Intelligent partition energy-saving control method for air conditioner pipe network system

Abstract

The invention discloses an intelligent partition energy-saving control method for an air conditioner pipe network system, which comprises the following steps of: 1, acquiring multi-source data, and acquiring environmental parameters, load parameters and meteorological parameters in real time; 2, dynamic partition modeling is carried out, and a pipe network is divided into intelligent partitions capable of being regulated and controlled independently based on regional load characteristic differences and real-time data; step 3, load prediction and optimization: predicting cooling and heating load demands, and generating a dynamic energy supply strategy; 4, cooperative control is carried out, and precise regulation and control of partitioned cooling / heating are achieved through a central controller in a coordinated mode; 5, feedback and self-adaptive adjustment are carried out, and partition boundaries and energy supply parameters are dynamically optimized based on real-time environment parameters and user comfort level feedback; according to the method, the inner area, the outer area and the functional area are divided through dynamic zoning modeling to predict the load demand, the problems of cold and heat offset and energy consumption waste caused by fixed zoning of a traditional air conditioning system are solved, and cooperative regulation and control are achieved through the variable air volume terminal, the variable frequency pump and the electric two-way valve.

Description

Technical Field

[0001] This invention relates to the field of air conditioning pipe network control technology, and in particular to an intelligent zoned energy-saving control method for air conditioning pipe network systems. Background Technology

[0002] Air conditioning piping systems are a core component of HVAC engineering, used for transporting chilled and hot water media. Their function is to efficiently distribute the cooling or heating energy generated by heat sources to various terminal devices (such as fan coil units and fresh air handling units) through a piping network, thereby regulating indoor temperature and maintaining a comfortable environment. Traditional air conditioning piping systems typically consist of chilled / hot water units, circulating water pumps, supply and return water pipes, valves, and accessories, employing a two- or four-pipe design and achieving hydraulic balance through parallel or parallel flow layouts. The chilled / hot water system is driven by a primary or secondary pump, combined with variable frequency control technology to adapt to varying flow demands and improve energy efficiency. In recent years, capillary network technology has been widely used as a new type of air conditioning terminal system, achieving cooling and heating through radiative heat transfer. This system features a large heat exchange area, low resistance, and significant energy-saving effects, and can be flexibly installed on walls, ceilings, or floors, significantly saving building space. Furthermore, the piping system also requires supporting cooling water circulation, condensate drainage, pipe insulation and corrosion protection measures, and equipment such as balancing valves and differential pressure bypass valves to ensure stable system operation.

[0003] In existing technologies, the control of air conditioning duct systems generally adopts a fixed zoning mode, that is, the zoning boundaries and energy supply strategies are pre-set based on the building structure. The differences in load characteristics between the inner and outer zones are not fully considered. The outer zone is significantly affected by solar radiation and indoor-outdoor temperature differences, while the load of the inner zone is mainly driven by heat dissipation from people and equipment. Traditional fixed zoning cannot dynamically adjust the energy supply strategies of the inner and outer zones, resulting in a "winter-summer cooling-heat offsetting" phenomenon in the outer zone during transitional seasons or extreme weather, leading to serious energy waste. Functional zone coordination is lacking: the differences in load demand and usage time between special functional areas such as conference rooms and computer rooms and ordinary office areas are not distinguished, often resulting in "energy supply to empty rooms" or "over-energy supply". Traditional methods rely on a single model to predict the load, which cannot effectively integrate the spatiotemporal characteristics of environmental parameters, load parameters and meteorological parameters. The prediction error generally exceeds 10%, making it difficult to support accurate energy supply. Traditional systems only adjust temperature and humidity through temperature sensors, ignoring the impact of carbon dioxide concentration, humidity, light intensity and other factors on human comfort. For example, when the conference room is crowded, the carbon dioxide concentration exceeds the standard, but the temperature may have dropped below the set value, resulting in an uncomfortable experience of "too cold but lack of oxygen". Therefore, this invention proposes an intelligent zoning energy-saving control method for air conditioning pipe network system to solve the problems existing in the prior art. Summary of the Invention

[0004] To address the aforementioned problems, the present invention aims to propose an intelligent zoning energy-saving control method for air conditioning pipe network systems. This method uses dynamic zoning modeling to divide the system into inner, outer, and functional zones to predict load demand, thus solving the problems of cooling and heating offsetting and energy waste caused by fixed zoning in traditional air conditioning systems. Furthermore, it addresses the issues of energy redundancy and response delay through the coordinated control of variable air volume terminals, variable frequency pumps, and electric two-way valves. Finally, it resolves the problems of uneven comfort and lag in manual adjustments through a closed-loop mechanism based on carbon dioxide concentration and user behavior feedback.

[0005] To achieve the objectives of this invention, the invention is implemented through the following technical solution: an intelligent zoned energy-saving control method for an air conditioning pipe network system, comprising the following steps:

[0006] Step 1: Multi-source data acquisition. Real-time collection of environmental parameters, load parameters, and meteorological parameters of each zone within the building is achieved through an Internet of Things (IoT) sensor network.

[0007] Step 2: Dynamic zoning modeling. Based on regional load characteristics differences and real-time data, a clustering algorithm is used to divide the pipeline network into independently controllable intelligent zones.

[0008] Step 3: Load forecasting and optimization. Combining historical data and machine learning models, predict the heating and cooling load demand of each zone and generate dynamic energy supply strategies.

[0009] Step 4: Collaborative control. The central controller coordinates the variable air volume terminals, electric two-way valves, and variable frequency pumps to achieve precise control of zoned cooling / heating.

[0010] Step 5: Feedback and Adaptive Adjustment. Based on real-time environmental parameters and user comfort feedback, dynamically optimize zone boundaries and power supply parameters to form closed-loop control.

[0011] Further improvements are made in the following: the dynamic zoning modeling in step two includes the division of inner and outer zones and the division of functional zones. The division of inner and outer zones is based on the heat transfer characteristics of the building envelope and the distribution of personnel density, and the k-means algorithm is used to divide the inner and outer zones. The division of functional zones is based on the area usage time and equipment type to set independent control strategies.

[0012] A further improvement lies in the fact that the k-means clustering process in the dynamic partitioning modeling includes:

[0013] S1. Initialize the centroids. Randomly select k data points as the initial cluster centers, ensuring that k ≤ n and that the initial centroids are not repeated.

[0014] S2. Assign partitions: Calculate the distance from each data point to the k centroids, and assign the data point to the cluster corresponding to the nearest centroid.

[0015] S3. Update the centroid by recalculating the mean load characteristics of the two types of partitions as the new centroid.

[0016] S4. Convergence judgment: After five iterations, the centroid stabilizes, and the inner and outer regions are divided according to the centroid characteristics.

[0017] A further improvement is made in the following: the machine learning model in step three includes an input layer, a hidden layer, and an output layer. The input layer is used to input environmental parameters, load parameters, and meteorological parameters. The hidden layer uses a convolutional neural network to extract spatiotemporal features. The output layer is used to generate the cooling / heating power and temperature and humidity setpoints for each zone.

[0018] A further improvement is that the hidden layer includes a CNN sub-layer and an LSTM sub-layer, wherein the CNN sub-layer is used to extract spatiotemporal features and the LSTM sub-layer is used to capture time series dependencies.

[0019] Further improvements are made in the following aspects: the coordinated control in step four includes variable air volume terminal adjustment, water system dynamic balancing, and variable frequency pump coordination. The variable air volume terminal adjustment adjusts the air supply volume according to the load side of the zone, and the air volume adjustment range is 30% to 100% of the rated value. The water system dynamic balancing adjusts the flow rate of hot and cold water through an electric two-way valve. The variable frequency pump coordination adjusts the speed of the variable frequency pump based on the feedback of water supply pressure and temperature difference.

[0020] A further improvement is that: in step five, the comfort feedback is assessed by evaluating comfort levels based on carbon dioxide concentration and user dwell time, and the temperature and humidity of each zone are dynamically adjusted.

[0021] The beneficial effects of this invention are as follows: This invention divides the load demand into inner zone, outer zone and functional zone by dynamic zoning modeling, which solves the problems of cooling and heating offset and energy waste caused by fixed zoning in traditional air conditioning systems. Through the coordinated control of variable air volume terminals, variable frequency pumps and electric two-way valves, it solves the problems of energy supply redundancy and response delay. Through the closed-loop mechanism of carbon dioxide concentration and user behavior feedback, it solves the problems of uneven comfort and lag in manual adjustment. Attached Figure Description

[0022] Figure 1 This is a flowchart of the steps of the present invention. Detailed Implementation

[0023] To enhance understanding of the present invention, the present invention will be further described in detail below with reference to embodiments. These embodiments are only used to explain the present invention and do not constitute a limitation on the scope of protection of the present invention.

[0024] The air conditioning piping system is the core network in HVAC engineering used to transport cold and hot media. Its core function is to efficiently distribute the cooling or heating generated by cold and heat sources to various terminal devices, thereby regulating indoor temperature and maintaining a comfortable environment. Energy-saving control of the air conditioning piping system achieves minimum energy consumption and maximum energy efficiency during the transport of cold and hot media through intelligent technology, dynamic adjustment strategies, and system optimization. Its core lies in solving problems such as hydraulic imbalance and low distribution efficiency in traditional piping systems. By combining cold and heat source coordination and data-driven management, a multi-dimensional energy-saving system is formed. This system mainly consists of a refrigerant circulation system, including the indoor unit evaporator, outdoor unit condenser, compressor, and valve components, which achieves heat absorption and release through refrigerant phase change. For example, after the evaporator absorbs indoor heat, the refrigerant is pressurized and heated by the compressor, and then released to the outdoor environment through the condenser. The air duct transmission system consists of air ducts, static pressure boxes, regulating valves, supply / return air vents, etc., responsible for transporting the treated air to various areas. The ductwork comes in various materials and requires a plenum chamber to balance the air pressure, while regulating valves control the airflow distribution. The water circulation system uses chiller / hot water units, circulating water pumps, and insulated pipes. The pumps drive water through evaporators or condensers to transfer heat or cold. Typical designs include two-pipe or four-pipe systems to accommodate variable flow requirements.

[0025] Current air conditioners generally use inverter technology to achieve energy-saving control. Inverter technology adjusts the compressor speed through an inverter to dynamically match the air conditioner's cooling / heating capacity, thereby significantly reducing energy consumption while ensuring comfort. Its specific working process can be divided into four main parts: core components, speed regulation mechanism, load response logic, and energy-saving mechanism. The core of inverter technology is the inverter, which converts the fixed-frequency AC power input from the power grid into variable-frequency AC power, thereby controlling the compressor speed. The compressor speed directly determines the cooling / heating output. The core advantage of inverter technology is that it adjusts the compressor speed in real time according to changes in indoor load, avoiding the drawbacks of frequent start-stop cycles in fixed-frequency air conditioners. The energy-saving effect of inverter technology mainly comes from avoiding the losses from frequent start-stop cycles and efficient operation matching the load. Fixed-frequency air conditioners control room temperature by "on / off" the compressor, which needs to overcome inertial torque each time it starts, consuming a lot of energy. In contrast, inverter air conditioners use a "low-frequency maintenance" mode to keep the compressor running continuously, avoiding start-up losses. The compressor's optimal efficiency point is usually within 70%-90% of its rated speed. Fixed-frequency air conditioners operate at full load (100% speed) or zero load (shutdown) for extended periods, resulting in lower efficiency. In contrast, variable-frequency air conditioners can adjust their speed according to the load, ensuring the compressor always operates within its high-efficiency range, thus further improving energy efficiency.

[0026] Based on this, according to Figure 1 As shown, this embodiment provides an intelligent zoned energy-saving control method for an air conditioning pipe network system, including the following steps:

[0027] Step 1: Multi-source data acquisition. Environmental, load, and meteorological parameters for each zone within the building are collected in real-time via an IoT sensor network. Environmental parameters are monitored in real-time by installing temperature, humidity, and carbon dioxide concentration sensors on the ceilings or walls of each zone. Load parameters are assessed by using infrared human body sensors to count occupants and current transformers to monitor equipment power and calculate heat dissipation. Meteorological parameters are collected by deploying an outdoor weather station on the roof to collect temperature, humidity, and solar radiation intensity data. This sensor data is then wirelessly transmitted to the central controller. A dynamic database is built using this multi-dimensional data, providing a foundation for subsequent zone control and management. High-frequency sampling ensures timely response to load changes.

[0028] Step 2: Dynamic zoning modeling. Based on regional load characteristics differences and real-time data, a clustering algorithm is used to divide the pipeline network into independently controllable intelligent zones. The inner / outer zones are dynamically divided through k-means clustering to adapt to load changes, distinguish the usage characteristics of different areas, achieve differentiated control, and avoid the cooling and heating offsetting caused by traditional fixed zoning.

[0029] Dynamic zoning modeling includes the division of inner and outer zones and the division of functional zones. The division of inner and outer zones is based on the heat transfer characteristics of the building envelope and the distribution of personnel density, and the k-means algorithm is used to divide the inner and outer zones. The division of functional zones is based on the usage time of the area and the type of equipment to set independent control strategies.

[0030] The k-means clustering process in dynamic partitioning modeling includes:

[0031] S1. Initialize the centroids. Randomly select k data points as the initial cluster centers, ensuring that k ≤ n and that the initial centroids are not repeated.

[0032] S2. Assign partitions: Calculate the distance from each data point to the k centroids, and assign the data point to the cluster corresponding to the nearest centroid.

[0033] S3. Update the centroid by recalculating the mean load characteristics of the two types of partitions as the new centroid.

[0034] S4. Convergence judgment: After five iterations, the centroid stabilizes, and the inner and outer regions are divided according to the centroid characteristics.

[0035] Step 3: Load forecasting and optimization. Combining historical data and machine learning models, predict the heating and cooling load demand of each zone and generate dynamic energy supply strategies. By combining spatiotemporal characteristics and time series dependencies, reduce prediction errors, provide data support for zoned energy supply, and discover redundant energy supply links through prediction to tap energy-saving potential.

[0036] The machine learning model consists of an input layer, hidden layers, and an output layer. The input layer takes in environmental parameters, load parameters, and meteorological parameters. The hidden layers use convolutional neural networks to extract spatiotemporal features. The output layer generates the cooling / heating power and temperature and humidity setpoints for each zone. The hidden layers include CNN sub-layers and LSTM sub-layers. The CNN sub-layer extracts spatiotemporal features, while the LSTM sub-layer captures time-series dependencies. The CNN sub-layer typically consists of the following components: convolutional layers, which scan the input data using multiple convolutional kernels to calculate the dot product of local regions and generate feature maps; activation functions, commonly ReLU, which introduces non-linear transformations and enhances the model's ability to express complex patterns; and pooling layers, which reduce the spatial dimensionality of the feature maps through max pooling or average pooling, reducing computation while retaining key information. The core component of the LSTM sub-layer is the memory unit, which includes the following gating mechanisms: a forget gate, which determines which old information to discard from the memory unit; an input gate, which controls the writing of new information; and an output gate, which generates the output based on the current state of the memory unit. First, the raw load, environmental and meteorological data are processed through a CNN sub-layer to extract local patterns within the time window. The feature sequence output by the CNN is then input into an LSTM sub-layer, which uses its memory units to capture long-term patterns in the time series.

[0037] Step 4: Collaborative control. The central controller coordinates the variable air volume terminals, electric two-way valves, and variable frequency pumps to achieve precise control of zoned cooling / heating.

[0038] The coordinated control includes variable air volume (VAV) terminal regulation, dynamic water system balancing, and variable frequency pump coordination. VAV terminal regulation adjusts the air volume based on zone load and ranges from 30% to 100% of the rated value. Dynamic water system balancing regulates hot and cold water flow through an electric two-way valve, independently controlling the flow rates. Hot water is supplied to external zones in winter, and cold water in summer. The electric two-way valve adjusts its opening based on temperature feedback to maintain the set temperature and humidity. Variable frequency pump coordination adjusts the pump speed based on supply water pressure and temperature difference feedback. The primary pump adjusts its speed based on the supply and return water temperature difference and the most unfavorable loop pressure difference, while the secondary pump dynamically adjusts its speed according to zone flow requirements to avoid large flow rates with small temperature differences.

[0039] Step 5: Feedback and Adaptive Adjustment. Based on real-time environmental parameters and user comfort feedback, dynamically optimize zone boundaries and power supply parameters to form a closed-loop control. Correct model deviations through real-time feedback and dynamically adjust parameters to meet individual needs.

[0040] Comfort feedback assesses comfort levels based on carbon dioxide concentration and user dwell time, dynamically adjusting temperature and humidity in different zones. If the carbon dioxide concentration in the meeting room exceeds 1200 ppm, the system automatically increases fresh air volume and adjusts the supply air volume.

[0041] This intelligent zoning energy-saving control method for the air conditioning network system collects environmental, load, and meteorological parameters of each zone within the building in real time through an IoT sensor network to construct a dynamic database. It uses a k-means clustering algorithm to dynamically divide the internal and external zones and functional areas, achieving differentiated control by combining the characteristics of the building envelope and usage. Based on a hybrid model that integrates spatiotemporal features and time series dependencies, it predicts the cooling and heating load demands of each zone and generates dynamic energy supply strategies. Through the coordinated regulation of variable air volume terminals, electric two-way valves, and variable frequency pumps, it achieves precise matching of air volume, hot and cold water flow rates, and water supply pressure within each zone. Finally, based on comfort feedback such as carbon dioxide concentration and user dwell time, it dynamically optimizes zone boundaries and energy supply parameters, effectively reducing energy consumption for cooling and heating offsetting and improving spatial comfort.

[0042] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent zoned energy-saving control of an air conditioning duct network system, characterized in that, Includes the following steps: Step 1: Multi-source data acquisition. Real-time collection of environmental parameters, load parameters, and meteorological parameters of each zone within the building is achieved through an Internet of Things (IoT) sensor network. Step 2: Dynamic zoning modeling. Based on regional load characteristics differences and real-time data, a clustering algorithm is used to divide the pipeline network into independently controllable intelligent zones. Step 3: Load forecasting and optimization. Combining historical data and machine learning models, predict the heating and cooling load demand of each zone and generate dynamic energy supply strategies. Step 4: Collaborative control. The central controller coordinates the variable air volume terminals, electric two-way valves, and variable frequency pumps to achieve precise control of zoned cooling / heating. Step 5: Feedback and Adaptive Adjustment. Based on real-time environmental parameters and user comfort feedback, dynamically optimize zone boundaries and power supply parameters to form closed-loop control.

2. The intelligent zoned energy-saving control method for an air conditioning pipe network system according to claim 1, characterized in that: The dynamic zoning modeling in step two includes the division of inner and outer zones and the division of functional zones. The division of inner and outer zones is based on the heat transfer characteristics of the building envelope and the distribution of personnel density, and the k-means algorithm is used to divide the inner and outer zones. The division of functional zones is based on the area usage time and equipment type to set independent control strategies.

3. The intelligent zoned energy-saving control method for an air conditioning pipe network system according to claim 2, characterized in that: The k-means clustering process in the dynamic partitioning modeling includes: S1. Initialize the centroids. Randomly select k data points as the initial cluster centers, ensuring that k ≤ n and that the initial centroids are not repeated. S2. Assign partitions: Calculate the distance from each data point to the k centroids, and assign the data point to the cluster corresponding to the nearest centroid. S3. Update the centroid by recalculating the mean load characteristics of the two types of partitions as the new centroid. S4. Convergence judgment: After five iterations, the centroid stabilizes, and the inner and outer regions are divided according to the centroid characteristics.

4. The intelligent zoned energy-saving control method for an air conditioning pipe network system according to claim 1, characterized in that: The machine learning model in step three includes an input layer, a hidden layer, and an output layer. The input layer is used to input environmental parameters, load parameters, and meteorological parameters. The hidden layer uses a convolutional neural network to extract spatiotemporal features. The output layer is used to generate the cooling / heating power and temperature and humidity setpoints for each zone.

5. The intelligent zoned energy-saving control method for an air conditioning pipe network system according to claim 4, characterized in that: The hidden layer includes a CNN sublayer and an LSTM sublayer. The CNN sublayer is used to extract spatiotemporal features, and the LSTM sublayer is used to capture time series dependencies.

6. The intelligent zoned energy-saving control method for an air conditioning pipe network system according to claim 1, characterized in that: The coordinated control in step four includes variable air volume terminal adjustment, water system dynamic balancing, and variable frequency pump coordination. The variable air volume terminal adjustment adjusts the air supply volume according to the zone load and the air volume adjustment range is 30% to 100% of the rated value. The water system dynamic balancing adjusts the flow rate of hot and cold water through an electric two-way valve. The variable frequency pump coordination adjusts the speed of the variable frequency pump based on the feedback of water supply pressure and temperature difference.

7. The intelligent zoned energy-saving control method for an air conditioning pipe network system according to claim 1, characterized in that: The comfort feedback in step five assesses comfort by considering carbon dioxide concentration and user dwell time, and dynamically adjusts the temperature and humidity of each zone.

Images