Air source heat pump heating method and system

Abstract

The invention provides an air source heat pump heating method and system, and relates to the technical field of heat pump system thermal control, and the method comprises the steps: collecting evaporator surface frost layer thickness, environment temperature and humidity parameters in real time, inputting the parameters to a pre-trained neural network algorithm model, and identifying a frosting rate by analyzing a multi-factor nonlinear relation; predicting a future frosting development trend, and outputting frosting thickness increase prediction and severity degree results; and according to the frosting thickness increase prediction and the severity degree result, reverse circulation defrosting is started in advance when the frost layer does not reach the critical thickness, and optimal control over the defrosting process is achieved by adjusting the opening degree of the four-way valve and the rotating speed of the draught fan, so that the system state after defrosting is obtained. The energy efficiency of the system is improved, and the heating comfort is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of thermal control of heat pump systems, and in particular to an air source heat pump heating method and system. Background Art

[0002] Air source heat pump heating systems have become the core technical solution for winter heating of self-built houses in rural areas of my country's extremely cold regions due to their advantages of not requiring fossil fuel combustion and high energy efficiency. In such scenarios, rural self-built houses generally lack municipal centralized heating pipeline coverage, and users are highly sensitive to heating costs and need to take into account unattended operation needs. The decentralized heating characteristics and energy-saving advantages of air source heat pumps can effectively adapt to the needs.

[0003] However, traditional control technology has some limitations. Some defrost controls based on fixed thresholds cannot adapt to dynamically changing temperature and humidity conditions, and are prone to defrost timing deviations. Parameter adjustment methods that rely on mechanism models are difficult to describe the complex nonlinear coupling relationship between ambient temperature and humidity, frosting rate, heating capacity, and grid voltage in the heat pump system, and cannot achieve precise control. Single parameter feedback control cannot take into account the interaction of multiple variables, and the control accuracy is insufficient. However, when applying neural network algorithms to rural air source heat pump heating scenarios in cold areas, there are still some challenges. For example, a single neural network model often focuses on solving local problems and fails to form a coordinated optimization overall control framework with other key links such as voltage compensation, load matching, and room control. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide an air source heat pump heating method and system to achieve system energy efficiency improvement and heating comfort optimization.

[0005] In order to solve the above technical problems, the technical solutions of the present invention are as follows: In a first aspect, an air source heat pump heating method is provided, the method comprising: Step 1: Real-time data collection of frost thickness, ambient temperature, and humidity on the evaporator surface is fed into a pre-trained neural network algorithm model. The model then analyzes the nonlinear relationship between multiple factors to identify the frost formation rate, predict future frost development trends, and output frost thickness growth predictions and severity results. Step 2: Based on the frost thickness growth prediction and severity results, reverse cycle defrosting is started in advance before the frost layer reaches a critical thickness. The defrost process is optimized by adjusting the four-way valve opening and the fan speed to obtain the system state after defrosting. Step 3: Based on the system status after defrosting, the grid voltage fluctuation is monitored in real time. Before the voltage drops below the normal operating range of the compressor, voltage compensation is initiated and the compressor frequency is dynamically adjusted to maintain stable system operation. Step 4: Based on the stable operation of the system, collect the supply and return water temperature data of the floor heating system. Combine the historical data with the real-time heat load demand to establish a thermal dynamic response relationship. By adjusting the speed of the buffer water tank circulation pump and the opening of the mixing valve, the matching result between the heat pump output and the heating demand is obtained. Step 5: Based on the matching results between the heat pump output and the heating demand, obtain temperature data at multiple locations through distributed temperature monitoring devices to establish the thermal field characteristic distribution of the heating area; generate system control parameters according to the thermodynamic characteristic parameters of each area, formulate a regional timed temperature control strategy based on the system control parameters, and control each heating circuit through Internet of Things communication to achieve energy efficiency optimization and comfortable heating.

[0006] Furthermore, the frost thickness, ambient temperature, and humidity parameters of the evaporator surface are collected in real time and input into a pre-trained neural network algorithm model. By analyzing the nonlinear relationship between multiple factors, the frost rate is identified, the future frost development trend is predicted, and the frost thickness growth prediction and severity results are output, including: Step 1.1, real-time collection of evaporator surface frost thickness, ambient temperature and humidity parameters, pre-processing the collected raw data parameters to obtain pre-processed data; Step 1.2: Based on the preprocessed data, the data is organized into a standardized input vector containing ambient temperature, humidity, and frost thickness in chronological order. This standardized input vector is then fed into a pretrained neural network model. The internal hidden layer nodes perform high-dimensional nonlinear mapping and feature extraction on the complex nonlinear interactions between the parameters in the input vector. This calculation results in implicit state features that reflect the current instantaneous frost formation condition. Based on these implicit state features, the output layer of the neural network model uses linear transformation and activation function processing to determine the precise frost formation rate under the current operating conditions. In step 1.3, based on the real-time frost formation rate and the time series data of the current ambient temperature and humidity, the neural network model performs a multi-step forward deduction calculation in the internal state space. It iteratively predicts the frost thickness increment at each moment in the future preset time window, accumulates it to the current thickness, and generates a frost thickness growth prediction curve with time as the independent variable, thereby obtaining the grade assessment result of the frost severity in the future period.

[0007] Furthermore, based on the frost thickness growth prediction and severity results, reverse cycle defrosting is started in advance before the frost layer reaches the critical thickness. The defrost process is optimized by adjusting the four-way valve opening and fan speed to obtain the system status after defrosting, including: Step 2.1: Based on the prediction and severity results, determine the frost growth trend and calculate the expected time to reach the critical thickness. Before the actual frost thickness reaches the critical thickness, generate a reverse cycle defrost start instruction in advance based on the prediction results; Step 2.2: Send the defrost start command to the air source heat pump unit control system to control the four-way valve to switch the refrigerant flow direction and start the reverse cycle defrost process; Step 2.3: During the defrost process, the opening rate of the four-way valve and the speed of the outdoor fan are dynamically adjusted according to the frost severity results to match the defrost heat requirements under different frost severity levels, thereby achieving optimized control of the defrost process. Step 2.4: monitor the evaporator surface temperature and system pressure parameters in real time to determine the degree of defrosting completion. When it is confirmed that the frost layer on the evaporator surface is completely cleared and the system parameters have returned to the normal operating range, generate a defrost end instruction, control the four-way valve and fan to return to the heating operation state, and obtain the system state after defrosting.

[0008] Furthermore, based on the system status after defrosting, the grid voltage fluctuation is monitored in real time, voltage compensation is initiated before the voltage falls below the normal operating range of the compressor, and the compressor frequency is dynamically adjusted to maintain stable system operation, including: Step 3.1: Based on the post-defrost system status, the grid voltage monitoring function is activated accordingly. The grid voltage data is collected in real time through the voltage sensor, and the voltage data is filtered and trend analyzed to predict voltage change trends. Step 3.2: before the grid voltage is predicted to drop to the minimum voltage threshold required for the normal operation of the compressor through analysis, a pre-compensation instruction is generated; Step 3.3: Send a pre-compensation instruction to the voltage compensation device to control it to start operation in advance to increase the power supply voltage and ensure that the compressor terminal voltage remains within the normal operating range; In step 3.4, while performing voltage compensation, the compressor frequency adjustment instruction is dynamically calculated and generated based on the amplitude and trend of the voltage fluctuation. The system load and power supply capacity are balanced by fine-tuning the compressor operating frequency to maintain the overall stable operation of the system.

[0009] Furthermore, based on the stable operation of the system, the supply and return water temperature data of the floor heating system is collected. By combining historical data with the real-time heat load demand, a thermal dynamic response relationship is established. By adjusting the speed of the buffer water tank circulation pump and the opening of the mixing valve, the matching result between the heat pump output and the heating demand is obtained, including: Step 4.1: Based on the maintained stable system operation state, start the floor heating system monitoring sequence and collect the supply water temperature and return water temperature data of the heating system in real time through the temperature sensor; Step 4.2: Call the stored historical heating data, which includes the supply and return water temperature difference, heat load, and system adjustment parameters under different outdoor working conditions, and integrate the historical data with the real-time supply and return water temperature data to obtain the integrated data; Step 4.3, based on the integrated data, calculate the corresponding relationship between the water supply and return temperature difference and the flow rate change per unit time, establish the real-time thermal dynamic response relationship of the system, and obtain the analysis results of the thermal dynamic response relationship; Step 4.4, compare the real-time heat load demand with the analysis results of the thermal dynamic response relationship, calculate the target rotating speed of the buffer tank circulating pump and the target opening degree of the mixing valve required to achieve supply-demand balance, and obtain the calculation results; Step 4.5, generate control instructions according to the calculation results, dynamically adjust the rotating speed of the buffer tank circulating pump to change the system circulating flow rate, and adjust the opening degree of the mixing valve to control the mixed water temperature, so that the output heat of the heat pump is coordinated with the real-time heating demand of the building, and the final matching result is obtained.

[0010] Further, based on the matching results of the heat pump output and the heating demand, the multi-position temperature data is obtained through the distributed temperature monitoring device, and the thermal field characteristic distribution of the heating area is established, including: Step 5.1, based on the matching results of the heat pump output and the heating demand, start the operation of the distributed temperature monitoring network, and collect real-time temperature data through temperature sensors arranged at multiple positions in the heating area; Step 5.2, clean and format the collected multi-position temperature data, eliminate abnormal data points, and associate the temperature data of each monitoring point with its corresponding spatial position information to obtain a standardized temperature distribution data set; Step 5.3, based on the obtained standardized temperature distribution data set, divide the heating area into several independent sub-areas according to the spatial structure, and calculate the representative temperature value of each sub-area using the weighted average algorithm to generate sub-area temperature distribution data; Step 5.4, based on the generated sub-area temperature distribution data, combined with the spatial structure characteristics and thermal characteristic parameters of the building, analyze the spatial correlation and variation law of the temperature data of each sub-area, and establish the thermal field characteristic distribution of the heating area.

[0011] Further, generate system control parameters according to the thermodynamic characteristic parameters of each area, develop a regional and timed temperature control strategy based on the system control parameters, and control each heating circuit through Internet of Things communication to realize energy efficiency optimization and comfortable heating, including: Step 5.5, based on the thermal field characteristic distribution, extract the thermodynamic characteristic parameters of each sub-area, including temperature stability, thermal inertia coefficient and thermal response time constant; Step 5.6, according to the extracted thermodynamic characteristic parameters, combined with the use function characteristics and comfort requirements of each sub-area, generate system control parameters for each sub-area; Step 5.7: Based on the generated system control parameters, combined with historical heating data and usage habits, formulate a regional timed temperature control strategy; In step 5.8, the developed regional timed temperature control strategy is distributed to the intelligent control devices of each heating circuit via IoT communication, achieving independent and precise control of each heating circuit; Step 5.9: Monitor the temperature changes and energy consumption data of each sub-area in real time, and dynamically optimize the regulation parameters and control strategies based on the actual operating results to achieve a balance between energy efficiency optimization and comfortable heating.

[0012] In a second aspect, an air source heat pump heating system comprises: The acquisition module is used to collect the frost thickness, ambient temperature, and humidity parameters of the evaporator surface in real time, and input them into a pre-trained neural network algorithm model. By analyzing the nonlinear relationship between multiple factors, the frost formation rate is identified, the future frost development trend is predicted, and the frost thickness growth prediction and severity results are output. Based on the frost thickness growth prediction and severity results, the reverse cycle defrost is started in advance before the frost layer reaches the critical thickness. The defrost process is optimized by adjusting the four-way valve opening and the fan speed to obtain the system state after defrosting. The calculation module is used to monitor grid voltage fluctuations in real time based on the system status after defrosting, initiate voltage compensation before the voltage falls below the normal operating range of the compressor, and dynamically adjust the compressor frequency to maintain stable system operation. Based on the stable operation of the system, the module collects the supply and return water temperature data of the floor heating system, combines historical data with real-time heat load demand, establishes a thermal dynamic response relationship, and obtains a matching result between the heat pump output and the heating demand by adjusting the buffer water tank circulation pump speed and the mixing valve opening. The processing module is used to obtain multi-location temperature data through distributed temperature monitoring devices based on the matching results of heat pump output and heating demand, and establish the thermal field characteristic distribution of the heating area; generate system control parameters according to the thermodynamic characteristic parameters of each area, formulate regional timing temperature control strategies based on the system control parameters, and control each heating circuit through Internet of Things communication to achieve energy efficiency optimization and comfortable heating.

[0013] According to a third aspect, a computing device includes: one or more processors; A storage device is 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 air source heat pump heating method.

[0014] In a fourth aspect, a computer-readable storage medium stores a program, which implements the air source heat pump heating method when executed by a processor.

[0015] The above solution of the present invention includes at least the following beneficial effects: By innovatively integrating neural network algorithms with full-process collaborative control logic, this system breaks through the technical bottlenecks of traditional systems. On the one hand, leveraging the neural network's ability to deeply explore multi-parameter nonlinear relationships, it accurately predicts evaporator frost trends and implements early defrost control, avoiding efficiency drops caused by late defrosting or energy waste caused by premature defrosting. Furthermore, through pre-compensation for voltage fluctuations and dynamic adjustment of compressor frequency, it addresses grid instability, reducing equipment downtime losses and the risk of heating interruptions. Furthermore, by establishing a thermal dynamic response relationship, it optimizes the matching of heat pump output with floor heating demand. Combined with distributed temperature monitoring, it constructs a thermal field distribution and develops a regional timed temperature control strategy. This reduces indoor temperature fluctuations, improves heating comfort, and implements differentiated temperature control tailored to the needs of different regions, reducing energy waste. Overall, this system achieves intelligent collaboration across all aspects, from frost prediction and voltage adaptation to heat load matching and regional temperature control, enhancing system operational stability in complex operating conditions such as those in rural areas with severe cold. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flow chart of an air source heat pump heating method provided by an embodiment of the present invention.

[0017] Figure 2 This is a schematic diagram of an air source heat pump heating system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0018] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0019] like Figure 1 As shown, an embodiment of the present invention provides an air source heat pump heating method, the method comprising the following steps: Step 1: Real-time data collection of frost thickness, ambient temperature, and humidity on the evaporator surface is fed into a pre-trained neural network algorithm model. The model then analyzes the nonlinear relationship between multiple factors to identify the frost formation rate, predict future frost development trends, and output frost thickness growth predictions and severity results. Step 2: Based on the frost thickness growth prediction and severity results, reverse cycle defrosting is started in advance before the frost layer reaches a critical thickness. The defrost process is optimized by adjusting the four-way valve opening and the fan speed to obtain the system state after defrosting. Step 3: Based on the system status after defrosting, the grid voltage fluctuation is monitored in real time. Before the voltage drops below the normal operating range of the compressor, voltage compensation is initiated and the compressor frequency is dynamically adjusted to maintain stable system operation. Step 4: Based on the stable operation of the system, collect the supply and return water temperature data of the floor heating system. Combine the historical data with the real-time heat load demand to establish a thermal dynamic response relationship. By adjusting the speed of the buffer water tank circulation pump and the opening of the mixing valve, the matching result between the heat pump output and the heating demand is obtained. Step 5: Based on the matching results between the heat pump output and the heating demand, obtain temperature data at multiple locations through distributed temperature monitoring devices to establish the thermal field characteristic distribution of the heating area; generate system control parameters according to the thermodynamic characteristic parameters of each area, formulate a regional timed temperature control strategy based on the system control parameters, and control each heating circuit through Internet of Things communication to achieve energy efficiency optimization and comfortable heating.

[0020] In the embodiment of the present invention, through the full-process intelligent design of prediction, regulation, matching and optimization, the core pain points of the traditional heating system are solved, and it has outstanding advantages in multiple dimensions. With the neural network algorithm as the core, the frost rate is accurately identified and the development trend is predicted by integrating multiple parameters such as the frost layer thickness of the evaporator, the ambient temperature and humidity, etc., so as to realize the early reverse cycle defrosting when the frost layer has not reached the critical thickness, thereby avoiding the problem of the traditional defrosting method causing a sudden drop in heat exchange efficiency due to being too late or causing energy waste due to being too early. The defrosting process is optimized by dynamically adjusting the opening of the four-way valve and the fan speed; after defrosting, the real-time monitoring and pre-compensation mechanism of the grid voltage fluctuation is relied on, combined with the dynamic adjustment of the compressor frequency, Avoid system shutdown or damage caused by voltage anomalies, and ensure operational continuity in unstable power grid scenarios such as rural areas in extremely cold regions; at the same time, by collecting the supply and return water temperatures of floor heating and combining historical data to establish a thermal dynamic response relationship, adjust the buffer water tank circulation pump and the mixing valve to achieve precise matching of heat pump output and heating demand, and then build the thermal field characteristic distribution of the heating area based on distributed temperature monitoring, formulate regional timed temperature control strategies and regulate each circuit through the Internet of Things to reduce indoor temperature fluctuations, improve heating comfort, and implement differentiated temperature control according to the needs of different regions to reduce energy consumption, ultimately achieving coordinated optimization of system energy efficiency, operational stability and user comfort.

[0021] In a preferred embodiment of the present invention, the above step 1 may include: Step 1.1 collects the frost thickness, ambient temperature, and humidity parameters of the evaporator surface in real time, and preprocesses the collected raw data parameters to obtain preprocessed data. Specifically, the following steps are performed: parameter collection is completed based on a preset sensor monitoring component. For the frost thickness of the evaporator surface, an infrared ranging sensor or a capacitive thickness detection device can be used to obtain real-time thickness data at a collection frequency of not less than 1 time per minute. For the ambient temperature and humidity parameters, synchronous collection is performed through an integrated temperature and humidity sensor arranged within 1 meter of the evaporator. The collection frequency is consistent with the frost thickness collection frequency to ensure the correspondence of the parameter time dimension.

[0022] Furthermore, considering that the original collected data may contain noise data or outliers caused by sensor fluctuations and external electromagnetic interference, the above-mentioned original data parameters need to be preprocessed: first, the continuously collected temperature, humidity and frost thickness data are smoothed by using the sliding average filter method to filter out high-frequency noise; secondly, the 3σ criterion is used to identify and eliminate abnormal data points that exceed the normal fluctuation range to avoid the interference of outliers on subsequent analysis; finally, the data after filtering and outlier elimination are normalized, and the values ​​of each parameter are mapped to the interval [0, 1] to obtain standardized preprocessed data.

[0023] Step 1.2: Based on the preprocessed data, a standardized input vector containing ambient temperature, humidity, and frost thickness is organized in chronological order. This standardized input vector is then input into a pretrained neural network model. The internal hidden layer nodes perform high-dimensional nonlinear mapping and feature extraction on the complex nonlinear interactions between the parameters in the input vector, calculating implicit state features that reflect the current instantaneous frost formation condition. Based on these implicit state features, the output layer of the neural network model uses linear transformation and activation function processing to obtain the precise frost formation rate under the current operating conditions. Specifically, the following steps are performed: Based on the obtained preprocessed data, the ambient temperature, humidity, and frost thickness data at the same collection moment are correlated and integrated in chronological order to construct a standardized input vector. Five sets of preprocessed data collected continuously within a fixed time interval (e.g., 5 minutes per data unit) are arranged in the order of temperature, humidity, and frost thickness at the five collection moments to form a 3×5 matrix input vector. This ensures that the vector contains multi-parameter information at a single moment and covers parameter variation trends within a short time series.

[0024] Subsequently, the standardized input vector is input into the pre-trained neural network model. The construction process of the neural network model is as follows: first, the multi-layer perceptron MLP is selected as the basic network structure, which includes an input layer, 2-3 hidden layers and an output layer; the number of neurons in the input layer matches the dimension of the standardized input vector, that is, 15 neurons, corresponding to a 3×5 matrix input, the number of neurons in each hidden layer is set to 32-64 according to the parameter coupling complexity, and 1 neuron is set in the output layer to output the frost rate. Secondly, historical frosting condition data are collected as training samples. The sample data covers the measured frost rates under different ambient temperature and humidity ranges and different frost layer thicknesses, and the sample data are preprocessed in the same way as step 1.1, and then the training is performed according to 7. :3 ratio into training set and validation set, then use mean square error as loss function, train the model through Adam optimization algorithm, calculate the second moment estimation of parameter gradient, adaptive learning rate term, dynamically adjust the update step size of each parameter, for parameters with gentle gradient change, appropriately increase the update step size to speed up convergence; for parameters with large gradient fluctuation, reduce the update step size to ensure stability, so as to achieve efficient optimization of model parameters, for example, the initial learning rate is set to 0.001, and the learning rate is decayed every 100 iterations, and the early stopping method is used. When the validation set loss does not decrease for 20 consecutive rounds, the training is terminated to avoid overfitting, until the prediction error of the model on the validation set is lower than the preset threshold, such as less than 5%, the training and construction of the model are completed.

[0025] In the process of model operation, the complex nonlinear interaction between temperature, humidity and frost thickness in the input vector is first processed by the internal hidden layer nodes. The hidden layer nodes perform weighted summation of the input parameters through the weight coefficients obtained by training, and combine them with the activation function Perform a nonlinear transformation, where is the input value of the hidden state feature received by the output layer node of the neural network after linear transformation, is the lower limit of the frosting rate, is the upper limit of the frosting rate, In order to adjust the parameters of the function slope, high-dimensional nonlinear mapping and extraction of multi-parameter coupling characteristics are realized, and finally implicit state characteristics that can reflect the current instantaneous frosting conditions are obtained, such as the slow growth characteristics in the early stage of frosting and the accelerated growth characteristics in the middle stage of frosting. Based on this implicit state characteristic, it is further transmitted to the output layer of the neural network model. The output layer compresses the dimension of the implicit state characteristic through linear transformation, and uses the activation function to map the output result to a reasonable frosting rate numerical range, thereby obtaining the accurate frosting rate under the current working conditions.

[0026] Step 1.3: Based on the real-time frost rate and the current ambient temperature and humidity time series data, the neural network model performs a multi-step forward deduction calculation in the internal state space. It iteratively predicts the frost thickness increment at each moment in the future preset time window, accumulates it to the current thickness, and generates a frost thickness growth prediction curve with time as the independent variable to obtain the grade assessment result of the frost severity in the future period. Specifically, it includes: based on the obtained real-time frost rate and combined with the current ambient temperature and humidity time series data, first, the real-time frost rate and the temperature and humidity time series data are synchronously input into the aforementioned neural network model, and the model starts a multi-step forward deduction calculation in its internal state space. Specifically, a future preset time window is set, such as the next 1 hour. The time window can be adjusted according to the actual frost sensitivity of the heating system, and the time window is divided into several sub-periods with equal time intervals, such as each sub-period is 10 minutes; then, the model uses the current frost rate as the initial value and combines the changing trend of the temperature and humidity time series data, such as temperature. The frost thickness increment within each sub-period is predicted through iterative calculation to determine whether the humidity continues to decrease or increase. In the first sub-period, the thickness increment within that period is calculated based on the current temperature and humidity conditions and the real-time frost formation rate. When entering the next sub-period, the predicted thickness of the previous sub-period is used as the current thickness. Combined with the temperature and humidity forecast values ​​of that sub-period, a new thickness increment is recalculated based on the changing trend of the temperature and humidity time series data. This process is repeated to complete the thickness increment prediction for all sub-periods in the future time window. The predicted thickness increments of each sub-period are then added to the current frost thickness to generate a frost thickness growth prediction curve with time as the horizontal axis and frost thickness as the vertical axis. Finally, according to the preset frost severity classification standard (e.g., light frost is a frost thickness of 0-1mm, moderate frost is 1-2mm, and heavy frost is greater than 2mm), the frost thickness values ​​at each future moment in the prediction curve are graded and matched to obtain the frost severity rating result for the future period.

[0027] In an embodiment of the present invention, noise interference and abnormal fluctuations in the original data are filtered out by real-time collection and preprocessing of the frost thickness, ambient temperature and humidity parameters on the evaporator surface; then, the standardized input vector is imported into a pre-trained neural network model, and with the help of deep mining and high-dimensional mapping of the complex nonlinear relationship between multiple parameters by the hidden layer nodes, the key features reflecting the instantaneous frosting condition are accurately extracted, and then the accurate frosting rate is obtained through output layer processing, which breaks through the limitation of the traditional single-parameter model that it is difficult to capture the influence of multiple factors, and greatly improves the accuracy of the frosting rate calculation; finally, based on the real-time frosting rate and environmental time series data, a multi-step forward deduction is performed to iteratively generate a frost thickness growth prediction curve and complete the severity level assessment, so as to accurately predict the future frosting development trend in advance, thereby enhancing the system's active prevention and control capabilities and adaptability to frosting problems.

[0028] In a preferred embodiment of the present invention, the above step 2 may include: Step 2.1: Based on the prediction and severity results, the frost layer growth trend is determined and the expected time until the critical thickness is reached is calculated. Before the actual frost layer thickness reaches the critical thickness, a reverse cycle defrost start command is generated in advance based on the prediction results. Specifically, the following steps are performed: Based on the output frost thickness growth prediction curve and the severity level assessment results, the slope of the frost layer thickness change over time is calculated using a trend fitting algorithm to determine whether the frost layer growth trend is accelerating, constant, or decelerating. Furthermore, based on a preset critical frost layer thickness value (e.g., 1.5 mm, which can be adjusted based on the evaporator model and heat exchange efficiency requirements), the actual measured frost layer thickness is subtracted from the critical thickness and then divided by the current frost formation rate to calculate the expected time until the frost layer reaches the critical thickness. Based on this, a lead time threshold is set, such as 20% of the expected time. When the difference between the actual frost layer thickness and the critical thickness is greater than the thickness value corresponding to the lead time, the reverse cycle defrost start command is triggered, ensuring that the defrost operation is initiated before the frost layer actually reaches the critical thickness.

[0029] Step 2.2, the defrosting start instruction is issued to the air source heat pump unit control system, and the four-way valve is switched to control the refrigerant flow direction, and the reverse cycle defrosting process is started, specifically including: based on the generated reverse cycle defrosting start instruction, the start instruction is issued to the main control system of the air source heat pump unit in the form of electrical signal through the internal communication bus of the system, such as RS485 bus, the main control system sends a switching signal to the drive module of the four-way valve after receiving the instruction, and the valve core of the four-way valve is moved from the heating working condition position to the defrosting working condition position, so as to change the circulating direction of the refrigerant in the system, so that the high temperature and high pressure refrigerant originally flowing to the indoor heat exchanger preferentially flows to the outdoor evaporator, and the frost layer on the surface of the evaporator is heated by using the condensation heat release of the refrigerant, so as to start the reverse cycle defrosting process, and the response time of the whole instruction issuing and four-way valve switching is controlled within 5 seconds, so as to ensure the rapid start of the defrosting process.

[0030] Step 2.3, in the defrosting process, the opening change rate of the four-way valve and the rotating speed of the outdoor fan are dynamically adjusted according to the frost severity result, so as to match the defrosting heat demand under different frost degrees, realize the optimization control of the defrosting process, specifically including: after starting the reverse cycle defrosting process, the system first calls the output frost severity level, such as light, medium and heavy, and adjusts the opening change rate of the four-way valve and the rotating speed of the outdoor fan in real time according to the preset level and parameter correspondence, when the light frost is determined, the opening of the four-way valve is gradually increased at a slow rate, and the outdoor fan is maintained at 60%-70% of the rated rotating speed; when the medium frost is determined, the opening change rate of the four-way valve is accelerated, and the fan rotating speed is reduced to 40%-50% of the rated rotating speed; when the heavy frost is determined, the opening change rate of the four-way valve is further improved, and the fan rotating speed is reduced to less than 20% of the rated rotating speed or even stopped.

[0031] Step 2.4, monitor the evaporator surface temperature and system pressure parameters in real time, judge the defrosting completion, and when it is confirmed that the frost layer on the evaporator surface is completely cleared and the system parameters return to the normal operating range, generate a defrosting end instruction, control the four-way valve and the fan to return to the heating operation state, and obtain the system state after defrosting. Specifically, it includes: while performing dynamic defrosting control, synchronously executing defrosting completion judgment and system recovery operations, collecting its surface temperature in real time through the temperature sensor arranged on the evaporator surface, and monitoring the pressure parameters of the high-pressure side of the system through the pressure transmitter; when the evaporator surface temperature is monitored to rise to 5 for 30 consecutive seconds, the defrosting is completed. ℃ or above, which can ensure that the frost layer is completely melted, and when the pressure on the high-pressure side of the system is stable within the normal operating range of 1.8-2.2MPa, it is determined that the frost layer has been completely cleared and the defrosting process has achieved the expected effect. At this time, the system generates a defrost end instruction, which is also sent to the unit control system through the internal communication bus to control the valve core of the four-way valve to switch from the defrosting condition position back to the heating condition position, and restore the speed of the outdoor fan to the set value during heating operation, such as the speed automatically adjusted according to the ambient temperature, so that the entire heat pump system re-enters the heating operation state. At this point, the system is in a stable operating state after defrosting.

[0032] In an embodiment of the present invention, through early predictions and severity results, the growth trend of the frost layer is accurately judged, the time to reach the critical thickness is calculated in advance, and a defrost instruction is generated before the frost layer actually reaches the critical value, effectively preventing the frost layer from being too thick and seriously affecting the heat exchange of the evaporator, thereby ensuring stable heating capacity; the instruction is quickly conveyed to the air source heat pump unit control system, and the four-way valve is manipulated to quickly switch the refrigerant flow direction and start defrosting, which greatly shortens the response time and reduces the risk of continued thickening of the frost layer; according to the severity of frost, the four-way valve opening change rate and the outdoor fan speed are flexibly adjusted to achieve accurate supply of defrost heat and avoid energy waste; by real-time monitoring of the evaporator surface temperature and system pressure parameters, the defrost completion status is accurately judged to ensure that the defrost is thorough and not excessive. After defrosting, the system can quickly restore stable heating, reduce indoor temperature fluctuations, and improve heating comfort and system operation efficiency.

[0033] In a preferred embodiment of the present invention, the above step 3 may include: Step 3.1, based on the system state after defrosting, and accordingly activate the grid voltage monitoring function, real-time acquisition of grid voltage data through voltage sensor, and voltage data filtering and trend analysis to predict voltage change trend, specifically including: first, the obtained system state after defrosting as a trigger condition, when the system confirms that the four-way valve and the fan have returned to the heating operation state, and the evaporator surface temperature and system pressure parameters are stable in the normal interval, send an activation signal to the grid voltage monitoring module to switch from standby state to working state; on this basis, through the voltage sensor installed in the compressor power supply circuit, real-time acquisition of instantaneous value data of grid input voltage at a frequency not less than 10 times / sec; then, the original voltage data collected is processed by moving average filtering method to filter out high-frequency noise caused by grid interference to obtain a smooth voltage curve; further, by analyzing the change slope and fluctuation amplitude of the voltage curve in the last 30 seconds, it is judged whether the voltage is currently rising, falling or stable, so as to predict the voltage change trend in the next 5-10 seconds.

[0034] Step 3.2, before analyzing and predicting that the grid voltage will drop to the minimum voltage threshold required for the normal operation of the compressor, generate a pre-compensation instruction, specifically including: based on the obtained voltage change trend prediction result, the system pre-sets the minimum voltage threshold required for the normal operation of the compressor, which is usually 85% of the rated voltage, which can be adjusted according to the compressor model, and compares the predicted future voltage value with the minimum threshold in real time; when it is found through trend analysis that the voltage will drop to the minimum threshold in the next 0.5-1 second, and this falling trend is persistent, such as showing voltage drop in 3 consecutive sampling periods, the central control unit of the system immediately generates a pre-compensation instruction, which contains the voltage amplitude information that needs to be compensated, to ensure that the compensation action is started before the voltage actually drops to the threshold.

[0035] Step 3.3, issue the pre-compensation instruction to the voltage compensation device to control it to run in advance to boost the power supply voltage and ensure that the voltage at the compressor end is maintained within the normal operating range, specifically including: based on the generated pre-compensation instruction, the pre-compensation instruction is issued to the voltage compensation device through the system's internal control bus, such as CAN bus; after receiving the instruction, the voltage compensation device runs according to the compensation amplitude information contained in the instruction, adjusts its output voltage gain, and injects compensation voltage into the compressor power supply circuit; wherein the size of the compensation voltage is determined according to the predicted voltage drop amplitude, so that the actual voltage at the input end of the compressor is always maintained within its normal operating range, i.e. 90%-110% of the rated voltage, to avoid abnormal operation of the compressor due to low voltage.

[0036] In step 3.4, while performing voltage compensation, dynamically calculate and generate compressor frequency adjustment instructions based on the amplitude and trend of voltage fluctuations. This fine-tuning of the compressor operating frequency balances the system load and power supply capacity, maintaining overall system stability. Specifically, while performing voltage compensation, the compressor frequency is adjusted simultaneously. The system records the amplitude of voltage fluctuations (i.e., the deviation between the actual voltage and the rated voltage) and the fluctuation trend, such as rapid drop, slow drop, or fluctuation amplitude, in real time. Based on a preset mapping relationship between voltage fluctuation and frequency adjustment (which is determined based on operating parameters provided by the compressor manufacturer and actual commissioning experience), the required compressor frequency adjustment amount for the current operating conditions is calculated. Subsequently, a corresponding frequency adjustment instruction is generated and issued to the compressor's variable frequency drive. This instruction fine-tunes the compressor's operating frequency to change its output power, thereby balancing the system load demand with the grid's power supply capacity. This coordinated operation of voltage compensation and frequency adjustment ultimately ensures that the heat pump system maintains stable operation despite grid voltage fluctuations, avoiding shutdowns or sudden drops in operating efficiency.

[0037] In an embodiment of the present invention, voltage monitoring is activated based on the system status after defrosting, and data is collected in real time by a voltage sensor. The voltage trend is predicted through filtering and trend analysis, so as to grasp the risk of power grid fluctuations in advance and avoid the impact of voltage drops on the equipment; before predicting that the voltage will drop to the minimum operating threshold of the compressor, a pre-compensation instruction is generated, so that the voltage compensation device intervenes in advance to increase the supply voltage, ensuring that the compressor terminal voltage is always within the normal range and preventing the compressor from shutting down due to undervoltage; at the same time, combined with the voltage fluctuation amplitude and trend, the compressor frequency is dynamically adjusted to balance the system load and power supply capacity, avoiding operational instability caused by mismatch between load and power supply. The entire process realizes early response and coordinated regulation of power grid fluctuations, reduces the risk of compressor failure, ensures continuous and stable operation of the system, avoids heating interruptions due to voltage problems, and takes into account both equipment reliability and heating continuity.

[0038] In a preferred embodiment of the present invention, the above step 4 may include: Step 4.1, based on the maintained stable operation state of the system, start the floor heating system monitoring sequence, and collect the water supply temperature and return water temperature data of the heating system in real time through the temperature sensor, specifically including: based on the maintained stable operation state of the system, confirm that the compressor terminal voltage is within the normal operating range, the compressor frequency is adjusted to a stable value, and there is no abnormal pressure fluctuation in the system, and then the start instruction of the floor heating system monitoring sequence can be triggered. After the monitoring sequence is started, it will drive the temperature sensors installed in specific positions of the water supply pipe and return pipe of the heating system, usually the water supply pipe near the heat pump outlet and the return pipe near the heat pump inlet, to enter the working state. The temperature sensor is preferably a platinum resistance sensor to ensure measurement accuracy. The sensor will obtain the instantaneous values ​​of the supply water temperature and the return water temperature in real time at a collection frequency of not less than 1 time / minute, and transmit the collected temperature data to the system for storage via wired or wireless communication.

[0039] Step 4.2 retrieves stored historical heating data. This historical data includes supply / return water temperature differences, heat loads, and system control parameters under different outdoor operating conditions. This data is then integrated with the real-time supply / return water temperature data to generate integrated data. Specifically, based on the continuous collection of real-time supply / return water temperature data, the system's central control unit sends a data call instruction to the data storage module, retrieving the stored historical heating data. This historical data must cover supply / return water temperature difference data under different outdoor operating conditions for at least one complete heating season, the actual building heat load under these conditions, and the system control parameters used to meet the heat load at that time, such as the buffer tank circulation pump speed and mixing valve opening. Subsequently, the retrieved historical heating data is integrated with the real-time supply / return water temperature data based on the correlation between outdoor operating conditions, supply / return water temperature difference, heat load, and control parameters. Specifically, the real-time supply / return water temperature data is matched and categorized with the supply / return water temperature difference data under the same or similar outdoor operating conditions in the historical data, generating integrated data that includes both real-time and historical information.

[0040] Step 4.3, based on the integrated data, calculate the corresponding relationship between the supply and return water temperature difference and the flow change per unit time, establish the real-time thermal dynamic response relationship of the system, and obtain the analysis results of the thermal dynamic response relationship, specifically including: based on the obtained integrated data, first extract the supply and return water temperature difference data and the corresponding system circulation flow data within a continuous time period, such as 1 hour, from the integrated data. The circulation flow data can be obtained by converting the speed of the buffer water tank circulation pump and the pump characteristic curve. Then, calculate the corresponding relationship between the change in the supply and return water temperature difference and the change in the circulation flow per unit time, such as 10 minutes, within the time period. For example, the corresponding change amplitude of the supply and return water temperature difference when the circulation flow increases by 10% is statistically analyzed. By analyzing and fitting multiple groups of such corresponding relationship data, the real-time thermal dynamic response relationship of the system under the current working conditions can be established. This relationship can clearly reflect the influence of the system circulation flow change on the supply and return water temperature difference, thereby forming the analysis results of the thermal dynamic response relationship.

[0041] Step 4.4, compare the real-time heat load demand with the analysis results of the thermal dynamic response relationship, calculate the target speed of the buffer water tank circulation pump and the target opening of the mixing valve required to achieve the supply and demand balance, and obtain the calculation results, specifically including: based on the analysis results of the thermal dynamic response relationship, the system needs to obtain the real-time heat load demand of the current building. This demand can be determined in two ways. One is to combine the real-time outdoor temperature and the indoor set temperature; the other is to monitor the deviation between the actual indoor temperature and the set temperature through the indoor temperature sensor, and reversely infer the required heat load supplement. Subsequently, the real-time heat load demand is compared with the obtained thermal dynamic response relationship analysis results to determine when Whether the heating capacity of the previous system can meet the real-time heat load demand. If the heating capacity is insufficient, it is necessary to determine the circulation flow adjustment range and water temperature adjustment range corresponding to the increased heating capacity; if the heating capacity is excessive, it is necessary to determine the adjustment range corresponding to the reduced heating capacity. According to the above judgment results, combined with the speed and flow characteristic curve of the buffer water tank circulation pump, the opening of the mixing valve and the water temperature adjustment characteristics, the target speed of the buffer water tank circulation pump required to achieve the balance between heat load supply and demand is calculated. For example, the current speed needs to be adjusted from 1500r / min to 1800r / min and the target opening of the mixing valve needs to be adjusted from 30% to 45%. Finally, a clear calculation result is obtained.

[0042] Step 4.5, generate control instructions according to the calculation results, dynamically adjust the speed of the buffer water tank circulation pump to change the system circulation flow, and adjust the opening of the mixing valve to control the mixed water temperature, so that the output heat of the heat pump is coordinated with the real-time heating demand of the building to obtain the final matching result, specifically including: according to the obtained calculation results, generate corresponding control instructions according to the calculation results, wherein the control instructions for the buffer water tank circulation pump include target speed information, and the control instructions for the mixing valve include target opening information, and then, send these control instructions to the variable frequency drive module of the buffer water tank circulation pump and the actuator of the mixing valve respectively, and the variable frequency drive module is connected to the actuator. After receiving the instruction, the output frequency will be adjusted to gradually adjust the speed of the circulation pump to the target speed, thereby changing the circulation flow of the system. The change in the circulation flow will directly affect the amount of heat transported through the floor heating pipeline per unit time; after receiving the instruction, the mixing valve actuator will adjust the opening of the valve core through the mechanical transmission structure, thereby controlling the mixing ratio of high-temperature supply water and low-temperature return water in the floor heating system, and then controlling the mixed water temperature delivered to the floor heating coil; through the coordinated dynamic adjustment of the circulation pump speed and the mixing valve opening, the heat output of the heat pump can be accurately adapted to the real-time heating needs of the building, and ultimately a matching result is obtained in which the heat pump output matches the heating demand.

[0043] In an embodiment of the present invention, floor heating monitoring is started based on the stable operation of the system, and supply and return water temperature data are collected in real time to ensure that the data can reflect the current heating situation; historical heating data is called and integrated with real-time data to avoid the limitations of single data; a thermal dynamic response relationship is established by calculating the relationship between the supply and return water temperature difference and the flow change, and the heat transfer law of the system is accurately grasped; the real-time heat load demand and response relationship analysis results are compared to calculate the target speed of the buffer water tank circulation pump and the target opening of the mixing valve to provide a clear basis for regulation; finally, the pump speed and valve opening are dynamically adjusted to make the heat output of the heat pump match the real-time heating demand of the building, avoid insufficient heat supply affecting comfort, and prevent excess heat supply causing energy waste.

[0044] In a preferred embodiment of the present invention, the above step 5 may include: Step 5.1: Based on the matching results of the heat pump output and heating demand, the distributed temperature monitoring network is activated. Real-time temperature data is collected through temperature sensors placed at multiple locations within the heating area. Specifically, based on the obtained matching results of the heat pump output and heating demand, the system's central control unit confirms that the heat pump output has been preliminarily coordinated with the building's real-time heating demand and that the system's heat output is stable. Only then can a startup command be issued to the distributed temperature monitoring network, switching the network from standby mode to data collection mode. The distributed temperature monitoring network consists of several temperature sensors, which must be arranged to cover key locations within the heating area. For example, sensors must be located in the center of the living room, near bedside tables and windows in each bedroom, in the kitchen operating area, and in the dry area of ​​the bathroom. Each independent space must contain at least two sensors, located close to the ground and 1.5 meters above the ground, to simulate the temperature of human activity areas. Digital temperature sensors are preferred to ensure data accuracy. Once activated, these sensors will collect real-time temperature data at the corresponding locations at a frequency of once every three minutes.

[0045] Step 5.2: Clean and format the collected multi-location temperature data, remove abnormal data points, and associate the temperature data of each monitoring point with its corresponding spatial location information to obtain a normalized temperature distribution data set. Specifically, based on the continuous collection of real-time temperature data at multiple locations, clean the received original temperature data. Specifically, first set a reasonable value range for the temperature data, usually 5°C-30°C, which can be adjusted according to the needs of the heating season. Values ​​outside this range are determined as abnormal data points, such as instantaneous ultra-high or ultra-low temperatures caused by sensor failure. For data points determined to be abnormal, they are not directly eliminated, but replaced with the average value of the normal data collected three times in a row to ensure the continuity of the data sequence. After cleaning, the data is formatted and the scattered temperature data is organized into structured data according to the fixed format of acquisition time, sensor number, corresponding position and temperature value. Furthermore, the temperature data of each sensor is associated with its pre-recorded spatial location information and marked in two ways: one is to record the functional name of the area where the sensor is located, and the other is to record its corresponding spatial coordinates. Through the association operation, a normalized temperature distribution data set containing three-dimensional information of time, position and temperature is finally formed.

[0046] Step 5.3, based on the obtained normalized temperature distribution dataset, the heating area is divided into several independent sub-regions according to the spatial structure, and the representative temperature value of each sub-region is calculated by using the weighted average algorithm to generate the sub-region temperature distribution data, which specifically includes: based on the obtained normalized temperature distribution dataset, further sub-region division and representative temperature calculation will be carried out, and in the specific operation, first, according to the actual spatial structure of the heating area, the sub-regions are divided, for example, based on the physical partition of the building such as walls, doors and windows, the whole heating area is divided into living room sub-region, master bedroom sub-region, secondary bedroom sub-region, kitchen sub-region, bathroom sub-region and several independent sub-regions, each of which corresponds to a group of associated temperature data in the normalized dataset. After the division is completed, the representative temperature value of each sub-region is calculated by using the weighted average algorithm, specifically, first, determine the weight according to the layout position of the sensor in each sub-region, for example, the weight of the sensor at a height of 1.5 m from the ground is set to 0.6, and the weight of the sensor close to the ground is set to 0.4; then multiply the temperature data of each sensor in the sub-region by its corresponding weight, and divide the sum by the total weight to obtain the representative temperature value of the sub-region. By executing the above calculation on all sub-regions one by one, the sub-region temperature distribution data containing the names of each sub-region and the corresponding representative temperature is finally generated.

[0047] Step 5.4, based on the generated sub-region temperature distribution data, combined with the spatial structure characteristics and thermal characteristic parameters of the building, the spatial correlation and variation law of the temperature data of each sub-region are analyzed, and the thermal field characteristic distribution of the heating area is established, which specifically includes: according to the generated sub-region temperature distribution data, the establishment of the thermal field characteristic distribution of the heating area is completed by combining the characteristics of the building itself. First, the system will call the pre-stored spatial structure characteristics and thermal characteristic parameters of the building, wherein the spatial structure characteristics include the wall thickness of each sub-region, the size and orientation of doors and windows, the floor height, etc., and the thermal characteristic parameters include the thermal conductivity of the wall insulation material, the heat transfer coefficient of the window, the thermal resistance of the roof and the ground, etc. Subsequently, based on these parameters, the spatial correlation of the temperature data of each sub-region is analyzed, for example, whether the temperature difference between the south-facing bedroom sub-region and the north-facing bedroom sub-region is caused by the different solar radiation heating caused by the orientation, whether there is mutual influence between the temperature of the living room sub-region and the adjacent kitchen sub-region, such as the heating effect of the kitchen on the living room temperature; at the same time, by continuously monitoring the change of the sub-region temperature with time, such as recording the representative temperature once an hour, the variation law is analyzed, for example, the corresponding drop amplitude of the temperature of each sub-region when the outdoor temperature drops by 2℃, to judge the difference in thermal stability of different sub-regions; through the comprehensive analysis of the spatial correlation and the variation law, the thermal field characteristic distribution that can directly reflect the temperature distribution difference, heat transfer characteristics and stability in the heating area is finally established in the form of temperature gradient chart or partition temperature distribution table.

[0048] In an embodiment of the present invention, distributed temperature monitoring is started based on the matching results between the heat pump and the heating demand, and data is collected by multi-position sensors to comprehensively capture the actual temperature distribution in the heating area, avoiding the limitations of single-point monitoring; the collected data is cleaned and formatted and associated with the spatial position to eliminate the interference of abnormal values ​​and ensure the accuracy and standardization of the temperature distribution data set; the sub-areas are divided according to the space and the representative temperature is calculated to make the temperature data more in line with the actual heating zoning demand and facilitate targeted regulation; the sub-area temperature correlation and law are analyzed in combination with the building structure and thermal parameters, and the thermal field characteristic distribution is established to clearly grasp the heat transfer characteristics and temperature change trends of each area.

[0049] In a preferred embodiment of the present invention, the above step 5 may include: Step 5.5, based on the thermal field characteristic distribution, extract the thermodynamic characteristic parameters of each sub-area. The thermodynamic characteristic parameters include temperature stability, thermal inertia coefficient and thermal response time constant. Specifically, based on the established thermal field characteristic distribution of the heating area, first, retrieve the representative temperature data of each sub-area for 24 consecutive hours from the thermal field characteristic distribution, and calculate the temperature stability based on this. The difference between the maximum and minimum temperatures of the sub-area in this time period is counted. The smaller the difference, the higher the temperature stability. At the same time, the frequency of temperature fluctuations is recorded as an auxiliary evaluation indicator of temperature stability. To extract the thermal inertia coefficient, it is necessary to combine the thermal characteristic parameters of the building, such as the wall thickness of the sub-area, the thermal conductivity of the insulation material and the temperature change data. When the system adjusts the heating output, the time required for the sub-area temperature to change from the initial value to the new stable value is recorded. The longer the time, the larger the thermal inertia coefficient. It can be determined by comparing the temperature response lag time of different sub-areas under the same heating adjustment amplitude. To extract the thermal response time constant, it is necessary to monitor the time required for the sub-area temperature to reach 63.2% of the final stable temperature thermodynamic characteristic after the heating is started or stopped. This time value is the thermal response time constant. By taking the average of the monitoring data of multiple rounds of heating start and stop processes, the accuracy of parameter extraction is ensured, and finally a parameter set is formed including the temperature stability level of each sub-area, the specific thermal inertia coefficient value and the thermal response time constant.

[0050] Step 5.6, according to the extracted thermodynamic characteristic parameters, combined with the use function characteristics and comfort requirements of each sub-region, generate system control parameters for each sub-region, including: based on the extraction of the thermodynamic characteristic parameters of each sub-region, combined with the use function characteristics and comfort requirements of each sub-region, generate dedicated system control parameters, first, the use function characteristics of each sub-region are determined, for example, the master bedroom and the secondary bedroom are sleep and rest areas, and the use period is concentrated in the thermodynamic characteristics 22:00-7:00; the living room is a daily activity area, and the use period is concentrated in the thermodynamic characteristics 8:00-22:00; the kitchen is a short-time operation area, and the use period is concentrated in the thermodynamic characteristics 1 hour before and after each meal; the bathroom is a washing area, and the use period is scattered and needs to maintain a basic temperature, at the same time, according to different function areas, corresponding comfort requirements are set: the temperature of the sleep area needs to be stable at 18-20℃, and the fluctuation amplitude is not more than ±0.5℃; the temperature of the activity area needs to be maintained at 22-24℃, and the fluctuation amplitude is not more than ±1℃; the temperature of the short-time use area can be controlled at 16-18℃, and the basic temperature is not lower than 15℃, then, combined with the thermodynamic characteristic parameters, the control parameters are generated, for the sub-regions with large thermal inertia coefficient and long thermal response time, such as the master bedroom with thick wall and good heat preservation, the heating start-stop advance is set to 30-40 minutes, to avoid the influence of temperature response lag on comfort; for the sub-regions with poor temperature stability and easy to be disturbed by the outside world, such as the secondary bedroom near the window, the temperature fluctuation threshold is set to ±0.3℃, and the sensitivity of the triggered control is higher; finally, the dedicated system control parameters of each sub-region are formed, including the target temperature range, the temperature fluctuation threshold, the heating start-stop advance, and the control sensitivity.

[0051] Step 5.7, based on the generated system control parameters, combined with historical heating data and usage habits, formulate a regional timed temperature control strategy, specifically including: based on the generated exclusive system control parameters for each sub-region, the system will combine historical heating data with user usage habits to formulate a regional timed temperature control strategy. First, the system calls the historical heating data stored in the data storage module. The data must cover the temperature adjustment records of each sub-region in at least one complete heating season, the energy consumption data of the corresponding time period, and the optimal heating parameters under different outdoor working conditions. At the same time, the user usage habits are obtained through the user interaction interface or historical operation records. For example, the user is accustomed to using the living room from 7:00-8:00 and 18:00-22:00 on weekdays and using the living room all day on weekends; and is accustomed to lowering the temperature of the master bedroom to 18 from 23:00-6:30. ℃. Then, based on time, each sub-area is divided into different control periods, each matched with a corresponding target temperature and control rules. For example, in the living room, during weekdays, the target temperature is set to 23℃ from 7:00-8:00 and 18:00-22:00, with a medium control sensitivity setting. From 22:00-7:00, the target temperature is set to 18℃, with a heating start-stop lead time of 20 minutes. On weekends, the target temperature is set to 23℃ from 10:00-21:00, with the same rules for weekdays for the rest of the period. For the master bedroom, the target temperature is set to 18℃ from 23:00-6:30, and to 20℃ from 6:30-23:00. Due to high thermal inertia, the heating start-stop lead time is set to 40 minutes to ensure that the temperature meets the target during user usage hours. This completes the timed temperature control strategy for all sub-areas.

[0052] Step 5.8, the formulated sub-regional timing temperature control strategy is issued to the intelligent control devices of each heating circuit through Internet of Things communication to realize independent and accurate control of each heating circuit, specifically including: after the completion of the sub-regional timing temperature control strategy, the strategy is issued and executed through Internet of Things communication, and in specific operation, the system central control unit first splits the formulated sub-regional timing strategy according to sub-regions, converts it into an instruction format thermodynamic characteristic recognizable by the intelligent control device, and the instruction content includes sub-region number, corresponding heating circuit identifier, start and end time of each control period, target temperature value, and adjustment instruction when the temperature deviation exceeds the limit, such as increasing the opening degree of the electric regulating valve by 10% when the temperature is 1°C lower than the target value; then, through industrial Internet of Things communication, preferably LoRa or NB-IoT communication protocol, the communication stability in low temperature and long distance environment is ensured, and the split instructions are issued to the intelligent control devices of each heating circuit, wherein each sub-region corresponds to an independent heating circuit, and the intelligent control devices in the circuit include electric regulating valves and intelligent temperature controllers, and the devices have been pre-bound with the identifiers of sub-regions and heating circuits. After receiving the instructions, the intelligent control devices automatically store and load the strategy, and perform control actions according to the time period and temperature requirements in the instructions, for example, automatically adjusting the opening degree of the electric regulating valve to change the heating capacity at the set time period, realizing independent and accurate control of each heating circuit, and avoiding mutual interference of different sub-regions due to control.

[0053] Step 5.9, monitor the temperature changes and energy consumption data of each sub-area in real time, dynamically optimize the control parameters and control strategies according to the actual operating results, and achieve a balance between energy efficiency optimization and comfortable heating, specifically including: while achieving independent and precise control of each heating circuit, the thermodynamic feature will ensure that the system is always in a balanced state of energy efficiency and comfort through real-time monitoring and dynamic optimization. Specifically, first start the two-dimensional monitoring mechanism. On the one hand, through the temperature sensors in each sub-area in the same way as step 5.1, the actual temperature data is collected at a frequency of 1 times / 5 minutes, and the actual temperature is compared with the target temperature set by the strategy in real time, and the temperature deviation duration is recorded; on the other hand, through the energy consumption metering devices installed in each heating circuit, such as ultrasonic heat meters, the energy consumption data is collected at a frequency of 1 times / hour. According to the data, the heat consumption per unit time of each sub-area is counted; then, the actual operation effect is evaluated according to the monitoring data. If the actual temperature of a sub-area continues to meet the target requirements and the energy consumption is at a relatively low level in the historical period, it means that the current control parameters are adapted to the strategy and no adjustment is required; if the sub-area frequently experiences temperature deviations, such as the heating lag in the sub-area with large thermal inertia, the heating start and stop advance time should be appropriately increased, such as from 30 minutes to 40 minutes; if the energy consumption is too high but the temperature has reached the standard, the target temperature upper limit should be appropriately lowered or the maximum opening of the electric control valve should be reduced. Through the cyclic optimization mechanism of thermodynamic characteristic monitoring, thermodynamic characteristic evaluation and adjustment, the system control parameters and sub-region timing strategies of each sub-area are dynamically updated, so as to finally achieve a two-way balance between system energy efficiency optimization and user heating comfort, and avoid energy waste or lack of comfort caused by fixed strategies.

[0054] In an embodiment of the present invention, thermodynamic parameters such as temperature stability of each sub-region are extracted from the distribution of thermal field characteristics to accurately grasp the differences in thermal characteristics of different regions; then, exclusive control parameters are generated in combination with the use functions and comfort requirements of the sub-region to avoid the drawbacks of one-size-fits-all thermodynamic control; then, a regional timed temperature control strategy is formulated in combination with historical data and usage habits to make the control more in line with actual needs; then, the strategy is sent to smart devices through the Internet of Things to achieve independent and precise control of each heating circuit to ensure that the temperature in different regions meets the standard; finally, the temperature and energy consumption are monitored in real time and dynamically optimized to avoid temperature fluctuations affecting comfort and reduce unnecessary energy consumption.

[0055] like Figure 2 As shown, an embodiment of the present invention further provides an air source heat pump heating system, comprising: The acquisition module is used to collect the frost thickness, ambient temperature, and humidity parameters of the evaporator surface in real time, and input them into a pre-trained neural network algorithm model. By analyzing the nonlinear relationship between multiple factors, the frost formation rate is identified, the future frost development trend is predicted, and the frost thickness growth prediction and severity results are output. Based on the frost thickness growth prediction and severity results, the reverse cycle defrost is started in advance before the frost layer reaches the critical thickness. The defrost process is optimized by adjusting the four-way valve opening and the fan speed to obtain the system state after defrosting. The calculation module is used to monitor grid voltage fluctuations in real time based on the system status after defrosting, initiate voltage compensation before the voltage falls below the normal operating range of the compressor, and dynamically adjust the compressor frequency to maintain stable system operation. Based on the stable operation of the system, the module collects the supply and return water temperature data of the floor heating system, combines historical data with real-time heat load demand, establishes a thermal dynamic response relationship, and obtains a matching result between the heat pump output and the heating demand by adjusting the buffer water tank circulation pump speed and the mixing valve opening. The processing module is used to obtain multi-location temperature data through distributed temperature monitoring devices based on the matching results of heat pump output and heating demand, and establish the thermal field characteristic distribution of the heating area; generate system control parameters according to the thermodynamic characteristic parameters of each area, formulate regional timing temperature control strategies based on the system control parameters, and control each heating circuit through Internet of Things communication to achieve energy efficiency optimization and comfortable heating.

[0056] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. An air source heat pump heating method, characterized in that: The method comprises: Step 1: Real-time data collection of frost thickness, ambient temperature, and humidity on the evaporator surface is fed into a pre-trained neural network algorithm model. The model then analyzes the nonlinear relationship between multiple factors to identify the frost formation rate, predict future frost development trends, and output frost thickness growth predictions and severity results. Step 2: Based on the frost thickness growth prediction and severity results, reverse cycle defrosting is started in advance before the frost layer reaches a critical thickness. The defrost process is optimized by adjusting the four-way valve opening and the fan speed to obtain the system state after defrosting. Step 3: Based on the system status after defrosting, the grid voltage fluctuation is monitored in real time. Before the voltage drops below the normal operating range of the compressor, voltage compensation is initiated and the compressor frequency is dynamically adjusted to maintain stable system operation. Step 4: Based on the stable operation of the system, collect the supply and return water temperature data of the floor heating system. Combine the historical data with the real-time heat load demand to establish a thermal dynamic response relationship. By adjusting the speed of the buffer water tank circulation pump and the opening of the mixing valve, the matching result between the heat pump output and the heating demand is obtained. Step 5: Based on the matching results between the heat pump output and the heating demand, obtain temperature data at multiple locations through distributed temperature monitoring devices to establish the thermal field characteristic distribution of the heating area; generate system control parameters according to the thermodynamic characteristic parameters of each area, formulate a regional timed temperature control strategy based on the system control parameters, and control each heating circuit through Internet of Things communication to achieve energy efficiency optimization and comfortable heating.

2. The air source heat pump heating method according to claim 1, characterized in that: The frost thickness, ambient temperature, and humidity parameters of the evaporator surface are collected in real time and input into a pre-trained neural network algorithm model. By analyzing the nonlinear relationship between multiple factors, the frost rate is identified, the future frost development trend is predicted, and the frost thickness growth prediction and severity results are output, including: Step 1.1, real-time collection of evaporator surface frost thickness, ambient temperature and humidity parameters, pre-processing the collected raw data parameters to obtain pre-processed data; Step 1.2: Based on the preprocessed data, the data is organized into a standardized input vector containing ambient temperature, humidity, and frost thickness in chronological order. This standardized input vector is then fed into a pretrained neural network model. The internal hidden layer nodes perform high-dimensional nonlinear mapping and feature extraction on the complex nonlinear interactions between the parameters in the input vector. This calculation results in implicit state features that reflect the current instantaneous frost formation condition. Based on these implicit state features, the output layer of the neural network model uses linear transformation and activation function processing to determine the precise frost formation rate under the current operating conditions. In step 1.3, based on the real-time frost formation rate and the time series data of the current ambient temperature and humidity, the neural network model performs a multi-step forward deduction calculation in the internal state space. It iteratively predicts the frost thickness increment at each moment in the future preset time window, accumulates it to the current thickness, and generates a frost thickness growth prediction curve with time as the independent variable, thereby obtaining the grade assessment result of the frost severity in the future period.

3. The air source heat pump heating method according to claim 2, characterized in that: Based on the frost thickness growth prediction and severity results, reverse cycle defrosting is started in advance before the frost layer reaches the critical thickness. The defrost process is optimized by adjusting the four-way valve opening and fan speed to obtain the system status after defrosting, including: Step 2.1: Based on the prediction and severity results, determine the frost growth trend and calculate the expected time to reach the critical thickness. Before the actual frost thickness reaches the critical thickness, generate a reverse cycle defrost start instruction in advance based on the prediction results; Step 2.2: Send the defrost start command to the air source heat pump unit control system to control the four-way valve to switch the refrigerant flow direction and start the reverse cycle defrost process; Step 2.3: During the defrost process, the opening rate of the four-way valve and the speed of the outdoor fan are dynamically adjusted according to the frost severity results to match the defrost heat requirements under different frost severity levels, thereby achieving optimized control of the defrost process. Step 2.4: monitor the evaporator surface temperature and system pressure parameters in real time to determine the degree of defrosting completion. When it is confirmed that the frost layer on the evaporator surface is completely cleared and the system parameters have returned to the normal operating range, generate a defrost end instruction, control the four-way valve and fan to return to the heating operation state, and obtain the system state after defrosting.

4. The air source heat pump heating method according to claim 3, characterized in that: Based on the system status after defrosting, the system monitors grid voltage fluctuations in real time, initiates voltage compensation before the voltage drops below the normal operating range of the compressor, and dynamically adjusts the compressor frequency to maintain stable system operation, including: Step 3.1: Based on the post-defrost system status, the grid voltage monitoring function is activated accordingly. The grid voltage data is collected in real time through the voltage sensor, and the voltage data is filtered and trend analyzed to predict voltage change trends. Step 3.2: before the grid voltage is predicted to drop to the minimum voltage threshold required for the normal operation of the compressor through analysis, a pre-compensation instruction is generated; Step 3.3: Send a pre-compensation instruction to the voltage compensation device to control it to start operation in advance to increase the power supply voltage and ensure that the compressor terminal voltage remains within the normal operating range; In step 3.4, while performing voltage compensation, the compressor frequency adjustment instruction is dynamically calculated and generated based on the amplitude and trend of the voltage fluctuation. The system load and power supply capacity are balanced by fine-tuning the compressor operating frequency to maintain the overall stable operation of the system.

5. The air source heat pump heating method according to claim 4, characterized in that: Based on the stable operation of the system, the supply and return water temperature data of the floor heating system is collected. By combining historical data with real-time heat load demand, a thermal dynamic response relationship is established. By adjusting the speed of the buffer water tank circulation pump and the opening of the mixing valve, the matching result between the heat pump output and the heating demand is obtained, including: Step 4.1: Based on the maintained stable system operation state, start the floor heating system monitoring sequence and collect the supply water temperature and return water temperature data of the heating system in real time through the temperature sensor; Step 4.2: Call the stored historical heating data, which includes the supply and return water temperature difference, heat load, and system adjustment parameters under different outdoor working conditions, and integrate the historical data with the real-time supply and return water temperature data to obtain the integrated data; Step 4.3: Based on the integrated data, calculate the corresponding relationship between the supply and return water temperature difference and the flow rate change per unit time, establish the real-time thermal dynamic response relationship of the system, and obtain the analysis results of the thermal dynamic response relationship; Step 4.4: Compare the real-time heat load demand with the analysis results of the thermal dynamic response relationship, calculate the target speed of the buffer water tank circulation pump and the target opening of the mixing valve required to achieve supply and demand balance, and obtain the calculation results; In step 4.5, control instructions are generated based on the calculation results. The speed of the buffer water tank circulation pump is dynamically adjusted to change the system circulation flow rate. At the same time, the opening of the mixing valve is adjusted to control the mixed water temperature. This ensures that the heat output of the heat pump is coordinated with the real-time heating demand of the building, resulting in the final matching result.

6. The air source heat pump heating method according to claim 5, characterized in that: Based on the matching results between heat pump output and heating demand, distributed temperature monitoring devices are used to obtain temperature data at multiple locations and establish the thermal field characteristic distribution of the heating area, including: Step 5.1: Based on the matching result between the heat pump output and the heating demand, the distributed temperature monitoring network is started to collect real-time temperature data through temperature sensors arranged at multiple locations in the heating area; Step 5.2: Clean and format the collected multi-location temperature data, remove abnormal data points, and associate the temperature data of each monitoring point with its corresponding spatial location information to obtain a normalized temperature distribution data set; Step 5.3: Based on the obtained normalized temperature distribution data set, the heating area is divided into several independent sub-areas according to the spatial structure. The representative temperature value of each sub-area is calculated using the weighted average algorithm to generate the sub-area temperature distribution data; In step 5.4, based on the generated sub-region temperature distribution data, combined with the spatial structural characteristics and thermal characteristic parameters of the building, the spatial correlation and change law of the temperature data of each sub-region are analyzed to establish the thermal field characteristic distribution of the heating area.

7. The air source heat pump heating method according to claim 6, characterized in that: Generate system control parameters based on the thermodynamic characteristic parameters of each area, formulate regional timed temperature control strategies based on the system control parameters, and control each heating circuit through IoT communication to achieve energy efficiency optimization and comfortable heating, including: Step 5.5, based on the thermal field characteristic distribution, extract the thermodynamic characteristic parameters of each sub-region, the thermodynamic characteristic parameters including temperature stability, thermal inertia coefficient and thermal response time constant; Step 5.6: Based on the extracted thermodynamic characteristic parameters, combined with the functional characteristics and comfort requirements of each sub-area, generate system control parameters for each sub-area; Step 5.7: Based on the generated system control parameters, combined with historical heating data and usage habits, formulate a regional timed temperature control strategy; In step 5.8, the developed regional timed temperature control strategy is distributed to the intelligent control devices of each heating circuit via IoT communication, achieving independent and precise control of each heating circuit; Step 5.9: Monitor the temperature changes and energy consumption data of each sub-area in real time, and dynamically optimize the regulation parameters and control strategies based on the actual operating results to achieve a balance between energy efficiency optimization and comfortable heating.

8. An air source heat pump heating system, which implements the air source heat pump heating method according to any one of claims 1 to 7, characterized in that: include: The acquisition module is used to collect the frost thickness, ambient temperature, and humidity parameters of the evaporator surface in real time, and input them into a pre-trained neural network algorithm model. By analyzing the nonlinear relationship between multiple factors, the frost formation rate is identified, the future frost development trend is predicted, and the frost thickness growth prediction and severity results are output. Based on the frost thickness growth prediction and severity results, the reverse cycle defrost is started in advance before the frost layer reaches the critical thickness. The defrost process is optimized by adjusting the four-way valve opening and the fan speed to obtain the system state after defrosting. The calculation module is used to monitor grid voltage fluctuations in real time based on the system status after defrosting, initiate voltage compensation before the voltage falls below the normal operating range of the compressor, and dynamically adjust the compressor frequency to maintain stable system operation. Based on the stable operation of the system, the module collects the supply and return water temperature data of the floor heating system, combines historical data with real-time heat load demand, establishes a thermal dynamic response relationship, and obtains a matching result between the heat pump output and the heating demand by adjusting the buffer water tank circulation pump speed and the mixing valve opening. The processing module is used to obtain multi-location temperature data through distributed temperature monitoring devices based on the matching results of heat pump output and heating demand, and establish the thermal field characteristic distribution of the heating area; generate system control parameters according to the thermodynamic characteristic parameters of each area, formulate regional timing temperature control strategies based on the system control parameters, and control each heating circuit through Internet of Things communication to achieve energy efficiency optimization and comfortable heating.

9. A computing device, characterized in that include: one or more processors; A storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, so that the one or more processors implement the air source heat pump heating method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a program, which, when executed by a processor, implements the air source heat pump heating method according to any one of claims 1 to 7.

Images