. This enhances the versatility and practicality of the formula; through real-time monitoring and dynamic adjustment, the system can quickly respond to environmental changes, such as pipe blockage, flow changes, etc., thereby maintaining the stability and efficiency of the air conditioning system.

[0026] In one embodiment of the present invention, the S22 includes: S221, using a pressure sensor installed in the air mixing chamber to monitor the pressure change in the air mixing chamber in real time, verifying the collected pressure data, and identifying and eliminating abnormal values; S222. Use time series analysis technology to analyze the changing trend of the air mixing chamber pressure and identify the laws and characteristics of pressure fluctuations; S223. According to the changing characteristics of the air mixing chamber pressure and the control requirements, the applicability of the fuzzy control strategy and the neural network control strategy is evaluated, and the optimal control strategy is selected; S224, initializing parameters of the selected control strategy, and verifying the adaptability and stability of the control strategy under different working conditions through simulation; S225. According to the change of the pressure in the air mixing chamber, the optimal ratio of fresh air to return air is calculated by using the control strategy, and the calculated ratio of fresh air to return air is converted into a control instruction of the actuator. The optimal ratio is calculated by the following formula:

[0027] Where e represents the error between the current pressure in the mixing chamber and the target pressure; Δe represents the rate of change of the pressure error; A min and A max Respectively represent the minimum and maximum value of fresh air ratio; K P and K d represents the proportional and differential gains of the fuzzy controller; K I Indicates the integral gain; A base Indicates the basic fresh air ratio when there is no pressure error; ΔA fuzzy The additional adjustment provided by the fuzzy logic controller is calculated using the fuzzy rule base, i.e. , and F total represents the comprehensive fuzzy logic function; in, and Represents fuzzy logic functions that map pressure errors and error change rates to adjustments to the fresh air ratio. These functions are usually defined by a fuzzy rule base as follows:

[0028] Among them, NB, NM, NS, ZE, PS, PM, and PB represent negative large, negative medium, negative small, zero, positive small, positive medium, and positive large, respectively; It mainly reflects the influence of the current size of the error on the control quantity, which is similar to the proportional (P) link in traditional PID control, but it realizes nonlinear adjustment through fuzzy logic; for example, when the error e is large, A larger adjustment will be output to quickly respond to deviations; It mainly reflects the influence of the changing trend of the error on the control quantity, which is similar to the differential (D) link in PID control; for example, when the error change rate When it is negative (NB), it means that the error is decreasing rapidly. The adjustment amount may be reduced to avoid system overshoot; The function is used to convert the precise input value into a fuzzy set, and then calculate the optimal ratio of fresh air and return air with the help of the fuzzy rule base.

[0029] S226. Monitor the adjusted air mixing chamber pressure in real time and evaluate the adjustment effect. If the expected target is not achieved, make iterative adjustments until the pressure is stable.

[0030] The working principle of the above technical solution is as follows: a high-precision pressure sensor is installed in the mixing chamber to monitor the pressure changes in the mixing chamber in real time; the sensor transmits the collected pressure data to the central control system through a signal; the central control system verifies the received pressure data to check the integrity and accuracy of the data; identifies and eliminates abnormal values ​​caused by sensor failure, data transmission errors, etc., to ensure the accuracy of subsequent analysis; uses time series analysis technology to perform statistical analysis on the historical data of the mixing chamber pressure to identify the trend and periodicity of pressure changes; analyzes the laws and characteristics of pressure fluctuations, such as fluctuation amplitude and frequency, to provide a basis for the formulation of subsequent control strategies; evaluates the applicability of fuzzy control strategies and neural network control strategies based on the changing characteristics of the mixing chamber pressure and control requirements; selects the optimal control strategy considering factors such as the real-time, stability, and robustness of the system; initializes the parameters of the selected control strategy, including the membership function of the fuzzy controller, the rule base, or the structure of the neural network , learning rate, etc.; the settings of these parameters will directly affect the performance and effect of the control strategy; the adaptability and stability of the control strategy under different working conditions are verified through simulation software; according to the simulation results, the parameters of the control strategy are adjusted until the best control effect is achieved; according to the changes in the pressure in the mixing chamber, the control strategy is used to calculate the best ratio of fresh air and return air; this ratio will ensure that the pressure in the mixing chamber is stable within the set range; the calculated fresh air and return air ratio is converted into control instructions of the actuator, such as adjusting the opening of the fresh air valve and the return air valve; these instructions will directly control the flow rate of fresh air and return air to achieve dynamic adjustment of the ratio; real-time monitoring of the adjusted mixing chamber pressure and evaluation of the adjustment effect; if the adjusted pressure is stable and meets the expected target, the control strategy is effective; if the adjusted pressure does not meet the expected target, iterative adjustment is performed; according to the real-time monitored pressure data, the parameters of the control strategy are adjusted or the control strategy is reselected; the above steps are repeated until the pressure is stable and meets the expected target.

[0031] The effect of the above technical solution is: by installing a high-precision pressure sensor to monitor the pressure change of the mixing chamber in real time, and verifying and eliminating abnormal values ​​of the collected data, the accuracy and reliability of the data are ensured. This provides accurate data support for the basic pressure control, so that the control strategy can more accurately reflect the actual situation of the pressure in the mixing chamber, thereby improving the accuracy of the control; the time series analysis technology is used to analyze the changing trend of the pressure in the mixing chamber, identify the law and characteristics of pressure fluctuations, and provide a scientific basis for the formulation of the control strategy. According to the changing characteristics and control requirements of the pressure in the mixing chamber, the optimal control strategy (fuzzy control or neural network control) is selected, and the parameters are initialized, and its adaptability and stability under different working conditions are verified by simulation. This refined control strategy can more accurately calculate the optimal ratio of fresh air and return air, realize dynamic adjustment of the ratio, thereby optimizing indoor air quality and meeting people's needs for a comfortable environment; by initializing the parameters of the control strategy and simulating and verifying it, it can be ensured that the control strategy has good adaptability and stability in practical applications. This helps to reduce system fluctuations and instability caused by improper control strategies and improve the overall stability and robustness of the system. At the same time, the adjusted mixing chamber pressure is monitored in real time to evaluate the adjustment effect. If the expected target is not achieved, iterative adjustments are made until the pressure is stable. This iterative adjustment mechanism ensures that the system can quickly adjust and remain stable when facing different working conditions and load changes; by accurately controlling the ratio of fresh air to return air, the air supply volume of the air conditioning system can be accurately adjusted, thereby avoiding unnecessary energy waste. At the same time, the optimization control strategy can also reduce the increase in energy consumption caused by system fluctuations and instability, and further improve energy utilization efficiency. This helps to reduce the operating costs of enterprises and improve economic benefits; the technical solution adopts a modular design, and each step and component is relatively independent, which is easy to maintain and expand. When the control strategy needs to be updated or improved, only the corresponding modules need to be adjusted and optimized without large-scale changes to the entire system. This reduces the difficulty and cost of system maintenance and upgrades, and improves the flexibility and scalability of the system. The above optimal ratio formula can calculate the optimal ratio of fresh air and return air in real time according to the error (e) and the change rate (Δe) between the current pressure and the target pressure in the mixing chamber, thereby realizing precise control of the pressure in the mixing chamber. Through the additional adjustment amount provided by the fuzzy logic controller (FLC), the system can further adapt to the complex and changeable operating environment and improve the robustness and flexibility of the control.By precisely controlling the ratio of fresh air to return air, the system can maximize the use of return air and reduce the consumption of fresh air while ensuring the stability of the pressure in the mixing chamber, thereby reducing energy consumption; this optimized ratio is particularly important when the outdoor air quality is poor or the fresh air supply is tight; stable mixing chamber pressure helps maintain the comfort of the indoor environment and avoids problems such as air flow turbulence and increased noise caused by pressure fluctuations; by precisely controlling the fresh air ratio, it can also ensure that the indoor air quality meets the standards and provide users with a good breathing environment. The integral term in the formula helps to eliminate static errors in the system and improve the steady-state accuracy of the system. The proportional term and differential term are used to improve the response speed and stability of the system, respectively, and prevent the system from over-adjusting or oscillating. The various parameters and fuzzy rule base in the formula can be adjusted and optimized according to actual conditions to make the system more in line with actual application needs.

[0032] In one embodiment of the present invention, the S222 includes: Use time series decomposition techniques, such as the STL method, to decompose the air mixing chamber pressure time series into trend terms, seasonal terms, and residual terms, and identify the long-term trend and seasonal fluctuation of pressure changes; Statistical tools such as autocorrelation function (ACF) and partial autocorrelation function (PACF) are used to identify the periodic components in the air mixing chamber pressure time series, such as daily cycle, weekly cycle, etc., and analyze their cycle length and fluctuation amplitude; Anomaly detection algorithms based on time series, such as statistics-based methods and machine learning-based methods, identify abnormal values ​​or mutation points in the time series of the air mixing chamber pressure; Based on the results of time series analysis, the typical patterns of pressure fluctuations in the air mixing chamber are identified and the causes of the fluctuations are analyzed; Extract key features from the air mixing chamber pressure time series, and select appropriate prediction models such as ARIMA model, LSTM neural network, etc. according to the characteristics of the air mixing chamber pressure time series for preliminary modeling; The prediction model is trained using historical data, and the trained prediction model is verified using the holdout method to evaluate the prediction accuracy and generalization ability of the model; the model is tuned based on the verification results.

[0033] The working principle of the above technical solution is: using the STL method, the mixing chamber pressure time series is decomposed into trend terms, seasonal terms and residual terms. The trend term reflects the long-term trend of the time series, the seasonal term reflects the seasonal fluctuation, and the residual term includes other changes except the trend and season. For example, suppose that there is an obvious seasonal fluctuation in the mixing chamber pressure time series due to seasonal changes (such as increased pressure caused by increased temperature in summer). Through STL decomposition, we can separate this seasonal fluctuation from the original time series, and obtain the long-term trend of pressure change (such as year-on-year increase or decrease) and residual terms (such as short-term fluctuations caused by random factors). Statistical tools such as autocorrelation function (ACF) and partial autocorrelation function (PACF) are used to identify the periodic components in the mixing chamber pressure time series, such as daily cycles, weekly cycles, etc., and analyze their cycle lengths and fluctuation amplitudes. For example, if there are daily periodic fluctuations in the mixing chamber pressure time series (such as pressure changes in the morning and evening every day), the ACF graph will show a high autocorrelation at the corresponding lag order. The PACF graph can further confirm the existence of such periodic fluctuations and determine its period length (such as 24 hours). At the same time, by analyzing the fluctuation amplitude of ACF and PACF, the intensity of periodic fluctuations can be understood. Anomaly detection algorithms based on time series, such as statistical methods (such as the 3σ principle) and machine learning methods (such as the isolated forest), identify outliers or mutation points in the time series of the pressure in the mixing chamber. For example, assume that the pressure value at a certain moment in the time series of the pressure in the mixing chamber suddenly increases or decreases, and there is a significant difference from the values ​​before and after. Through statistical methods (such as the 3σ principle), we can calculate the mean and standard deviation of the time series, and regard values ​​exceeding 3 times the standard deviation as outliers. Alternatively, through machine learning methods (such as the isolated forest), we can train a model to identify abnormal patterns in time series and mark out abnormal values ​​or mutation points. According to the results of time series analysis, typical patterns of pressure fluctuations in the mixing chamber, such as periodic fluctuations, trend changes, random fluctuations, etc., are identified, and the causes of their occurrence are analyzed. For example, if there are obvious periodic fluctuations in the time series of the mixing chamber pressure, and this fluctuation is closely related to the changes in external environmental factors (such as temperature and humidity), it can be inferred that this periodic fluctuation is caused by external environmental factors. In addition, if there are trend changes in the time series (such as increasing or decreasing year by year), it may be related to factors such as aging and insufficient maintenance of the air conditioning system. Extract key features (such as fluctuation amplitude, cycle length, trend slope, etc.) from the time series of the mixing chamber pressure, and select a suitable prediction model for preliminary modeling based on the characteristics of the time series. Commonly used prediction models include ARIMA models, LSTM neural networks, etc. For example, if the time series of the mixing chamber pressure has stable periodic and trend changes and no obvious nonlinear characteristics, the ARIMA model can be selected for modeling.If there are complex nonlinear relationships and long-term dependencies in the time series, LSTM neural network can be selected for modeling. In the preliminary modeling stage, the parameters and structure of the model need to be set according to the characteristics of the time series and the extracted features. Use historical data to train the prediction model, and verify the trained model by the holdout method (i.e., divide the data set into a training set and a test set). Evaluate the prediction accuracy and generalization ability of the model, and tune the model according to the verification results. For example, the historical data of the air mixing chamber pressure time series is divided into a training set and a test set, where the training set is used to train the prediction model and the test set is used to verify the prediction effect of the model. The prediction accuracy of the model is evaluated by calculating the error (such as mean square error MSE) between the predicted value and the actual value of the model on the test set. If the prediction error is large, the model needs to be tuned, such as adding features, adjusting the model structure or parameters, etc., to improve the prediction ability and generalization ability of the model.

[0034] The effects of the above technical solutions are as follows: by using time series decomposition technology (such as the STL method), the time series of the mixing chamber pressure is decomposed into trend terms, seasonal terms and residual terms, which can clearly identify the long-term trend of pressure changes, seasonal fluctuations and random disturbances, and help to deeply understand the dynamic characteristics of the mixing chamber pressure; using statistical tools such as autocorrelation function (ACF) and partial autocorrelation function (PACF), the periodic components in the mixing chamber pressure time series, such as daily cycles, weekly cycles, etc., can be accurately identified, and their cycle lengths and fluctuation amplitudes can be analyzed, providing an important basis for the subsequent prediction model construction; the anomaly detection algorithm based on time series can accurately identify the mixing chamber pressure. Outliers or mutation points in the time series may be caused by system failures, external interference and other factors. By timely detecting and processing these outliers, the accuracy and robustness of the prediction model can be improved; according to the results of time series analysis, the typical patterns of pressure fluctuations in the mixing chamber can be identified, such as periodic fluctuations, trend changes, random fluctuations, etc., and the causes of their occurrence can be analyzed, which is helpful for formulating targeted control strategies and prediction models; extract key features from the time series of the mixing chamber pressure, and select a suitable prediction model for preliminary modeling based on its characteristics, and then train and verify the model through historical data to evaluate the prediction accuracy and generalization ability of the model. Tuning the model according to the verification results can further improve the prediction performance of the model; through the above technical solutions, a more accurate and reliable prediction model for the mixing chamber pressure can be constructed, providing strong support for the control and regulation of the mixing chamber pressure. This helps to maintain the stability of the mixing chamber pressure and improve the comfort and energy efficiency of the indoor environment. The results of the prediction model can provide guidance for the actual control of the mixing chamber pressure, such as providing a basis for decision-making in adjusting the ratio of fresh air to return air and adjusting the valve opening, thereby optimizing the control system of the mixing chamber pressure.

[0035] In one embodiment of the present invention, S3 includes: S31. Regularly read various technical parameters of the refrigeration unit through the Modbus network protocol, and clean and format the read data; S32. Analyze the relationship between technical parameters and energy consumption through data mining techniques (such as association analysis and classification algorithms) to identify key factors affecting refrigeration efficiency; S33, using extreme value search algorithm combined with historical operation data and current environmental load, dynamically adjust the operation strategy of the refrigeration unit to maximize the overall COP (energy efficiency ratio); S34. Based on the cooling demand forecast, the predictive control algorithm is used to intelligently control the addition and subtraction operations of the host, and the preventive maintenance strategy of the refrigeration unit is implemented. The failure risk is predicted based on data analysis, and maintenance is carried out in advance.

[0036] The working principle of the above technical solution is: various technical parameters are read from the refrigeration unit regularly through the Modbus network protocol. These parameters include but are not limited to cooling capacity, power consumption, compressor status, condenser pressure, evaporator pressure, etc. The Modbus protocol is a serial communication-based protocol that is widely used in industrial control systems. It defines a communication mode between master and slave devices, allowing a master device (such as a computer or PLC) to communicate with a slave device (such as a sensor, actuator, etc.). The read data is cleaned, including removing duplicate data, filling missing data, etc.; the data is formatted to meet the requirements of subsequent analysis. This may include data standardization, that is, converting data of different dimensions into a unified dimension for comparison and analysis. Data mining techniques, such as association analysis and classification algorithms, are applied to conduct in-depth analysis of the relationship between technical parameters and energy consumption; association analysis can reveal potential connections between technical parameters, such as the relationship between cooling capacity and power consumption; classification algorithms can classify the status of the refrigeration unit into different categories, such as normal, faulty, etc., for subsequent analysis and processing. Through data mining technology, key factors affecting refrigeration efficiency are identified; these factors may include compressor status, condenser pressure, evaporator pressure, etc.; focusing on monitoring and optimizing these key factors can improve the energy efficiency ratio (COP) of the refrigeration unit; using the extreme value search algorithm, combined with historical operating data and current environmental load, the operation strategy of the refrigeration unit is dynamically adjusted; the extreme value search algorithm is an adaptive control algorithm that can search for control variables that make the system output reach the maximum or minimum value without knowing the specific form of the loss function; according to the results of the extreme value search algorithm, the operating parameters of the refrigeration unit, such as compressor speed, condenser fan speed, etc., are adjusted; the goal is to make the energy efficiency ratio (CO P) while meeting the current cooling demand; using predictive control algorithms, predict future cooling demand based on historical data and current environmental loads; this helps to adjust the operating strategy of the refrigeration unit in advance to avoid energy waste caused by over-cooling or under-cooling; intelligently control the addition and subtraction operations of the host according to the cooling demand prediction results; when it is predicted that the cooling demand will increase, start the standby unit in advance; when it is predicted that the cooling demand will decrease, shut down some units in a timely manner; predict failure risks based on data analysis and perform maintenance in advance; this includes monitoring and predicting the degree of wear of key components, as well as regular inspection and replacement of aging components; through preventive maintenance strategies, the efficient and stable operation of the refrigeration unit can be ensured and its service life can be extended.

[0037] The effects of the above technical solutions are as follows: various technical parameters of the refrigeration unit are read regularly through the Modbus network protocol to ensure the timeliness and accuracy of the data; the read data is cleaned and formatted, including data deduplication, data filling, data standardization, etc., to improve the quality and availability of the data; data mining technology (such as association analysis, classification algorithm) is used to conduct in-depth analysis of the relationship between technical parameters and energy consumption, and identify the key factors affecting refrigeration efficiency; this helps to optimize the refrigeration unit in a targeted manner, thereby improving its energy efficiency ratio (COP); the extreme value search algorithm is used to combine historical operating data and current environmental loads to dynamically adjust the operating strategy of the refrigeration unit; this dynamic adjustment can ensure that the refrigeration unit can maintain a high energy efficiency ratio under different working conditions, thereby reducing energy consumption; according to the cooling demand forecast, the predictive control algorithm is used to intelligently control the main Add or subtract machines; this can avoid energy waste caused by over-cooling or insufficient cooling, and realize intelligent scheduling and optimized operation of refrigeration units; predict failure risks based on data analysis and perform maintenance in advance; this can significantly reduce the failure rate of refrigeration units, extend their service life, and reduce downtime and maintenance costs caused by failures; through the above technical solutions, the energy efficiency ratio of refrigeration units is significantly improved, and energy consumption is effectively reduced; this helps to reduce the operating costs of enterprises and improve overall operating benefits; the implementation of preventive maintenance strategies can reduce the failure rate and downtime of refrigeration units; this helps to ensure the stable operation of the refrigeration system and improve the reliability and stability of the system; through intelligent data reading, processing and analysis methods, remote monitoring and intelligent management of refrigeration units can be realized; this helps to improve the management efficiency of enterprises and reduce labor costs.

[0038] In one embodiment of the present invention, the S33 includes: S331, selecting an extreme value search algorithm according to the characteristics of the refrigeration unit, and initializing the algorithm parameters; S332. Define an objective function with the goal of maximizing the overall COP, and use historical operating data and iterative calculations to continuously find a combination of refrigeration unit operating parameters that can achieve the optimal value of the objective function, such as condenser fan speed and evaporator water pump flow; S333. S334, evaluating the performance of the extreme value search algorithm through cross-validation, and optimizing the algorithm according to the evaluation results; S335, preprocessing the historical operation data, extracting key features from the preprocessed historical operation data, and using a feature selection method to screen out the features that have the greatest impact on the overall COP; S336. Use a clustering algorithm to group historical operation data, identify operation modes under different working conditions, monitor environmental parameters in real time, and use time series analysis to predict the refrigeration load in the future; S337. According to the load forecast result, an early warning threshold is set, and when the forecast load exceeds or falls below a certain range, a corresponding early warning mechanism is triggered; S338. Formulate an adjustment plan for the refrigeration unit operation strategy based on the optimization results of the extreme value search algorithm, the analysis results of the historical operation data, and the prediction results of the current environmental load; The formulated operation strategy is converted into control instructions and sent to the refrigeration unit in real time through the Modbus network protocol. The adjusted refrigeration unit operating parameters and overall COP are monitored in real time, and the adjustment effect is evaluated. If the expected goal is not achieved, iterative adjustments are made based on the feedback information until the optimal operating state is achieved.

[0039] The working principle of the above technical solution is: according to the characteristics and requirements of the refrigeration unit, select a suitable extreme value search algorithm. These algorithms may include gradient ascent / descent method, genetic algorithm, particle swarm optimization algorithm, etc. Each algorithm has its unique advantages and applicable scenarios, and needs to be selected according to actual conditions; initialize the parameters of the selected algorithm. For example, for the gradient ascent / descent method, it is necessary to set the learning rate; for the genetic algorithm, it is necessary to set the population size, crossover probability, mutation probability, etc.; for the particle swarm optimization algorithm, it is necessary to set the number of particles, inertia weight, learning factor, etc. The setting of these parameters will directly affect the performance and convergence speed of the algorithm; define the objective function with the overall COP maximization as the goal. This function takes the operating parameters of the refrigeration unit (such as condenser fan speed, evaporator water pump flow, etc.) as input and outputs the overall COP; using historical operating data, it continuously searches for the combination of refrigeration unit operating parameters that makes the objective function reach the optimal value through iterative calculation. In each iteration, the algorithm calculates the objective function value based on the current parameter combination, and updates the parameter combination according to the algorithm rules until the optimal solution is found or the upper limit of the number of iterations is reached; the performance of the extreme value search algorithm is evaluated through cross-validation. This includes indicators such as the convergence speed, search accuracy, and robustness of the algorithm. The evaluation results will serve as the basis for algorithm tuning; the algorithm will be tuned according to the evaluation results. This may include adjusting algorithm parameters, improving algorithm structure, or introducing new algorithm strategies. The purpose of tuning is to improve the performance and search efficiency of the algorithm; preprocessing operations are performed on historical operation data, including data cleaning, denoising, and normalization. These operations are intended to improve the quality and availability of data; feature selection methods (such as mutual information, recursive feature elimination, etc.) are used to filter out the features that have the greatest impact on the overall COP from the preprocessed historical operation data. These features will serve as the basis for subsequent analysis and optimization; clustering algorithms are used to group historical operation data and identify operating modes under different working conditions. These modes will help the system better understand the operating laws and characteristics of the refrigeration unit; real-time monitoring of environmental parameters (such as ambient temperature, humidity, etc.), and use time series analysis to predict the refrigeration load in the future. This helps the system make adjustments in advance to cope with possible load changes; set warning thresholds based on load forecast results. When the predicted load exceeds or falls below a certain range, the corresponding early warning mechanism is triggered; when the predicted load exceeds or falls below a certain range, the corresponding early warning mechanism is triggered; for example, assuming that the rated load of a refrigeration unit is 1000 kW, historical data show that the load mean is 800 kW and the standard deviation is 100 kW; based on the comprehensive design parameters and historical statistics, the early warning threshold range is set to 600 kW (800 kW - 2×100 kW) to 1000 kW (rated load); when the predicted load exceeds 1000 kW, a high load early warning is triggered, and when it is below 600 kW, a low load early warning is triggered.This helps the system to promptly identify potential problems and take corresponding countermeasures; based on the optimization results of the extreme value search algorithm, the analysis results of historical operating data, and the prediction results of the current environmental load, formulate adjustment plans for the operation strategy of the refrigeration unit. These plans include compressor start and stop strategies, condenser fan speed adjustment strategies, evaporator water pump flow adjustment strategies, etc.; the formulated operation strategies are converted into control instructions and sent to the refrigeration unit in real time through the Modbus network protocol. This ensures that the system can respond and adjust the operating status of the refrigeration unit in a timely manner; monitor the adjusted refrigeration unit operating parameters and overall COP in real time, and evaluate the adjustment effect. If the expected goal is not achieved, iterative adjustments are made based on the feedback information until the optimal operating state is achieved. This process is a closed-loop feedback system that can continuously optimize and improve the operating efficiency and energy efficiency ratio of the refrigeration unit.

[0040] The effect of the above technical solution is: through the application of the extreme value search algorithm, the technical solution can accurately find the combination of operating parameters of the refrigeration unit that maximizes the overall COP. Compared with traditional fixed parameter operation or simple manual adjustment, this refined adjustment method can significantly improve the energy efficiency ratio of the refrigeration unit, thereby reducing energy consumption and improving energy utilization efficiency. The clustering algorithm in the technical solution can identify the operating mode under different working conditions, which enables the system to automatically adjust the operating strategy according to different working conditions. This adaptive ability enhances the stability and reliability of the system and reduces the system fluctuations and failure risks caused by changes in working conditions; the future refrigeration load is predicted through time series analysis, and the warning threshold is set. When the predicted load exceeds or falls below a certain range, the corresponding warning mechanism is triggered. This intelligent warning function can detect potential load changes in advance, provide system maintenance personnel with sufficient time for troubleshooting and preventive maintenance, and avoid downtime and maintenance costs caused by failures; the control instructions in the technical solution are sent to the refrigeration unit in real time through the Modbus network protocol, realizing remote monitoring and intelligent management of the refrigeration unit. This management method not only improves management efficiency and reduces labor costs, but also improves the automation level of the system, allowing the system to respond more quickly to environmental changes and user needs; through accurate load forecasting and optimized operation strategies, this technical solution can reasonably allocate refrigeration resources according to actual load demand, avoiding energy waste caused by over-refrigeration or insufficient refrigeration. This resource optimization allocation method not only helps energy conservation and emission reduction, but also reduces the operating costs of enterprises and improves economic benefits; this technical solution integrates a variety of advanced technologies such as extreme value search algorithms, data mining technology, and time series analysis, which promotes technological innovation and industrial upgrading in the refrigeration industry. This innovation not only improves the performance and energy efficiency of refrigeration units, but also provides useful reference and reference for intelligent management and energy conservation and emission reduction in other industries.

[0041] In one embodiment of the present invention, the S4 includes: S41. According to the cooling water temperature difference, the real-time air conditioning load and the outdoor wet-bulb temperature, an intelligent algorithm (such as a fuzzy control algorithm or a neural network algorithm) is used to calculate the optimal air volume and the number of operating fans of the cooling tower; the optimal air volume is calculated by the following formula:

[0042] The number of running units is obtained by the following formula:

[0043] Among them, k1, k2, k3 are coefficients determined based on system characteristics and experience; T in Indicates the inlet temperature of cooling water; T out Indicates the outlet temperature of cooling water; T wb Indicates the outdoor wet bulb temperature; N max Indicates the maximum number of fans that can be operated in the cooling tower; Q max Indicates the maximum air volume of a single fan; L indicates the real-time load of the air conditioner; Indicates rounding up; S42, dynamically adjust the fan speed through frequency conversion speed control; and monitor the fan operation status; S43. Remotely monitor and warn of faults of cooling tower fans based on the Internet of Things technology, and adjust the supply of cold and heat sources in real time in combination with the load status of the fresh air unit; S44. Based on the outdoor temperature and humidity changes and on-site temperature and other parameters, a supply air temperature prediction model is established, and a predictive control algorithm is used to dynamically adjust the supply air temperature.

[0044] The working principle of the above technical solution is as follows: calculation of the optimal air volume and number of running units of cooling tower fans; input parameters, including cooling water temperature difference, real-time air conditioning load, and outdoor wet bulb temperature. Intelligent algorithms (such as fuzzy control algorithm and neural network algorithm) are used to analyze and process the input parameters. The output results calculate the optimal air volume and number of running units of cooling tower fans. Through variable frequency speed control, the fan speed is dynamically adjusted to adapt to different cooling needs. When the cooling demand increases, the fan speed is increased to enhance the cooling effect. When the cooling demand decreases, the fan speed is reduced to reduce energy consumption. Real-time monitoring of the fan operating status, including speed, vibration, temperature and other parameters. Through data analysis, potential failure risks can be discovered in time to prevent failures. Based on the Internet of Things technology, the cooling tower fan is remotely monitored and fault warning is carried out; the operating status and environmental parameters of the fan are collected in real time through sensors and communication equipment; the data is transmitted to the cloud platform for analysis and processing to achieve remote monitoring and fault warning; combined with the load status of the fresh air unit, the supply of cold and heat sources is adjusted in real time; when the load of the fresh air unit increases, the supply of cold and heat sources is increased to meet the demand; when the load of the fresh air unit decreases, the supply of cold and heat sources is reduced to save energy; through wind and water linkage control, the efficient and stable operation of the air conditioning system is ensured. When the temperature sensor in the strong cooling zone detects that the temperature is higher than the preset value, in the wind system, the control system increases the opening of the electric proportional integral regulating air valve in the corresponding area to increase the air supply volume; at the same time, in the water system, the frequency of the refrigeration pump is increased, the flow of chilled water is increased, and the opening of the wind cabinet water valve is increased, so that more cold energy is transferred to the air to quickly reduce the temperature in the strong cooling zone. When the temperature reaches the preset value, the wind system and the water system fine-tune the operating parameters according to the temperature fluctuation to maintain temperature stability. For weak cooling areas, when the temperature rises, the wind system appropriately increases the air valve opening, but the increase is smaller than that of the strong cooling area, and the temperature is adjusted by a smaller increase in air volume; the water system appropriately increases the frequency of the refrigeration pump and the opening of the water valve to provide an appropriate amount of cooling. Similarly, when the temperature reaches the set value, fine-tuning is performed. The comfort area has high requirements for temperature stability. When the temperature deviates from the preset value, the air system and the air valve opening, as well as the refrigeration pump frequency and water valve opening of the water system are adjusted slightly to avoid excessive temperature fluctuations that affect the comfort of passengers and staff. Based on parameters such as outdoor temperature and humidity changes and on-site temperature, a supply air temperature prediction model is established; a prediction model is established by analyzing historical data and real-time data; the model can predict the supply air temperature change trend in the future; a predictive control algorithm is used to dynamically adjust the supply air temperature; according to the results of the prediction model, the supply air temperature is adjusted in advance to adapt to environmental changes; by optimizing the supply air temperature, passenger comfort and energy saving effects are improved.

[0045] The effects of the above technical solutions are as follows: by adopting intelligent algorithms such as fuzzy control algorithms and neural network algorithms, the optimal air volume and number of cooling tower fans can be accurately calculated according to parameters such as cooling water temperature difference, real-time air conditioning load and outdoor wet bulb temperature. This refined adjustment method avoids the energy waste caused by traditional fixed parameter operation and significantly improves the energy efficiency ratio of the system; through variable frequency speed regulation technology, the fan speed is dynamically adjusted to adapt to different cooling needs. This adjustment method not only maintains the efficient operation of the cooling tower, but also significantly reduces energy and water consumption, achieving the goal of energy saving and consumption reduction; real-time monitoring of the fan operation status, including key parameters such as speed, vibration, and temperature, timely discovers potential fault risks, and prevents faults from occurring. This preventive maintenance method improves the stability and reliability of the system and reduces downtime and maintenance costs caused by faults; based on the Internet of Things technology, remote monitoring and fault warning of cooling tower fans are carried out. This remote management method enables operation and maintenance personnel to grasp the operating status of the system in real time, respond to and handle abnormal situations in a timely manner, and further enhance the stability and reliability of the system; combined with the load status of the fresh air unit, the supply of cold and heat sources is adjusted in real time to achieve wind and water linkage control. This control method enables the air conditioning system to be intelligently adjusted according to actual needs, avoiding energy waste caused by excessive cooling or heating, and improving the overall energy efficiency of the system; through the application of Internet of Things technology and intelligent algorithms, intelligent management of cooling tower fans and air conditioning systems is realized. This management method not only improves the automation level of the system, but also reduces labor costs, making operation and maintenance work more efficient and convenient; based on parameters such as outdoor temperature and humidity changes and on-site temperature, a supply air temperature prediction model is established. This model can predict the supply air temperature change trend in the future and provide a scientific basis for system adjustment; the predictive control algorithm is used to dynamically adjust the supply air temperature to adapt to environmental changes. This adjustment method not only improves the comfort of passengers, but also maximizes the energy saving effect. By accurately controlling the supply air temperature, energy waste and passenger discomfort caused by temperature fluctuations are avoided. The above optimal air volume calculation formula comprehensively considers multiple factors such as cooling water temperature difference, air conditioning real-time load and outdoor wet bulb temperature. The formula can calculate the optimal air volume that is more in line with actual needs, thereby improving the cooling efficiency of the cooling tower; accurate air volume control helps to avoid excessive or insufficient operation of the fan, thereby reducing energy consumption and operating costs. The above operating number calculation formula can calculate the minimum number of operating units required based on the optimal air volume and the maximum air volume of a single fan, thereby ensuring that the cooling tower maintains a high operating efficiency while meeting the cooling demand. By limiting the maximum number of operating units, the formula can prevent too many fans from running at the same time, thereby avoiding resource waste and increased energy consumption.

[0046] In one embodiment of the present invention, the S44 includes: Collect historical data of key parameters such as outdoor temperature and humidity, field temperature, and supply air temperature, and perform preprocessing operations. Use feature selection methods such as mutual information and recursive feature elimination to screen out the features that have the greatest impact on supply air temperature prediction. At the same time, use data mining technology to explore the potential relationship between features. Select a prediction model based on the characteristics of the data and prediction requirements, and use the preprocessed data to train the model; Verify the trained prediction model through cross-validation to evaluate the prediction performance and generalization ability of the model; optimize the model based on the verification results; Based on the supply air temperature prediction model, a predictive control algorithm is designed to predict the supply air temperature using real-time data. Based on the prediction results and the optimal supply air temperature setting value calculated by the control algorithm, the supply air temperature of the air conditioning system is adjusted in real time through the control system. Monitor the adjusted air supply temperature and passenger comfort feedback in real time to evaluate the adjustment effect; if the expected target is not achieved, make iterative adjustments based on the feedback information to optimize the predictive control algorithm and air supply temperature setting value; Collect passenger comfort feedback on the air-conditioning system, including perceived evaluations of temperature, humidity, wind speed, etc., and develop optimization strategies based on passenger comfort evaluation results.

[0047] The working principle of the above technical solution is: collect historical data of key parameters such as outdoor temperature and humidity, field temperature, and supply air temperature; perform preprocessing operations on the collected data, including data cleaning, missing value processing, outlier detection, etc.; use feature selection methods (such as mutual information, recursive feature elimination, etc.) to screen out the features that have the greatest impact on supply air temperature prediction; explore the potential relationship between features to provide a basis for subsequent modeling; select a suitable prediction model based on the characteristics of the data and prediction requirements. For example, for time series data, you can choose ARIMA, LSTM and other models; for complex nonlinear relationships, you can choose support vector machine, random forest and other machine learning models, or convolutional neural network, recurrent neural network and other deep learning models; use the preprocessed data to train the selected model to obtain a preliminary prediction model; verify the trained prediction model through cross-validation and other methods to evaluate the prediction performance and generalization ability of the model; according to the verification results, tune the model to improve the prediction accuracy and stability; based on the supply air temperature prediction model, design a predictive control algorithm, such as MPC (model predictive control) algorithm, adaptive control algorithm, etc.; the algorithm should take into account multiple factors such as changes in outdoor temperature and humidity, changes in on-site temperature, and passenger comfort requirements, and dynamically Adjust the supply air temperature set value; use real-time data to predict the supply air temperature, and adjust the supply air temperature of the air conditioning system in real time through the control system based on the prediction results and the optimal supply air temperature set value calculated by the control algorithm; monitor the adjusted supply air temperature and passenger comfort feedback in real time; evaluate the adjustment effect, and if the expected goal is not achieved, make iterative adjustments based on the feedback information to optimize the predictive control algorithm and supply air temperature set value; collect passenger comfort feedback on the air conditioning system, including perceived evaluations of temperature, humidity, wind speed, etc.; based on the passenger comfort evaluation results, formulate optimization strategies, such as adjusting the supply air temperature range, optimizing the air supply method (such as up and down air supply, side air supply, etc.), and increasing humidity control; integrate the optimization strategy into the predictive control algorithm to further improve passenger comfort and energy saving effects.

[0048] The effects of the above technical solutions are: by using feature selection methods such as mutual information and recursive feature elimination, the features that have the greatest impact on the supply air temperature prediction are screened out, and the potential relationship between the features is mined using data mining technology, which helps to improve the accuracy and robustness of the prediction model; according to the characteristics of the data and the prediction requirements, the appropriate prediction model is selected for training. Whether it is a time series analysis model (such as ARIMA, LSTM), a machine learning model (such as support vector machine, random forest) or a deep learning model (such as convolutional neural network, recurrent neural network), the prediction accuracy can be improved to a certain extent. Through cross-validation and model tuning, the prediction performance and generalization ability of the model can be further ensured; based on the supply air temperature prediction model, a predictive control algorithm (such as MPC algorithm, adaptive control algorithm, etc.) is designed, which can comprehensively consider multiple factors such as outdoor temperature and humidity changes, on-site temperature changes, and passenger comfort requirements, and dynamically adjust the supply air temperature set value. This intelligent control method can achieve precise control of the air conditioning system, avoid excessive cooling or heating, and thus significantly reduce energy consumption; by real-time monitoring of the adjusted supply air temperature and passenger comfort feedback, the adjustment effect is evaluated, and the predictive control algorithm and supply air temperature set value are iteratively adjusted and optimized based on the feedback information. This real-time feedback mechanism can ensure that the operation of the air-conditioning system is always maintained in the optimal state, further improving the energy-saving effect; collecting passengers' feedback on the comfort of the air-conditioning system, including temperature, humidity, wind speed and other perception evaluations, helps to understand the actual needs of passengers; based on the results of passenger comfort evaluation, formulate optimization strategies, such as adjusting the air supply temperature range, optimizing the air supply mode (such as up and down air supply, side air supply, etc.), and increasing humidity control. These optimization strategies can further improve passenger comfort and improve passenger satisfaction; through predictive control algorithms and real-time monitoring systems, the air-conditioning system is automatically controlled. This automatic control method reduces manual intervention and improves the operating efficiency and stability of the system; combined with advanced technologies such as data mining, machine learning / deep learning, the air-conditioning system is intelligently managed. This intelligent management method can realize real-time monitoring and early warning of the system status, discover potential problems in advance and take corresponding measures to ensure the safe and stable operation of the system.

[0049] In one embodiment of the present invention, S5 includes: S51. Calculate the optimal air valve opening through intelligent algorithm according to the changes of indoor and outdoor environment; S52, based on the air valve adjustment strategy combining remote control and local control, the air valve opening is adjusted, and the operating status of the air valve is monitored in real time through the Internet of Things technology; S53. Through the centralized intelligent management system of central air-conditioning equipment, real-time monitoring of equipment status, fault warning and remote operation and maintenance are carried out.

[0050] The working principle of the above technical solution is as follows: the system first monitors the changes in indoor and outdoor environments in real time through a sensor network, including key parameters such as temperature, humidity, and CO2 concentration; it uses intelligent optimization algorithms such as genetic algorithms and particle swarm algorithms to calculate the optimal air valve opening in the current environment based on the environmental parameters monitored in real time; these algorithms use iterative searches to find the optimal air valve opening configuration that can balance indoor and outdoor air circulation and maintain the air quality in the subway station; the calculated optimal air valve opening will serve as the basis for subsequent air valve adjustments; the system adopts an air valve adjustment strategy that combines remote control with local control; remote control allows operation and maintenance personnel to remotely adjust the air valve through the central management system to improve the flexibility and response speed of the adjustment; local control ensures that When remote control fails, the air valve can still be basically adjusted through the local controller to ensure the reliability of the system; using the Internet of Things technology, the system can monitor the operating status of the air valve in real time, including key parameters such as opening, speed, and power consumption; through data analysis, the system can promptly detect abnormal conditions in the operation of the air valve and prevent failures from occurring; through a centralized intelligent management system, the system can monitor the operating status of central air-conditioning equipment in real time, including the working conditions of key components such as compressors, condensers, and evaporators; the system can issue early warnings for equipment failures based on real-time monitoring data and notify operation and maintenance personnel in advance to handle them; operation and maintenance personnel can remotely debug, maintain, and troubleshoot equipment through the remote operation and maintenance platform, reducing the number of on-site repairs and improving operation and maintenance efficiency.

[0051] The effects of the above technical solutions are: through intelligent algorithms such as genetic algorithms and particle swarm algorithms, the optimal air valve opening is calculated in real time according to changes in indoor and outdoor environments (such as temperature, humidity, CO2 concentration, etc.); this intelligent adjustment method can accurately control the opening of the air valve, thereby effectively balancing indoor and outdoor air circulation and avoiding over-ventilation or under-ventilation. By optimizing air circulation, the freshness and quality of the air in the subway station can be maintained, providing passengers with a more comfortable environment; balanced air circulation helps to reduce the concentration of pollutants in the subway station, such as harmful gases such as CO2 and formaldehyde; maintaining good air quality in the subway station helps to improve passenger comfort and satisfaction, and also helps the safety and health of subway station operations. The air valve adjustment strategy combining remote control with local control improves the flexibility of air valve adjustment; remote control allows operation and maintenance personnel to remotely adjust the air valve through the central management system without having to go to the site for operation, which improves work efficiency; local control is used as a backup plan to ensure that when remote control fails, the air valve can still be basically adjusted through the local controller, ensuring the reliability of the system. Through the Internet of Things technology, the operating status of the air valve can be monitored in real time, including key parameters such as opening, speed, and power consumption; this real-time monitoring method helps to promptly detect abnormal conditions in the operation of the air valve, such as blockage, failure, etc., so that measures can be taken in advance to prevent or repair them; the application of Internet of Things technology has improved the reliability and safety of air valve regulation; through the centralized intelligent management system of central air-conditioning equipment, real-time monitoring of equipment status, fault warning and remote operation and maintenance are realized; this centralized management method helps operation and maintenance personnel to fully understand the operating status of the equipment and promptly discover and deal with potential faults; through the remote operation and maintenance function, operation and maintenance personnel can debug, maintain and troubleshoot the equipment without going to the site, reducing operation and maintenance costs and time costs; real-time monitoring of equipment status and early warning and operation and maintenance can help to promptly discover and deal with potential problems of the equipment, thereby extending the service life of the equipment; at the same time, the intelligent management system also helps to optimize the operating parameters of the equipment, improve the operating efficiency and energy-saving performance of the equipment; through the centralized intelligent management system, it can reduce operation and maintenance costs and time costs.

[0052] One embodiment of the present invention, as Figure 2 As shown, a full-link energy-saving system for central air conditioning in a subway station based on wind balance, the system includes: Data collection module: collects environmental data (such as temperature, humidity, passenger flow, etc.) inside and outside the subway station in real time through the sensor network, and uploads the collected environmental data to the central air-conditioning full-link energy-saving control platform; Dynamic adjustment module: The central air-conditioning full-link energy-saving control platform dynamically adjusts the air volume of the supply and return air ducts through fan control based on the collected data; Parameter reading module: reads all technical parameters of the refrigeration unit through the Modbus network protocol, uses the extreme value search algorithm to optimize the overall COP of the central air conditioner, and controls the host addition and subtraction according to the cooling demand; Supply adjustment module: adjust the air volume and number of cooling tower fans according to the cooling water temperature difference, real-time air conditioning load, outdoor wet-bulb temperature and other related conditions, and adjust the supply of cold and heat sources in real time according to the load status of the fresh air unit, perform wind and water joint adjustment, and adjust the supply air temperature according to the corresponding parameters such as outdoor temperature and humidity changes and on-site temperature; Opening adjustment module: adjust the air valve opening according to the changes in indoor and outdoor environment, and conduct centralized intelligent management of central air-conditioning equipment; Angle adjustment module: The opening angle of the electric proportional-integral damper in the corresponding area is adjusted based on the feedback signals from the high-precision temperature sensors installed in different areas of the platform layer.

[0053] The working principle of the above technical solution is as follows: a dense sensor network is arranged inside and outside the subway station, and these sensors can capture key environmental data such as temperature, humidity, and passenger flow inside and outside the subway station in real time. These data are uploaded to the central air-conditioning full-link energy-saving control platform in real time through a high-speed communication network. The control platform is a central system that integrates advanced algorithms and data processing capabilities, which can quickly analyze and process massive data; the central air-conditioning full-link energy-saving control platform dynamically adjusts the air volume of the supply and return air ducts according to the collected environmental data, including fine-tuning parameters such as fan speed and damper opening. By avoiding excessive entry of fresh air or excessive discharge of return air, it can ensure that the air circulation in the subway station not only meets the comfort requirements of passengers, but also maximizes the use of energy; through the Modbus network protocol, the central air-conditioning full-link energy-saving control platform can read all technical parameters of the refrigeration unit, including operating status, energy consumption, refrigeration efficiency, etc. Using the extreme value search algorithm, the platform can optimize the overall COP (energy efficiency ratio) of the central air-conditioning. According to the actual cooling demand of the subway station, the platform can intelligently control the operation of the host's adder and subtractor to ensure that the cooling efficiency is always in the optimal state; the cooling tower is an important part of the central air-conditioning system of the subway station, and its operating efficiency directly affects the energy consumption of the entire system. The central air-conditioning full-link energy-saving control platform intelligently adjusts the air volume and number of cooling tower fans according to the cooling water temperature difference, the real-time air-conditioning load, and the outdoor wet-bulb temperature. At the same time, the platform also adjusts the supply of cold and heat sources in real time according to the load status of the fresh air unit to achieve wind and water joint adjustment. In addition, the platform can also fine-tune the supply air temperature according to the corresponding parameters such as the changes in outdoor temperature and humidity and the on-site temperature to ensure that the air environment in the subway station is always in the best state; according to the changes in the indoor and outdoor environment, the central air-conditioning full-link energy-saving control platform can intelligently adjust the air valve opening by installing an electric proportional integral regulating air valve at the outlet position of each air valve, and setting high-precision temperature sensors in different areas such as the strong cold zone, weak cold zone, and comfort level of the platform layer. These temperature sensors can monitor the temperature changes in the area in real time and accurately, and feed the data back to the control system. The control system accurately adjusts the opening angle of the electric proportional integral regulating air valve in the corresponding area according to the feedback signals from these temperature sensors and the preset temperature values ​​in different areas, so as to achieve effective control of the temperature in different areas. When the air outlet is close to the wind cabinet, the area will be easier to reach the set temperature due to the large air volume. At this time, the system automatically reduces the opening angle of the electric regulating air valve so that the excess air volume can be transported to the farther air outlet area, thereby ensuring that the temperature of the farther air outlet area can also reach the set value. At the same time, a lower temperature value is set for the strong cooling zone. When the temperature feedback from the temperature sensor in the strong cooling zone does not meet the standard, the system appropriately increases the opening of the air valve in the area; and for the weak cooling zone and the comfort zone, the air valve opening is adjusted according to their respective preset temperatures.This not only ensures that the air circulation in the subway station meets the comfort needs of passengers, but also maximizes the use of energy. At the same time, the platform also centrally manages the central air-conditioning equipment, including equipment operation status monitoring, fault warning and remote control. Through centralized management, the overall optimization and energy saving and consumption reduction of the central air-conditioning system of the subway station can be achieved.

[0054] The effects of the above technical solutions are as follows: environmental data inside and outside the subway station are collected in real time through the sensor network and uploaded to the energy-saving control platform, so that the system can be accurately adjusted according to the actual situation; the dynamic adjustment of the fan control ensures the pressure balance of the mixing chamber, avoids excessive entry of fresh air or excessive discharge of return air, thereby reducing energy waste; the extreme value search algorithm is used to optimize the overall COP of the central air-conditioning, making the operation of the refrigeration unit more efficient and further improving the energy utilization efficiency; due to the improvement in energy utilization efficiency, the energy consumption of the central air-conditioning system of the subway station is significantly reduced, thereby reducing operating costs; centralized intelligent management makes the operation of the equipment more stable, reducing downtime and maintenance costs caused by equipment failures; by accurately adjusting the supply air temperature and air volume, the air environment in the subway station is made more comfortable, It meets the needs of passengers; Feng Shui coordination ensures that the supply of cold and heat sources matches the changes in the environment inside and outside the subway station, further improving the comfort of passengers; this technical solution contributes to energy conservation and emission reduction by improving energy utilization efficiency and reducing energy consumption; reduced energy consumption means reduced fossil fuel consumption and greenhouse gas emissions, which is beneficial to environmental protection and sustainable development; centralized intelligent management makes the operation of the subway station central air-conditioning system more intelligent and automated, reducing manual intervention and operational difficulty; real-time monitoring and early warning functions enable managers to discover and solve problems in a timely manner, improving management efficiency; dynamic adjustment of fan control, refrigeration unit optimization, Feng Shui coordination and other measures make the operation of the subway station central air-conditioning system more stable and reliable; reducing downtime caused by equipment failure and the impact on subway operations.

[0055] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A full-link energy-saving method for central air conditioning in subway stations based on wind balance, characterized in that: The method comprises: S1. Collect environmental data inside and outside the subway station in real time through the sensor network, and upload the collected environmental data to the central air-conditioning full-link energy-saving control platform; S2. The central air conditioning full-link energy-saving control platform dynamically adjusts the air volume of the supply and return air ducts through fan control based on the collected data; S3, read all technical parameters of the refrigeration unit through the Modbus network protocol, use the extreme value search algorithm to optimize the overall COP of the central air conditioner, and control the host addition and subtraction according to the cooling demand; S4. According to relevant conditions, adjust the air volume and number of cooling tower fans, adjust the supply of cold and hot sources in real time according to the load status of the fresh air unit, conduct wind and water joint adjustment, and adjust the supply air temperature according to the corresponding parameters; S5. Adjust the air valve opening according to the changes in the indoor and outdoor environment, and conduct centralized intelligent management of central air-conditioning equipment; S6. The opening angle of the electric proportional-integral regulating air valve in the corresponding area is adjusted based on the feedback signals from the high-precision temperature sensors installed in different areas of the platform layer.

2. According to claim 1, the full-link energy-saving method for central air conditioning in subway stations based on wind balance is characterized in that: Said S1 comprises: S11, collect environmental data through a high-precision sensor network deployed inside and outside the subway station; S12. Preliminary processing of the collected environmental data is performed through data verification and anomaly detection mechanisms, and outliers are eliminated; S13, uploading the preliminarily processed environmental data to the central air-conditioning full-link energy-saving control platform via a wired or wireless network; S14, storing the real-time data and the historical data respectively through the real-time database and the historical database established on the control platform; S15. Use machine learning algorithms to mine and analyze stored data, identify environmental change trends, and predict future load demands.

3. According to claim 2, the full-link energy-saving method for central air conditioning in subway stations based on wind balance is characterized in that: The S15 comprises: S151. Based on the characteristics of the environmental data, select features that have an important impact on the load demand forecast; perform feature extraction, and perform normalization or standardization processing on the feature data; S152. Select a machine learning algorithm based on the characteristics of historical data and forecasting requirements; S153, using the processed stored data to train the selected machine learning model, and verifying the trained model through cross-validation to evaluate the generalization ability of the model on unknown data; S154. Use time series analysis or cluster analysis methods to identify the changing trends of environmental parameters in the stored data, and predict the load demand in the future based on the trained machine learning model; S155, quantitatively evaluate the prediction results, fine-tune the model according to the evaluation results, and formulate energy-saving strategies based on the prediction results; S156. Feedback the prediction results to the central air-conditioning full-link energy-saving control platform in real time, and adjust the operating status of the air-conditioning system in real time according to the predicted load demand; S157. Collect feedback data based on the actual operating results of the air-conditioning system and continuously optimize and iterate the prediction model.

4. According to claim 1, the full-link energy-saving method for central air conditioning in subway stations based on wind balance is characterized in that: The S2 comprises: S21. Based on the collected environmental data, the current air volume values ​​of the supply air duct and the return air duct are calculated using a fluid mechanics model; and the fan speed and inverter output are dynamically adjusted through a PID control algorithm; S22, monitor the pressure of the air mixing chamber in real time, and use fuzzy control strategy or neural network control strategy to dynamically adjust the ratio of fresh air and return air according to pressure changes; S23. Based on the pressure balance simulation model, simulate the wind pressure changes under different working conditions and verify the effectiveness of the control strategy; S24. Based on the adaptive learning mechanism, the control parameters are continuously optimized according to the operating data, and regular pressure balance tests are carried out to evaluate the system performance.

5. According to claim 4, the full-link energy-saving method for central air conditioning in subway stations based on wind balance is characterized in that: The S22 comprises: S221, using a pressure sensor installed in the air mixing chamber to monitor the pressure change in the air mixing chamber in real time, verifying the collected pressure data, and identifying and eliminating abnormal values; S222. Use time series analysis technology to analyze the changing trend of the air mixing chamber pressure and identify the laws and characteristics of pressure fluctuations; S223. According to the changing characteristics of the air mixing chamber pressure and the control requirements, the applicability of the fuzzy control strategy and the neural network control strategy is evaluated, and the control strategy is selected; S224, initializing parameters of the selected control strategy, and verifying the adaptability and stability of the control strategy under different working conditions through simulation; S225, according to the change of the pressure in the air mixing chamber, the optimal ratio of fresh air to return air is calculated by using the control strategy, and the calculated ratio of fresh air to return air is converted into a control instruction of the actuator; S226. Monitor the adjusted air mixing chamber pressure in real time and evaluate the adjustment effect. If the expected target is not achieved, make iterative adjustments until the pressure is stable.

6. According to claim 1, the full-link energy-saving method for central air conditioning in subway stations based on wind balance is characterized in that: The S3 includes: S31. Regularly read various technical parameters of the refrigeration unit through the Modbus network protocol, and clean and format the read data; S32. Analyze the relationship between technical parameters and energy consumption through data mining technology to identify key factors affecting refrigeration efficiency; S33, using extreme value search algorithm combined with historical operation data and current environmental load, dynamically adjust the operation strategy of the refrigeration unit to maximize the overall COP; S34. Based on the cooling demand forecast, the predictive control algorithm is used to intelligently control the addition and subtraction operations of the host, and the preventive maintenance strategy of the refrigeration unit is implemented. The failure risk is predicted based on data analysis, and maintenance is carried out in advance.

7. The full-link energy-saving method for central air conditioning in subway stations based on wind balance according to claim 6 is characterized in that: The S33 comprises: S331, selecting an extreme value search algorithm according to the characteristics of the refrigeration unit, and initializing the algorithm parameters; S332, define an objective function with the goal of maximizing the overall COP, and use historical operating data and iterative calculation to continuously find a combination of operating parameters of the refrigeration unit that makes the objective function reach the optimal value; S333, evaluating the performance of the extreme value search algorithm through cross-validation, and optimizing the algorithm according to the evaluation results; S334, preprocessing the historical operation data, extracting key features from the preprocessed historical operation data, and using a feature selection method to screen out the features that have the greatest impact on the overall COP; S335. Use a clustering algorithm to group historical operation data, identify operation modes under different working conditions, monitor environmental parameters in real time, and use time series analysis to predict the refrigeration load in the future; S336. According to the load forecast result, an early warning threshold is set, and when the forecast load exceeds or falls below a certain range, a corresponding early warning mechanism is triggered; S337. Formulate an adjustment plan for the refrigeration unit operation strategy based on the optimization results of the extreme value search algorithm, the analysis results of the historical operation data, and the prediction results of the current environmental load; S338. Convert the formulated operation strategy into control instructions, send them to the refrigeration unit in real time through the Modbus network protocol, monitor the adjusted operation parameters and overall COP of the refrigeration unit in real time, evaluate the adjustment effect, and if the expected goal is not achieved, make iterative adjustments based on the feedback information until the optimal operation state is achieved.

8. The full-link energy-saving method for central air conditioning in subway stations based on wind balance according to claim 1 is characterized in that: The S4 comprises: S41. Calculate the optimal air volume and number of cooling tower fans using an intelligent algorithm based on the cooling water temperature difference, the real-time air conditioning load, and the outdoor wet-bulb temperature; S42, dynamically adjust the fan speed through frequency conversion speed control, and monitor the fan operation status; S43. Remotely monitor and warn of faults of cooling tower fans based on the Internet of Things technology, and adjust the supply of cold and heat sources in real time in combination with the load status of the fresh air unit; S44. Based on the changes in outdoor temperature and humidity and the on-site temperature, a supply air temperature prediction model is established, and a predictive control algorithm is used to dynamically adjust the supply air temperature.

9. The full-link energy-saving method for central air conditioning in subway stations based on wind balance according to claim 1 is characterized in that: The S5 comprises: S51. Calculate the optimal air valve opening through intelligent algorithm according to the changes of indoor and outdoor environment; S52, based on the air valve adjustment strategy combining remote control and local control, the air valve opening is adjusted, and the operating status of the air valve is monitored in real time through the Internet of Things technology; S53. Through the centralized intelligent management system of central air-conditioning equipment, real-time monitoring of equipment status, fault warning and remote operation and maintenance are carried out.

10. The full-link energy-saving system of central air conditioning in subway stations based on wind balance is characterized by: The system comprises: Data collection module: collects environmental data inside and outside the subway station in real time through the sensor network, and uploads the collected environmental data to the central air-conditioning full-link energy-saving control platform; Dynamic adjustment module: The central air-conditioning full-link energy-saving control platform dynamically adjusts the air volume of the supply and return air ducts through fan control based on the collected data; Parameter reading module: reads all technical parameters of the refrigeration unit through the Modbus network protocol, uses the extreme value search algorithm to optimize the overall COP of the central air conditioner, and controls the host addition and subtraction according to the cooling demand; Supply adjustment module: adjust the air volume and number of cooling tower fans according to the cooling water temperature difference and related conditions, adjust the supply of cold and hot sources in real time according to the load status of the fresh air unit, conduct wind and water joint adjustment, and adjust the supply air temperature according to the corresponding parameters; Opening adjustment module: adjust the air valve opening according to the changes in indoor and outdoor environment, and conduct centralized intelligent management of central air-conditioning equipment; Angle adjustment module: The opening angle of the electric proportional-integral damper in the corresponding area is adjusted based on the feedback signals from the high-precision temperature sensors installed in different areas of the platform layer.

Images