An intelligent frost control method and system for an air source heat pump system

The intelligent defrosting control method for air source heat pump systems, which utilizes multi-sensor fusion and predictive models, solves the problems of delayed and misjudgment in frost layer detection under humid and cold climates, achieving efficient and accurate defrosting control and improving system stability and energy efficiency.

CN121761539BActive Publication Date: 2026-11-10BEIJING NATIONAL BUILDING GREEN & LOW CARBON TECHNOLOGY INNOVATION CENTER CO LTD +1
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
CN202511867878.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-12-11
Publication Date
2026-11-10
Estimated Expiration
2045-12-11

AI Technical Summary

Technical Problem

Existing air source heat pump systems suffer from problems such as delayed defrost detection, misjudgment, delayed response, and insufficient utilization of external information in cold and humid climates, resulting in inaccurate defrost control and affecting system stability and energy efficiency.

Method used

By employing a multi-sensor fusion and prediction model, a multi-dimensional state feature vector is constructed by collecting evaporator pressure and temperature data, compressor operating parameters, and ambient humidity. Combined with a frost layer identification model and a trend prediction model, the system can achieve real-time identification and trend warning of frost layer status and dynamically adjust the defrosting strategy.

Benefits of technology

It significantly improves the accuracy and response speed of frost layer judgment, reduces the frequency of defrosting, improves system energy efficiency, enhances user thermal comfort, and has good engineering adaptability and industrialization prospects.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121761539B_ABST
    Figure CN121761539B_ABST
Patent Text Reader

Abstract

The application belongs to the technical field of heating, ventilation and air conditioning, and relates to an intelligent frost detection control method and system for an air source heat pump system. The method comprises the following steps: S1: collecting refrigerant pressure and temperature data at the inlet and outlet of an evaporator, and collecting real-time operating condition parameters, environmental parameters and meteorological prediction parameters in a future period; S2: preprocessing the collected data and constructing a multi-dimensional state feature vector; S3: obtaining a current frost layer grade index L1; S4: obtaining a frost layer grade trend index L2 in the future period; S5: obtaining a comprehensive frost detection index S _defrost , and determining whether to perform a defrosting operation on the basis of the comprehensive frost detection index S; and S6: when the defrosting operation is needed, scheduling a defrosting control strategy from a preset defrosting strategy library on the basis of the current frost layer grade index L1, the real-time operating condition parameters and the environmental parameters, and performing the defrosting operation on the basis of the scheduled defrosting control strategy. The method can improve the annual energy efficiency and operation reliability of the air source heat pump system in a humid and cold climate.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of HVAC technology and relates to an intelligent defrosting control method and system for air source heat pump systems. Specifically, this invention relates to an intelligent defrosting control method based on multi-sensor information fusion and prediction models, applied to the automatic defrosting management of air source heat pump systems under humid and cold climate conditions. It can be widely used in heating equipment for residential, commercial, and industrial buildings and belongs to the cross-technology category of energy-saving control of heat pump systems and the application of artificial intelligence algorithms. Background Technology

[0002] Air source heat pump systems, as a highly energy-efficient and environmentally friendly combined cooling and heating system, have seen increasingly widespread application in building heating, electrification retrofitting, and renewable energy systems in recent years. Especially under the impetus of the "dual-carbon" strategy, they have become one of the important pathways to replace traditional coal-fired heating. In the hot-summer, cold-winter regions of southern China, air source heat pumps are widely used for winter heating in residential buildings, commercial buildings, and public buildings. However, the winter climate in these regions is characterized by low temperatures and high humidity, with outdoor temperatures often ranging from 0°C to 8°C and relative humidity exceeding 80%. Under these conditions, the outdoor heat exchanger (evaporator) of the heat pump is highly susceptible to frost formation. Frost buildup not only significantly reduces heat exchange efficiency but also increases compressor load, causes system instability, and can even lead to frequent unit shutdowns and decreased comfort. Therefore, accurately determining the timing of frost formation, quickly and efficiently completing defrosting, and minimizing the interference of defrosting on normal heating have become key technical issues in the engineering and promotion of this type of product.

[0003] Existing air source heat pump systems generally employ defrosting strategies based on single parameters such as temperature difference (e.g., the temperature difference between the inlet and outlet air of the evaporator) or pressure difference (changes between high and low pressure). Some products also use timed defrosting as a supplement. While these methods are simple to implement, frosting behavior is greatly influenced by multiple factors such as ambient temperature, humidity, and system load. A single parameter often cannot accurately reflect the state of the frost layer, leading to significant misjudgments or delays. For example, in scenarios where humidity rises rapidly at night in winter, the heat pump may have already frosted, but the temperature difference may not have reached the set threshold, resulting in a delayed defrosting. Conversely, if the outdoor environment is dry and cold or the system is operating at low load, falsely triggering defrosting due to temperature fluctuations not only wastes energy but also affects the continuity of indoor heating.

[0004] To improve the adaptability and accuracy of defrosting strategies, some patents have explored the introduction of multi-source sensor fusion and multi-parameter defrosting judgment strategies. For example, Guizhou University proposed a method for estimating frost amount based on data such as air humidity, heat exchanger temperature, and air volume in CN110285617A, which has on-demand defrosting capabilities. Shenzhen Paiwo New Energy Technology Co., Ltd. adopted a capacitive frost layer thickness sensor and introduced ultrasonic vibration-assisted defrosting in CN113654282A to enhance defrosting efficiency and reduce user perception interference. Chongqing University proposed a combined strategy of frost core identification and localized enhanced defrosting in CN109442824B, which adjusts the flow direction of the heat medium by controlling valves to achieve frost layer distribution-responsive defrosting and avoid localized frost recurrence. In addition, Xinyu Group Co., Ltd. integrated more than six parameters, including ambient temperature and humidity and compressor suction pressure, in CN119022403B to construct a combined defrosting criterion system, which significantly improved the robustness of the system's frost detection response.

[0005] Although existing air source heat pump defrosting control technology has made progress in terms of sensor types and logic rules, it still has significant shortcomings in actual operation, mainly including the following aspects: First, current mainstream defrosting detection methods mostly rely on a single physical parameter (such as temperature difference, pressure difference, or time) for judgment, failing to comprehensively consider the coupled effects of multiple factors involved in the defrosting process of heat exchangers, such as temperature and humidity changes, system flow fluctuations, and load changes. This leads to widespread defrosting detection delays or misjudgments. In actual operating conditions, it is easy for frost to significantly affect heat exchange efficiency without triggering defrosting, or for defrosting to be mistakenly triggered due to parameter fluctuations even before frost has formed, reducing system stability and energy efficiency.

[0006] Secondly, although some patents have introduced multi-sensor data fusion methods, they mostly remain within the framework of empirical threshold settings and expert rules, lacking modeling of the frost development process and dynamic response adjustment mechanisms, and thus failing to achieve trend-oriented control logic. Especially under wet and cold climate conditions, frost formation is characterized by suddenness and rapidity, and relying solely on the current state makes it difficult to respond in a timely manner, easily leading to a vicious cycle of frequent defrosting or severe frost adhesion.

[0007] Furthermore, currently available technologies rarely consider using weather forecast data as one of the inputs for frost detection control, which prevents the system from using information about upcoming climate changes (such as sudden drops in temperature and soaring humidity) to adjust its operating strategy in advance, thus losing the opportunity to improve energy efficiency and stability through feedforward prediction.

[0008] In addition, the existing defrosting detection control method is weakly coupled with the defrosting execution logic. Defrosting detection is only used to trigger the defrosting switch. There is no linkage mechanism to optimize the defrosting strategy according to the frost layer distribution pattern. It is difficult to support the coordinated control of multiple defrosting modes (such as hot gas bypass, directional heating, and soft switching) and the system energy-saving target.

[0009] In summary, the defrosting control of current air source heat pump systems still faces technical bottlenecks such as insufficient defrosting accuracy, slow response, lack of predictive ability, poor algorithm adaptability, and low utilization of external information. There is an urgent need to develop an intelligent defrosting control method based on multi-sensor fusion, combined with weather forecasting and time series modeling, and with trend prediction capabilities, in order to significantly improve the annual energy efficiency and operational reliability of air source heat pumps in humid and cold climates. Summary of the Invention

[0010] To address the shortcomings of existing technologies, this invention proposes an intelligent defrosting control method and system for air source heat pump systems. Based on multi-sensor fusion, combined with weather forecasting and time series modeling, and possessing trend prediction capabilities, it can significantly improve the annual energy efficiency and operational reliability of air source heat pumps in humid and cold climates.

[0011] To achieve the above objectives, the present invention provides the following technical solution: A smart defrosting control method for an air source heat pump system, characterized by comprising the following steps: S1: Collect refrigerant pressure and temperature data at the evaporator inlet and outlet of the air source heat pump system, and collect real-time operating parameters, environmental parameters and weather forecast parameters for future periods of the air source heat pump system. S2: Preprocess the collected data and construct a multidimensional state feature vector based on the preprocessed data; S3: Input the multidimensional state feature vector into the frost layer recognition model to obtain the current frost layer level index L1; S4: Input the multidimensional state feature vector and the meteorological prediction parameters for the future time period into the trend prediction model to obtain the frost layer level trend index L2 for the future time period. S5: Weight the current frost level index L1 and the frost level trend index L2 to obtain the comprehensive frost detection index S. _defrost Based on the comprehensive frost detection index S _defrost Determine whether defrosting is required; S6: When defrosting is required, a defrosting control strategy is scheduled from the preset defrosting strategy library based on the current frost level index L1, real-time operating condition parameters and environmental parameters, and the defrosting operation is performed based on the scheduled defrosting control strategy.

[0012] Preferably, the method further includes the following steps: S7: When the defrosting control strategy is completed and the preset defrosting termination condition is met, the system enters the reset state, automatically clears the collected data and re-initializes the data acquisition process, and enters a new monitoring cycle.

[0013] Preferably, in step S1, the refrigerant pressure and temperature data at the evaporator inlet and outlet of the air source heat pump system are collected by a first pressure-temperature integrated sensor and a second pressure-temperature integrated sensor respectively installed on the low-pressure side and the high-pressure side of the air source heat pump system. The collected real-time operating parameters of the air source heat pump system include the operating frequency, speed and current load of the compressor and fan. The collected environmental parameters include ambient temperature and ambient humidity. The collected meteorological forecast parameters for the future period are hourly weather forecast data for the next 1 to 6 hours, which include temperature, relative humidity, wind speed, dew point temperature and atmospheric pressure.

[0014] Preferably, in step S2, the constructed multidimensional state feature vector includes the evaporator inlet and outlet temperature difference ΔT, the refrigerant side pressure difference ΔP, the compressor operating frequency f, the compressor current change rate di / dt, the wet cooling factor WCI composed of ambient humidity and ambient temperature, and the fan speed change rate dn / dt.

[0015] Preferably, in step S3, the frost layer identification model is a multi-class identification model constructed based on random forest or gradient boosting decision tree algorithms.

[0016] Preferably, in step S4, the trend prediction model is a time series regression model based on a recurrent neural network structure.

[0017] Preferably, in step S5, S _defrost = α·L1+ β·L2, where α and β are weighting factors, satisfying α +β = 1.

[0018] Furthermore, the present invention also provides an intelligent defrosting control system for an air source heat pump system, characterized in that it includes: The data acquisition module is used to collect refrigerant pressure and temperature data at the evaporator inlet and outlet of the air source heat pump system, as well as real-time operating parameters, environmental parameters, and weather forecast parameters for future periods of the air source heat pump system. The data preprocessing and multidimensional state feature vector construction module is used to preprocess the collected data and construct multidimensional state feature vectors based on the preprocessed data. The current frost level index determination module is used to input the multidimensional state feature vector into the frost layer recognition model to obtain the current frost level index L1. The frost layer growth trend prediction module is used to input the multidimensional state feature vector and meteorological prediction parameters for future periods into the trend prediction model to obtain the frost layer level trend index L2 for future periods. The comprehensive frost detection index calculation and defrosting judgment module is used to weight and combine the current frost level index L1 and the frost level trend index L2 to obtain the comprehensive frost detection index S. _defrost Based on the comprehensive frost detection index S _defrost Determine whether defrosting is required; The defrosting control strategy scheduling and execution module is used to schedule defrosting control strategies from the preset defrosting strategy library based on the current frost level index L1, real-time operating condition parameters and environmental parameters when defrosting operation is required, and to perform defrosting operation based on the scheduled defrosting control strategies.

[0019] Furthermore, the present invention also provides an intelligent defrosting control device for an air source heat pump system, characterized in that it includes: One or more processors; Memory, used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the intelligent defrosting control method for air source heat pump systems described in this invention. Finally, the present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that, when the program is executed by a processor, it implements the steps of the intelligent defrosting control method for an air source heat pump system described in the present invention.

[0020] Compared with existing technologies, the intelligent defrosting control method for air source heat pumps based on multi-sensor fusion and prediction models of the present invention has the following significant technical advantages and application value in terms of defrosting judgment mechanism, control strategy response efficiency, and system operating energy efficiency: 1. This invention utilizes pressure and temperature data from both sides of the evaporator, compressor and fan operating parameters, and ambient temperature and humidity to construct a multi-dimensional state feature vector. Combined with a frost layer identification model and a trend prediction model, it significantly improves the accuracy and foresight of frost detection and judgment, enabling real-time identification and trend warning of frost conditions. It can proactively respond 10 to 30 minutes in advance, significantly reducing the lag in defrosting actions.

[0021] 2. This invention, by introducing a dynamic comprehensive defrosting index and a defrosting control strategy scheduling mechanism, can effectively reduce unnecessary defrosting cycles. Actual test results show that the frequency of defrosting decreases by approximately 20-30%, energy consumption per defrost cycle decreases by 10-25%, and the annual operating energy efficiency ratio (APF) can be improved by 8-12%. In terms of user experience, the more gentle defrosting mode control avoids heating interruptions and fluctuations in outlet air temperature caused by traditional defrosting, thus improving thermal comfort.

[0022] 3. This invention possesses excellent compatibility and engineering adaptability. All sensors used are common configurations found in heat pump equipment, the control algorithm can be integrated into existing motherboards or edge control modules, and the required meteorological data can be accessed through standard API interfaces. It features low deployment costs, strong adaptability, and ease of widespread application in both residential and commercial air source heat pump products.

[0023] In summary, this invention not only significantly improves the operational stability and energy efficiency of air source heat pump systems under humid and cold conditions, but also possesses good product integration and industrialization prospects, and has high technological promotion value and socio-economic benefits. Attached Figure Description

[0024] Figure 1 This is a flowchart of the intelligent defrosting control method for an air source heat pump system according to the present invention.

[0025] Figure 2 This is a schematic diagram of the structure of an air source heat pump system that employs the intelligent defrosting control method of the present invention.

[0026] Figure 3 This is a schematic diagram of the intelligent defrosting control system for the air source heat pump system of the present invention. Detailed Implementation

[0027] Before detailing any embodiment of the invention, it should be understood that the invention, in its application, is not limited to the details of the construction and arrangement of the components set forth in the following description or illustrated in the following figures. The invention can have other embodiments and can be practiced or carried out in various ways. Furthermore, it should be understood that the wording and terminology used herein are for descriptive purposes and should not be considered limiting. The use of “comprising” or “having” and variations thereof is intended to cover the items set forth below and their equivalents, as well as any additional items. Unless otherwise specified or limited, the terms “installation,” “connection,” “support,” and “linkage,” and variations thereof are used broadly and cover both direct and indirect installation, connection, support, and linking. Moreover, “connection” and “linkage” are not limited to physical or mechanical connections or links. Furthermore, firstly, in the disclosure of this invention, the terms "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the above terms should not be construed as limiting this invention. Secondly, the term "a" should be understood as "at least one" or "one or more," that is, in one embodiment, the number of an element can be one, while in another embodiment, the number of the element can be multiple. The term "a" should not be construed as a limitation on the quantity.

[0028] In the field of air source heat pump systems, current defrosting control strategies primarily focus on judging the current state, with most failing to establish future trend prediction mechanisms. They lack time-series modeling of the frost formation process and have not effectively incorporated external data sources such as weather forecasts for predictive control. While preliminary attempts at frost detection based on machine learning or AI algorithms have been made in the literature, they are still rare in published patents. A systematic approach integrating data-driven, physical modeling, and external prediction remains lacking. Therefore, it is necessary to develop an intelligent frost detection control method that integrates multi-dimensional data and possesses the ability to predict frost trends and provide control output, and to engineer its deployment in heat pump control systems to meet the practical needs of efficient and reliable defrosting control for air source heat pumps operating under complex conditions in humid and cold climates.

[0029] Figure 1 A flowchart of the intelligent defrosting control method for an air source heat pump system according to the present invention is shown. Figure 1 As shown, the intelligent defrosting control method for an air source heat pump system of the present invention includes the following steps: S1: Data Acquisition.

[0030] Data on refrigerant pressure and temperature at the evaporator inlet and outlet of the air source heat pump system are collected, along with real-time operating parameters, environmental parameters, and weather forecast parameters for future periods.

[0031] Specifically, such as Figure 2 As shown, the air source heat pump system includes components such as a compressor 1, a condenser 2 (serving as an indoor heat exchanger), an electronic expansion valve 3, an evaporator 4 (serving as an outdoor heat exchanger), and a fan 5. In this invention, to achieve the perception and trend prediction of frosting conditions, a first PT sensor (pressure-temperature integrated sensor) 7 and a second PT sensor 6 are respectively arranged at the inlet and outlet of the evaporator 4 to acquire refrigerant pressure and temperature data at the evaporator inlet and outlet, thereby facilitating the calculation of the evaporator's heat exchange performance and frosting trend.

[0032] Furthermore, this invention also includes a control motherboard 11 that establishes a communication connection with the controllers of the compressor 1 and the fan 5. The control motherboard 11 is used to acquire real-time operating parameters such as compressor operating frequency, current, voltage, and fan speed. Simultaneously, an integrated ambient temperature and humidity sensor 9 is installed on the air inlet side to monitor the current outdoor air temperature and relative humidity 8. Moreover, the control motherboard 11 is equipped with an IoT communication module, which connects to a cloud-based meteorological platform 10 via an API interface to periodically and synchronously acquire hourly meteorological forecast data for the next 1 to 6 hours, including temperature, relative humidity, wind speed, dew point temperature, and atmospheric pressure.

[0033] S2: Data preprocessing and construction of multidimensional state feature vectors.

[0034] The collected data is preprocessed, and a multidimensional state feature vector is constructed based on the preprocessed data.

[0035] In this invention, the data collected in step S1 is first preprocessed. This preprocessing includes data cleaning and standardization operations, specifically including moving average filtering to smooth short-term fluctuations, outlier identification and removal to eliminate sensor drift interference, and dimensional unification based on interval normalization or Z-score methods. These specific preprocessing methods are existing technologies and, for simplicity, will not be described in detail here.

[0036] Based on the preprocessed data, a multidimensional state feature vector F is constructed. In this invention, the constructed multidimensional state feature vector F includes, but is not limited to, parameters such as: evaporator inlet and outlet temperature difference ΔT, refrigerant side pressure difference ΔP, compressor operating frequency f, compressor current change rate di / dt, the wet-cold index (WCI) composed of ambient humidity and temperature, and fan speed change rate dn / dt. This multidimensional state feature vector F serves as input to the subsequent state recognition model and trend prediction model, used in the training and inference stages to ensure that the input data has physical meaning and discriminative power, thus helping to improve the accuracy of frost layer identification and prediction.

[0037] S3: Current frost level index determination.

[0038] The multidimensional state feature vector is input into the frost layer recognition model to obtain the current frost layer level index L1.

[0039] In this invention, the multidimensional state feature vector F constructed in step S2 is input into the frost layer recognition model, which is a multi-class recognition model trained by a supervised learning algorithm. Preferably, the frost layer recognition model is an ensemble learning structure constructed using a Random Forest or Gradient Boosting Decision Tree (GBDT) algorithm. This frost layer recognition model takes the multidimensional state feature vector F as input and learns the nonlinear mapping relationship between the system's operating state and the degree of frost through a tree structure, outputting the current frost level index L1. L1 can be a discrete level label, such as "0 - no frost", "1 - light frost", "2 - moderate frost", or "3 - severe frost", or it can be a continuous score used to characterize the confidence range of the severity of frost. This output serves as the first channel basis for the intelligent frost judgment control logic, directly reflecting the current frost state of the heat exchanger surface and providing a basis for subsequent comprehensive decision-making.

[0040] S4: Forecast of frost layer growth trend.

[0041] The multidimensional state feature vector and the meteorological forecast parameters for future periods are input into the trend prediction model to obtain the frost layer level trend index L2 for future periods.

[0042] In this invention, the multidimensional state feature vector F constructed in step S2 and the meteorological prediction parameters for future time periods are used as joint inputs to the trend prediction model. The trend prediction model is used to predictively model the frost evolution trend of the air source heat pump system within a time window of Δt. Preferably, the trend prediction model is a time series regression model based on a recurrent neural network structure, including but not limited to Long Short-Term Memory (LSTM) networks or Gated Recurrent Unit (GRU) networks, which have the ability to handle nonlinear dynamic sequences.

[0043] The input to the trend prediction model includes the operating characteristic sequence of the air source heat pump system over a past period (from time tn to the current time t), that is, the constructed multidimensional state characteristic vector sequence F. t-n ~ F t and the weather parameter series for future periods. _t+1 ~ Weather _t+Δt The output is the frost level trend index L2 within the predicted time window. The frost level trend index L2 is a confidence assessment indicator reflecting the short-term development trend of frost, and can be represented as discrete levels such as "tending towards no frost," "tending towards light frost," or "tending towards heavy frost," or a continuous probability prediction score, used to provide early warning before frost forms. This output serves as the second channel for the intelligent frost judgment control logic, providing predictive support for subsequent comprehensive determination of whether to perform defrosting actions.

[0044] S5: Comprehensive frost detection index calculation and defrosting judgment.

[0045] The current frost level preparation index L1 and the frost level trend index L2 are weighted and combined to obtain the comprehensive frost detection index S. _defrost Based on the comprehensive frost detection index S _defrost Determine whether defrosting is necessary.

[0046] In this invention, a dual-channel fusion-type frost detection decision mechanism is constructed, which weights and combines the current frost level index L1 output in step S3 with the frost level trend index L2 output in step S4 to construct a comprehensive frost detection index S. _defrost This is used to uniformly determine whether to perform a defrosting operation. The comprehensive defrosting index S is used to uniformly determine whether to perform a defrosting operation. _defrost Calculate using the following formula: S _defrost = α·L1+ β·L2.

[0047] Here, α and β are weighting factors set or dynamically adjusted by the system, satisfying α + β = 1, which respectively reflect the relative importance that the air source heat pump system attaches to the current state identification and future trend prediction results.

[0048] Furthermore, the present invention further sets a dynamic frost detection threshold S. _th The dynamic frost threshold S _th The comprehensive defrost index S can be dynamically adjusted based on factors such as the operating environment of the air source heat pump system, historical defrost records, and user-defined priorities (e.g., energy saving or comfort). _defrost ≥ S _th When a risk is identified that requires defrosting, the system will automatically enter the defrosting control process.

[0049] This invention effectively balances current operating conditions with short-term development trends, reduces misjudgment rate and lag risk, and significantly improves the accuracy and economy of defrosting.

[0050] S6: Defrosting control strategy scheduling and execution.

[0051] When defrosting is required, a defrosting control strategy is scheduled from the preset defrosting strategy library based on the current frost level index L1, real-time operating condition parameters, and environmental parameters, and the defrosting operation is performed based on the scheduled defrosting control strategy.

[0052] In this invention, the comprehensive frost detection index S _defrost ≥ S _thWhen the defrost trigger condition is met, a defrost control strategy will be scheduled, and the optimal defrost method will be selected and matched from the preset defrost strategy library based on the current frost level index L1, the temperature and humidity parameters of the external environment, and the real-time operating conditions of the air source heat pump system (including compressor frequency, evaporation temperature, wind speed, etc.). The preset defrost strategy library includes, but is not limited to, the following three typical defrost methods: (1) Reverse Cycle Defrost (RCD): By controlling the reversing valve to change the direction of refrigerant flow, the condenser is temporarily switched to the evaporator to release heat for defrosting; (2) Gas Bypass Defrost (GBD): The compressor exhaust is introduced into the evaporator for directional heating through the gas bypass valve; (3) Shutdown Defrost (SD): The machine is shut down under specific environmental or fault conditions to wait for natural defrosting or forced heating.

[0053] Then, commands can be executed according to the matched optimal defrosting method, and the action status of each actuator can be controlled in a coordinated manner, including but not limited to switching of reversing valves, opening and closing of solenoid valves, setting of compressor operating frequency, fan start and stop logic, defrosting duration management, etc., forming a defrosting control sequence.

[0054] S7: Reset and control closed loop establishment.

[0055] Once the defrosting control strategy is completed and the preset defrosting termination condition is met, the system enters a reset state, automatically clears the collected data, reinitializes the data acquisition process, and enters a new control cycle.

[0056] In this invention, after the defrosting control strategy is completed and the preset defrosting termination condition is met, the system enters a reset state, automatically clearing temporary variables used for state identification and trend prediction in the previous cycle, including but not limited to: multidimensional state feature vector cache, model prediction state, defrosting flag, and strategy execution flag. Simultaneously, the data acquisition process is reinitialized, and a new monitoring cycle begins.

[0057] This invention constructs a complete intelligent defrosting control closed loop by sequentially executing data acquisition, status identification, trend prediction, defrosting decision-making, strategy execution, and system reset, ensuring that the air source heat pump system has continuous, dynamically adaptive, and highly efficient defrosting capabilities under humid and cold conditions.

[0058] Meanwhile, in this invention, the hardware and software architecture supports edge computing. The frost layer identification model and trend prediction model can run on the MCU controller or embedded edge node, or they can be trained by a remote server and then deployed to the device side. Weather data is automatically synchronized daily through a standard REST API to ensure the accuracy of the input boundary conditions of the trend prediction model.

[0059] The air source heat pump system of the present invention is applicable to vapor compression cycle heat pump devices using R410A, R32, R290, CO2 or other common or alternative refrigerants. The intelligent defrosting control method and system of the present invention do not depend on a specific working fluid and have good versatility and scalability.

[0060] In summary, this invention improves the accuracy and response speed of defrosting judgment by constructing an intelligent defrosting control method that integrates real-time operating data and future meteorological information. It has good engineering adaptability and energy-saving effect, and is suitable for the operation control of various air source heat pump equipment in complex and cold climates.

[0061] Figure 3 A schematic diagram of the intelligent defrosting control system for the air source heat pump system of the present invention is shown. Figure 3 As shown, the intelligent defrosting control system for the air source heat pump system of the present invention includes: 1. Data acquisition module.

[0062] The data acquisition module is used to collect refrigerant pressure and temperature data at the evaporator inlet and outlet of the air source heat pump system, as well as real-time operating parameters, environmental parameters, and meteorological forecast parameters for future periods of the air source heat pump system.

[0063] 2. Data preprocessing and multidimensional state feature vector construction module.

[0064] The data preprocessing and multidimensional state feature vector construction module is used to preprocess the collected data and construct multidimensional state feature vectors based on the preprocessed data.

[0065] 3. Current Frost Level Index Determination Module The current frost level index determination module is used to input the multidimensional state feature vector into the frost layer recognition model to obtain the current frost level index L1.

[0066] 4. Frost layer growth trend prediction module.

[0067] The frost growth trend prediction module is used to input the multidimensional state feature vector and meteorological prediction parameters for future periods into the trend prediction model to obtain the frost level trend index L2 for future periods.

[0068] 5. Comprehensive frost detection index calculation and defrosting judgment module.

[0069] The comprehensive frost detection index calculation and defrosting judgment module is used to weight and combine the current frost level index L1 and the frost level trend index L2 to obtain the comprehensive frost detection index S. _defrost Based on the comprehensive frost detection index S _defrost Determine whether defrosting is necessary.

[0070] 6. Defrosting control strategy scheduling and execution module.

[0071] The defrosting control strategy scheduling and execution module is used to schedule defrosting control strategies from the preset defrosting strategy library based on the current frost level index L1, real-time operating condition parameters and environmental parameters when defrosting operation is required, and to perform defrosting operation based on the scheduled defrosting control strategies.

[0072] Furthermore, the intelligent defrosting control system for the air source heat pump system of the present invention may further include: 7. Reset and control closed-loop establishment module.

[0073] The reset and control closed-loop establishment module is used to enter the reset state after the defrosting control strategy has been executed and the preset defrosting termination condition has been reached. It automatically clears the collected data and re-initializes the data acquisition process to enter a new monitoring cycle.

[0074] Furthermore, the present invention also provides an intelligent defrosting control device for an air source heat pump system, comprising: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the intelligent defrosting control method for an air source heat pump system of the present invention. Finally, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the intelligent defrosting control method for an air-source heat pump system of the present invention.

[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention. Those skilled in the art can modify or make equivalent substitutions to the technical solutions of the present invention based on the concept of the present invention, without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. A smart defrosting control method for an air source heat pump system, characterized in that, Includes the following steps: S1: Collect refrigerant pressure and temperature data at the evaporator inlet and outlet of the air source heat pump system, and collect real-time operating parameters, environmental parameters and weather forecast parameters for future periods of the air source heat pump system. S2: The collected data is preprocessed, and a multidimensional state feature vector is constructed based on the preprocessed data. The preprocessing includes data cleaning and standardization operations, specifically including moving average filtering to smooth short-term fluctuations, outlier identification and removal to eliminate sensor drift interference, and dimension unification processing based on interval normalization or Z-score methods. The constructed multidimensional state feature vector includes the evaporator inlet and outlet temperature difference ΔT, refrigerant side pressure difference ΔP, compressor operating frequency f, compressor current change rate di / dt, the wet cooling factor WCI composed of ambient humidity and ambient temperature, and the fan speed change rate dn / dt. S3: Input the multidimensional state feature vector into the frost layer recognition model to obtain the current frost layer level index L1; S4: Input the multidimensional state feature vector and the meteorological prediction parameters for the future time period into the trend prediction model to obtain the frost layer level trend index L2 for the future time period. S5: Weight the current frost level index L1 and the frost level trend index L2 to obtain the comprehensive frost detection index S. _defrost Based on the comprehensive frost detection index S _defrost Determine whether defrosting is required; S6: When defrosting is required, a defrosting control strategy is scheduled from the preset defrosting strategy library based on the current frost level index L1, real-time operating condition parameters and environmental parameters, and the defrosting operation is performed based on the scheduled defrosting control strategy.

2. The intelligent defrosting control method for an air source heat pump system according to claim 1, characterized in that, Further steps include: S7: When the defrosting control strategy is completed and the preset defrosting termination condition is met, the system enters the reset state, automatically clears the collected data and re-initializes the data acquisition process, and enters a new control cycle.

3. The intelligent defrosting control method for an air source heat pump system according to claim 1 or 2, characterized in that, In step S1, the refrigerant pressure and temperature data at the evaporator inlet and outlet of the air source heat pump system are collected by the first pressure-temperature integrated sensor and the second pressure-temperature integrated sensor respectively installed on the low-pressure side and the high-pressure side of the air source heat pump system. In addition, the real-time operating condition parameters of the air source heat pump system collected include the operating frequency, speed and current load of the compressor and fan. The environmental parameters collected include the ambient temperature and ambient humidity. The meteorological forecast parameters collected for the future period are hourly weather forecast data for the next 1 to 6 hours, which include temperature, relative humidity, wind speed, dew point temperature and atmospheric pressure.

4. The intelligent defrosting control method for an air source heat pump system according to claim 3, characterized in that, In step S3, the frost layer identification model is a multi-class identification model constructed based on random forest or gradient boosting decision tree algorithms.

5. The intelligent defrosting control method for an air source heat pump system according to claim 4, characterized in that, In step S4, the trend prediction model is a time series regression model based on a recurrent neural network structure.

6. The intelligent defrosting control method for an air source heat pump system according to claim 5, characterized in that, In step S5, S _defrost = α·L1+ β·L2, where α and β are weighting factors, satisfying α + β = 1.

7. An intelligent defrosting control system for an air source heat pump system, characterized in that, include: The data acquisition module is used to collect refrigerant pressure and temperature data at the evaporator inlet and outlet of the air source heat pump system, as well as real-time operating parameters, environmental parameters, and weather forecast parameters for future periods of the air source heat pump system. The data preprocessing and multidimensional state feature vector construction module is used to preprocess the collected data and construct a multidimensional state feature vector based on the preprocessed data. The preprocessing includes data cleaning and standardization operations, specifically including moving average filtering to smooth short-term fluctuations, outlier identification and removal to eliminate sensor drift interference, and dimensional unification processing based on interval normalization or Z-score methods. The constructed multidimensional state feature vector includes the evaporator inlet and outlet temperature difference ΔT, the refrigerant side pressure difference ΔP, the compressor operating frequency f, the compressor current change rate di / dt, the wet cooling factor WCI composed of ambient humidity and ambient temperature, and the fan speed change rate dn / dt. The current frost level index determination module is used to input the multidimensional state feature vector into the frost layer recognition model to obtain the current frost level index L1. The frost layer growth trend prediction module is used to input the multidimensional state feature vector and meteorological prediction parameters for future periods into the trend prediction model to obtain the frost layer level trend index L2 for future periods. The comprehensive frost detection index calculation and defrosting judgment module is used to weight and combine the current frost level index L1 and the frost level trend index L2 to obtain the comprehensive frost detection index S. _defrost Based on the comprehensive frost detection index S _defrost Determine whether defrosting is required; The defrosting control strategy scheduling and execution module is used to schedule defrosting control strategies from the preset defrosting strategy library based on the current frost level index L1, real-time operating condition parameters and environmental parameters when defrosting operation is required, and to perform defrosting operation based on the scheduled defrosting control strategies.

8. An intelligent defrosting control device for an air source heat pump system, characterized in that, include: One or more processors; Memory, used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the intelligent defrosting control method for an air source heat pump system as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the intelligent defrosting control method for an air source heat pump system as described in any one of claims 1-6.

Citation Information

Patent Citations

  • A method and system for defrosting an air source heat pump

    CN109442824B

  • Air source heat pump defrosting control method

    CN110285617A

  • Air source heat pump defrosting system and defrosting judgment method

    CN113654282A

  • A heat pump air conditioner defrosting control method, device and equipment

    CN119022403B

  • Vehicle air outlet intelligent adjusting method and system based on machine learning

    CN120096286A