A heat pipe ground source heat pump system and its operation and management method

Abstract

This application relates to a heat pipe energy pile and a heat pipe ground source heat pump system and its operation and management method, solving the technical problems of existing ground source heat pumps requiring additional power and consuming high energy, and being difficult to adapt to bidirectional energy supply in winter and summer. The system includes a pile foundation, heat pipes, a heat exchanger, and a heat pump unit. The heat pipe is integrated into the pile foundation and is a closed tube with a capillary wick, vacuumed, and filled with a phase change working fluid. It consists of an evaporation section at the bottom / side of the pile foundation and a condensation section at the top. The evaporation section exchanges heat with the soil, the condensation section is connected to the heat exchanger, and the heat pump ground source side is connected to the heat exchanger. The capillary wick drives the working fluid to flow back autonomously, realizing heat transfer or release. This application has the following effects: it resolves the contradiction between low energy consumption and year-round energy supply in existing technologies, achieves efficient bidirectional energy supply in winter and summer without additional power, and significantly reduces the energy consumption of building cold and heat source systems.

Description

Technical Field

[0001] This invention relates to the field of building energy conservation technology, and in particular to a heat pipe ground source heat pump system and its operation and management method. Background Technology

[0002] Shallow geothermal energy, as a renewable and clean energy source, has the advantages of wide distribution, stable operation, and high utilization efficiency. It has become an important research direction for building energy conservation and sustainable development. The ground source heat pump system is its core development and utilization form. It can meet the building heating, cooling and domestic hot water needs by exchanging energy with the soil through underground heat exchange pipelines.

[0003] Currently, the mainstream ground source heat pump systems adopt buried pipe heat exchange methods such as single U-tube, double U-tube, or spiral tube. In areas with scarce land resources, researchers have proposed the "energy pile" technology, which combines building pile foundations with ground source heat pumps, enabling the pile foundations to have both load-bearing and water circulation heat exchange functions. At the same time, heat pipe technology, due to its high thermal conductivity, lack of mechanical moving parts, and ease of maintenance, has shown the potential to improve heat transfer efficiency in the field of ground source heat pumps.

[0004] However, existing technologies have significant shortcomings: traditional buried pipe systems and water circulation energy piles rely on fluid circulation to transfer heat, requiring additional energy to maintain fluid flow. Furthermore, they are prone to unstable heat exchange efficiency due to changes in pipeline resistance and fluctuations in fluid flow, resulting in high energy costs over the long term. Some technical solutions that attempt to combine heat pipes are limited by the heat transfer characteristics of existing heat pipes, and can only achieve unidirectional heat transfer under single operating conditions (such as absorbing heat from the soil in winter). They cannot meet the reverse demand of releasing building waste heat to the soil in summer, making it difficult for the system to cover bidirectional energy supply scenarios for heating and cooling throughout the year. Summary of the Invention

[0005] In order to resolve the contradiction between low energy consumption and year-round energy supply in existing technologies, achieve efficient energy supply in both winter and summer without additional power, and significantly reduce the energy consumption of building heating and cooling systems, this application provides a heat pipe ground source heat pump system and its operation and management method.

[0006] Firstly, this application provides a heat pipe ground source heat pump system, which adopts the following technical solution:

[0007] A heat pipe ground source heat pump system, comprising:

[0008] Pile foundation;

[0009] At least one heat pipe is attached to the pile foundation. The heat pipe is a heat pipe with a capillary structure and is a closed tube body with an internal vacuum and filled with a phase change working fluid.

[0010] The heat pipe includes an evaporation section located at the lower part and / or side of the pile foundation, and a condensation section located at the upper part of the pile foundation;

[0011] The heat exchanger is in fluid communication with the condenser section of the heat pipe;

[0012] The heat pump unit has its ground source side in fluid communication with the heat exchanger;

[0013] The evaporation section of the heat pipe is used for heat exchange with the soil around the pile, and the condensation section of the heat pipe transfers heat to the heat pump unit or releases heat from the heat pump unit to the soil through a heat exchanger.

[0014] The capillary wick structure is configured to drive the liquid phase change working fluid to autonomously reflux between the evaporation and condensation sections via capillary force.

[0015] By adopting the above technical solutions, the heat pipe of this system can drive the phase change working fluid to flow back autonomously by capillary force due to its capillary wick structure, without the need for additional power, thus significantly reducing energy consumption; the evaporation section exchanges heat efficiently with the soil around the pile, and the condensation section interacts with the heat pump unit through the heat exchanger. In winter, it can transfer soil heat to the heat pump, and in summer, it can release the waste heat of the heat pump to the soil, realizing bidirectional heat transfer in winter and summer, stably adapting to the annual energy supply demand, and improving the system's operating efficiency and stability.

[0016] Secondly, this application provides an operation and management method for a heat pipe ground source heat pump system, which adopts the following technical solution:

[0017] A method for operating and managing a heat pipe ground source heat pump system includes:

[0018] The system collects real-time operating and environmental data of the heat pipe ground source heat pump system through a preset sensor matrix, and performs data preprocessing to form a system operating status feature set. The operating data includes heat pipe operating pressure and temperature, inlet and outlet temperatures of the medium in the heat exchanger, and operating power and load rate of the heat pump unit. The environmental data includes building indoor temperature, building load information, outdoor meteorological parameters, and soil temperature distribution.

[0019] The system compares the building's indoor temperature and outdoor meteorological parameters, which are part of the system's operational status feature set, with preset mode determination thresholds to determine the system's current operating mode, which includes heating mode and cooling mode. Simultaneously, based on the time series data of soil temperature distribution, the system simulates the changes in the soil temperature field within a set future period using a heat conduction model. The magnitude and rate of soil temperature rise and fall in the simulation results are compared with preset safety thresholds, and the risk level of soil thermal imbalance is quantified based on the degree to which the thresholds are exceeded.

[0020] Based on the determined operating mode and the assessed soil thermal imbalance risk level, the corresponding target control parameter set is called from the preset control parameter library to generate control commands;

[0021] The generated control commands are sent to the corresponding actuators to drive the heat pump unit, the device for regulating the flow rate of the heat exchanger medium, and the heat pipe array to operate according to the commands.

[0022] By adopting the above technical solution, multi-dimensional data can be collected in real time through a sensor matrix to fully understand the system operation and environmental status; accurately determine the heating / cooling mode and quantify the risk of soil thermal imbalance; then generate control commands according to the corresponding parameters to drive the equipment, which can adapt to winter and summer conditions, improve system operating efficiency, avoid soil thermal imbalance, and ensure long-term stable operation of the system. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the overall structure of a heat pipe ground source heat pump system according to an embodiment of this application.

[0024] Figure 2 This is a schematic diagram of the arrangement of heat pipes combined with pile foundation according to another embodiment of this application.

[0025] Figure 3 This is a schematic diagram of the arrangement of heat pipes combined with pile foundation according to another embodiment of this application.

[0026] Figure 4 This is a schematic diagram of the working principle of the heat pipe in an embodiment of this application.

[0027] Figure 5 This is a flowchart illustrating an operation and management method for a heat pipe ground source heat pump system according to an embodiment of this application. Detailed Implementation

[0028] The present application will be further described in detail below with reference to the accompanying drawings.

[0029] Reference Figure 1 The present application discloses a heat pipe ground source heat pump system, which mainly consists of pile foundation, heat pipe, heat exchanger and heat pump unit, forming a complete geothermal energy utilization system.

[0030] As a load-bearing component of the building foundation, the pile foundation uses cast-in-place concrete piles or prestressed pipe piles (such as prestressed high-strength concrete pipe piles, i.e., PHC piles). At least one heat pipe is integrated with the pile foundation to form a structurally and functionally integrated component. Specifically, the heat pipe can be arranged in various ways: such as... Figure 1 As shown, heat pipes can be directly installed inside the pile foundation, running longitudinally through the pile body; as Figure 2 As shown, the heat pipe can be connected to the U-shaped or W-shaped heat exchange pipes pre-embedded in the pile foundation to form a composite heat exchange system; such as Figure 3 As shown, the heat pipe can extend below the pile foundation and directly contact the deeper soil; in addition, the heat pipe can also be directly coupled to the building floor structure, with its condensation section embedded in the floor concrete.

[0031] This heat pipe is a sealed metal tube internally evacuated and filled with a phase change working fluid. Its inner wall has a capillary wick structure, which is configured to drive the liquid phase change working fluid to autonomously reflux between the evaporation and condensation sections via capillary force. For details, please refer to [reference needed]. Figure 4 The portion of the heat pipe located in the contact area with the underground soil constitutes the evaporation section, the outer wall of which is covered with a thermally conductive filler layer or equipped with thermally conductive fins; the portion of the heat pipe extending upwards above the ground surface constitutes the condensation section.

[0032] The condenser section of the heat pipe is fluidly connected to the heat exchanger via a sealed pipeline, forming a closed loop. A plate heat exchanger is preferred, with its other side connected to the ground source circuit of the heat pump unit. The load side of the heat pump unit is connected to the building's indoor terminal system to complete the final energy supply.

[0033] When the system is working, it relies on the phase change cycle of the working fluid inside the heat pipe to achieve efficient heat transfer, as detailed below:

[0034] During winter heating, the heat pipe evaporation section absorbs heat from the soil, causing the liquid working fluid inside the pipe to evaporate and vaporize. The generated steam rises to the condensation section under the action of pressure difference, where it releases latent heat. This heat is transferred to the ground source side medium of the heat pump unit through a heat exchanger. The medium, carrying heat, enters the heat pump unit and is further heated by the heat pump, ultimately supplying heating to the building. The condensed liquid working fluid, aided by the capillary force generated by the capillary structure, autonomously flows back to the evaporation section, completing a non-powered heat exchange cycle.

[0035] In summer cooling conditions, the above process proceeds in reverse: the waste heat generated by the heat pump unit is carried to the heat exchanger by the ground source medium; the heat is transferred to the condensing section of the heat pipe through the heat exchanger, causing the working fluid inside the heat pipe to evaporate; the steam transports the heat to the underground evaporation section, and finally releases it into the cooler soil to achieve heat dissipation; the condensed liquid working fluid also flows back to the condensing section through the capillary force of the capillary wick, ensuring the stable operation of the refrigeration cycle.

[0036] Reference Figure 5 This application also provides a method for operating and managing a heat pipe ground source heat pump system, including:

[0037] Step S100: Real-time acquisition of operating data and environmental data of the heat pipe ground source heat pump system through a preset sensor matrix, followed by data preprocessing to form a system operating status feature set. The operating data includes heat pipe operating pressure and temperature, inlet and outlet temperatures of the medium in the heat exchanger, and operating power and load rate of the heat pump unit. The environmental data includes building indoor temperature, building load information, outdoor meteorological parameters, and soil temperature distribution.

[0038] The system comprises several components: a sensor matrix, consisting of various sensors distributed across key locations within the heat pipe ground source heat pump system to collect real-time operational and environmental data; data preprocessing, which involves cleaning, standardizing, and filtering the raw data to remove noise and outliers, fill in missing data, standardize data formats, and extract relevant information to improve data quality and usability; and a system operating status feature set, which, after preprocessing, comprehensively reflects the system's current operating status, including both operational and environmental data.

[0039] Heat pipe operating pressure and temperature: High-precision pressure and temperature sensors are installed in the evaporation and condensation sections of the heat pipe to collect real-time pressure and temperature data of the working fluid inside the heat pipe. Inlet and outlet temperatures of the medium in the heat exchanger: Temperature sensors are installed on the inlet and outlet pipes of the heat exchanger to measure the inlet and outlet temperatures of the medium in the heat exchanger. Operating power and load rate of the heat pump unit: The operating power and load rate of the heat pump unit are collected in real-time using power sensors and load rate monitoring devices inside the heat pump unit.

[0040] Building Indoor Temperature: Temperature and humidity sensors are installed at different locations within the building to collect indoor temperature data in real time. Building Load Information: Power consumption of each energy-consuming device is monitored in real time through the building's power distribution system, and the usage time of the energy-consuming devices is statistically analyzed to obtain building load information. Outdoor Meteorological Parameters: Outdoor meteorological parameters such as temperature, humidity, wind speed, and solar radiation are collected in real time through weather stations installed outside the building. Soil Temperature Distribution: Multiple temperature sensors are pre-embedded in the soil around the pile foundation to form a soil temperature monitoring network, collecting soil temperature distribution data in real time.

[0041] Step S200: The building's indoor temperature and outdoor meteorological parameters in the system's operating status feature set are compared with preset mode determination thresholds to determine the current operating mode of the system, which includes heating mode and cooling mode. At the same time, based on the time series data of soil temperature distribution, the soil temperature field changes within a set period are simulated through a heat conduction model. The magnitude and rate of soil temperature rise and fall in the simulation results are compared with preset safety thresholds. The risk level of soil thermal imbalance is quantified based on the degree to which the threshold is exceeded.

[0042] The system includes: a mode determination threshold (a preset temperature threshold used to determine whether the system is currently in heating or cooling mode); a heat conduction model (a mathematical model used to simulate changes in the soil temperature field, based on Fourier's law of heat conduction); and a safety threshold (preset thresholds for the magnitude and rate of soil temperature change used to assess the risk level of soil thermal imbalance).

[0043] The process regarding the operating mode is explained below:

[0044] The system first acquires indoor building temperature and outdoor meteorological parameters through a preset sensor matrix, and then compares this data with preset mode determination thresholds to determine whether the system is currently in heating or cooling mode. Specifically, if the indoor building temperature is lower than the preset heating threshold (e.g., 18℃) and the outdoor temperature is lower than the low temperature threshold (e.g., 5℃), it is determined to be in heating mode; conversely, if the indoor building temperature is higher than the preset cooling threshold (e.g., 26℃) and the outdoor temperature is higher than the high temperature threshold (e.g., 30℃), it is determined to be in cooling mode.

[0045] The process for analyzing the risk level of soil thermal imbalance is described below:

[0046] The system utilizes time-series data of soil temperature distribution to simulate soil temperature field changes over a predetermined period (e.g., 24 hours) using a heat conduction model. It then compares the magnitude and rate of soil temperature change in the simulation results with preset safety thresholds. Based on the degree to which the thresholds are exceeded, the system quantifies the risk level of soil thermal imbalance, classifying it into low risk (both magnitude and rate of change are within the safety threshold), medium risk (exceeding the threshold by less than 30%), and high risk (exceeding the threshold by more than 30%).

[0047] The construction process of the heat conduction model is roughly as follows: First, soil temperature data and corresponding environmental parameters, such as outdoor temperature and humidity, are collected under different geological conditions and environments. This data is acquired in real time using temperature sensors embedded in the soil and weather stations. Then, this data is used to calibrate key parameters in the model (such as soil thermal conductivity and specific heat capacity), optimizing the parameters by minimizing the error between the model's predicted values ​​and actual measurements. Finally, the accuracy and reliability of the model are verified through methods such as cross-validation to ensure that it can accurately predict changes in soil temperature.

[0048] Step S300: Based on the determined operating mode and the assessed soil thermal imbalance risk level, the corresponding target control parameter set is called from the preset control parameter library to generate control instructions.

[0049] The system comprises the following components: Control Parameter Library: A pre-set collection of optimal control parameters for different operating modes and risk levels. Target Control Parameter Set: A specific set of control parameters retrieved from the control parameter library based on the current operating mode and risk level. Control Commands: Specific instructions generated by the system based on the target control parameter set to adjust the equipment's operating status.

[0050] For a detailed explanation of the process, please refer to steps S310 to S350, which will not be repeated here.

[0051] In step S400, the generated control command is sent to the corresponding actuator to drive the heat pump unit, the device for adjusting the flow rate of the heat exchanger medium, and the heat pipe array to operate according to the command.

[0052] Actuator: The hardware device in the system used to execute control commands, such as heat pump units, heat pipe arrays, etc.

[0053] The general process is described below:

[0054] 1. Control Command Transmission: Based on the control commands generated in step S300, the system transmits the commands to the corresponding actuators via a communication interface (such as a PLC controller). These commands include adjustments to the operating power of the heat pump unit, flow regulation of the circulating water pump, and start / stop control of the heat pipe array.

[0055] 2. Actuator Response: After receiving the control command, the actuator adjusts its operating status according to the command requirements. For example, a heat pump unit adjusts its operating power according to the command, a circulating water pump adjusts its flow rate, and a heat pipe array starts or stops some heat pipes according to the command.

[0056] The risk levels of soil thermal imbalance include:

[0057] Step S210: Based on soil temperature distribution data, calculate the current soil thermal state characteristic value using a preset soil thermal state characteristic value calculation method, and compare it with a preset primary warning threshold to determine whether there is a risk of soil thermal imbalance. If not, proceed to step S220; if yes, proceed to step S230.

[0058] Among them, the soil thermal state characteristic value is a quantitative indicator that reflects the current soil thermal state, calculated from soil temperature distribution data using a specific algorithm. The primary warning threshold is a preset benchmark value used to determine whether there is a risk of soil thermal imbalance.

[0059] The process is described below:

[0060] 1. Data Collection: The system collects real-time soil temperature data from temperature sensors embedded in the soil. This data includes temperature measurements at different depths and locations. Assuming a total of N sensors and each sensor measures a temperature of T... i (i=1,2,…,N).

[0061] 2. Eigenvalue calculation:

[0062] Average temperature: Calculate the average temperature measured by all sensors using the following formula: ,in, It is the average temperature.

[0063] Temperature variance: Calculates the variance of temperature data to assess the uniformity of temperature distribution. The formula is:

[0064] ,in, It is the temperature variance.

[0065] Temperature gradient: Calculate the temperature gradient between adjacent sensors to assess the drasticness of temperature changes. The formula is: ;in, It is the temperature gradient between the i-th sensor and the (i+1)-th sensor. and These are the temperatures measured by the (i+1)th and ith sensors, respectively. It is the depth difference between the two sensors.

[0066] 3. Risk Assessment: The calculated soil thermal state characteristic values ​​(average temperature, temperature variance, temperature gradient) are compared with preset primary warning thresholds. If any characteristic value exceeds the preset primary warning threshold, the system will determine that there is a risk of soil thermal imbalance and execute step S230; if all characteristic values ​​are within the threshold range, the system will execute step S220.

[0067] Step S220: Based on the magnitude and rate of soil temperature rise and fall in the simulation results, compare them with the preset safety threshold, and quantify the risk level of soil thermal imbalance according to the degree to which the threshold is exceeded.

[0068] Among them, soil temperature variation range: the maximum change in soil temperature within a set period, reflecting the range of soil temperature fluctuation. Soil temperature change rate: the amount of change in soil temperature per unit time, reflecting the severity of soil temperature change. Safety threshold: the preset maximum allowable values ​​for soil temperature variation range and rate, used to determine the risk level of soil thermal imbalance.

[0069] The necessary process is described below:

[0070] 1. Data Collection and Simulation: The system collects current soil temperature distribution data from soil temperature sensors and simulates soil temperature field changes over a set period (e.g., 24 hours) using a heat conduction model. The simulation results include predicted soil temperature values ​​at different depths and locations.

[0071] 2. Calculate the magnitude and rate of change:

[0072] Variation range: The maximum change in soil temperature during the simulation period is calculated using the following formula: ,in, It is the soil temperature at the end of the simulation. It simulates the soil temperature at the start.

[0073] Rate of change: The maximum rate of change of soil temperature is calculated using the following formula: ,in, and It refers to soil temperature at continuous points in time. It is a time interval.

[0074] 3. Risk Level Quantification: The calculated magnitude and rate of soil temperature change are compared with preset safety thresholds to quantify the risk level of soil thermal imbalance.

[0075] Low risk: The magnitude and rate of change are both within the safe threshold.

[0076] Medium risk: The magnitude or rate of change exceeds the safety threshold by less than 30%.

[0077] High risk: The magnitude or rate of change exceeds the safety threshold by more than 30%.

[0078] Step S230: Extract the time-series characteristic data of building load from the system operation status feature set, including load intensity variation curves, load duration distribution, and load fluctuation frequency characteristics. Perform spatial gradient analysis on the soil temperature distribution data to obtain the temperature gradient distribution characteristics.

[0079] Among them, the building load time-series characteristic data reflects the characteristic data of building energy consumption changing over time, including load intensity variation curves, load duration distribution, and load fluctuation frequency characteristics. The load intensity variation curve describes the change of building load over time, reflecting the dynamic trend of load change.

[0080] Load duration distribution: Statistically analyzes the duration distribution of building load at different levels. Load fluctuation frequency characteristics: Analyzes the frequency and amplitude of building load fluctuations per unit time. Temperature gradient distribution characteristics: Characteristics obtained through spatial gradient analysis of soil temperature distribution data, reflecting the spatial variation of soil temperature.

[0081] The process is described below:

[0082] The system acquires real-time building load data from a pre-set sensor matrix, including indoor building temperature and the operating power of heat pump units. Using this data, the system calculates the building load intensity variation curve, load duration distribution, and load fluctuation frequency characteristics.

[0083] Load intensity variation curve: Through time series analysis, a curve showing the change of building load over time is plotted. The formula is as follows: ,in, It is the total load at time t. Let be the power of the i-th device at time t, and N be the total number of devices.

[0084] Load duration distribution: The duration at different load levels is statistically analyzed using the following formula: ,in, δ is the duration of load level L, T is the total time, and δ is the indicator function, which is 1 when L(t) = L and 0 otherwise.

[0085] Load fluctuation frequency characteristics: The formula for calculating the frequency and amplitude of load fluctuations per unit time is as follows: Where F is the load fluctuation frequency and ϵ is the preset fluctuation threshold.

[0086] Spatial gradient analysis of soil temperature distribution data: The system acquires soil temperature distribution data from temperature sensors embedded in the soil. Assume a total of M sensors, with each sensor measuring a temperature value of T. j (j=1,2,…,M).

[0087] Calculate the temperature gradient between adjacent sensors to obtain the temperature gradient distribution characteristics. The formula is:

[0088] ;

[0089] in, It is the temperature gradient between the j-th sensor and the (j+1)-th sensor. and Δz represents the temperature measured by the (j+1)th and j-th sensors, respectively, and Δz is the depth difference between the two sensors.

[0090] Step S240: The time-series characteristic data of building load and the simulation results of soil temperature field are fused to form a load-temperature coupled feature set.

[0091] The load-temperature coupling feature set combines time-series data of building load with soil temperature field simulation results to form a comprehensive feature set reflecting changes in both building load and soil temperature, used for multi-dimensional risk assessment. Data fusion integrates data from different sources to extract more comprehensive information for more accurate risk assessment.

[0092] The general process is described below:

[0093] 1. Data Preparation: The time-series characteristic data of building load obtained in step S230 includes load intensity variation curves, load duration distribution, and load fluctuation frequency characteristics. The soil temperature field simulation results obtained in step S220 include the amplitude and rate of change of soil temperature.

[0094] 2. Data Alignment: Align the time-series characteristics of building load with the soil temperature field simulation results on the same time series to ensure that they are compared and analyzed at the same point in time. For example, if both building load data and soil temperature data are recorded hourly, align them to the same point in time.

[0095] 3. Data Fusion: A weighted average method is used to combine building load and soil temperature data. The specific formula is as follows: ;in, Here, L(t) is the coupled characteristic value at time t, L(t) is the building load, and T(t) is the soil temperature. and These are weighting coefficients, adjusted according to actual needs. The selection of these weighting coefficients can be determined based on the degree of influence of load and temperature on system operation. For example, if the building load has a greater impact on system operation, then a weighting coefficient can be assigned. Higher weight.

[0096] 4. Feature set construction: The fused data is used to form a load-temperature coupled feature set, which contains the coupled feature values ​​at each time point.

[0097] Step S250: Perform a multi-dimensional risk assessment based on the load-temperature coupling feature set: Calculate the load intensity change rate based on the load intensity change curve, and compare the load intensity change rate with a preset load impact threshold to determine the load impact level; compare the load duration distribution with a preset cumulative effect threshold to determine the load accumulation level; calculate the non-uniformity coefficient based on the temperature gradient distribution characteristics, and compare it with a preset distribution uniformity threshold to determine the temperature distribution anomaly level.

[0098] The load impact level is determined by comparing the load intensity change rate with a preset load impact threshold, indicating the degree of impact of load changes on system operation. The load accumulation level is determined by comparing the load duration with a preset cumulative effect threshold, indicating the degree of impact of the load on the long-term operation of the system. The temperature distribution anomaly level is determined by comparing the temperature gradient distribution non-uniformity coefficient with a preset distribution uniformity threshold, indicating the degree of soil temperature distribution anomaly.

[0099] The system performs a multi-dimensional risk assessment based on the load-temperature coupling feature set. The specific process is as follows:

[0100] 1. Load impact level assessment:

[0101] Calculating the load intensity change rate: The system calculates the load intensity change rate based on the load intensity change curve. The formula is: Load intensity change rate = ,in, It is the change in load intensity. This refers to the time interval. The system compares the calculated load intensity change rate with a preset load impact threshold. Based on the magnitude of the change rate, the load impact level is divided into three levels: low, medium, and high. For example, if the change rate is less than 30% of the threshold, it is determined to be a low impact; if the change rate is between 30% and 60% of the threshold, it is determined to be a medium impact; and if the change rate exceeds 60% of the threshold, it is determined to be a high impact.

[0102] 2. Load accumulation level assessment:

[0103] Statistical load duration: The system calculates the duration of load at different levels based on the load duration distribution. The formula is: ;in, δ is the duration of load level L, T is the total time, and δ is the indicator function, which is 1 when L(t) = L and 0 otherwise.

[0104] The system compares the statistically obtained load duration with a preset cumulative threshold. Based on the duration, the load accumulation level is divided into three levels: low, medium, and high. For example, if the duration is less than 30% of the threshold, it is judged as low accumulation; if the duration is between 30% and 60% of the threshold, it is judged as medium accumulation; and if the duration exceeds 60% of the threshold, it is judged as high accumulation.

[0105] 3. Assessment of Temperature Distribution Anomaly Level:

[0106] Calculating the non-uniformity coefficient: Based on the characteristics of the temperature gradient distribution, the system calculates the non-uniformity coefficient of the temperature gradient distribution. The formula is: Non-uniformity coefficient = ;in, It is the standard deviation of the temperature gradient. It is the average value of the temperature gradient.

[0107] Comparison of distribution uniformity threshold: The system compares the calculated non-uniformity coefficient with a preset distribution uniformity threshold. Based on the magnitude of the non-uniformity coefficient, the temperature distribution anomaly level is divided into three levels: low, medium, and high. For example, if the non-uniformity coefficient is less than 30% of the threshold, it is judged as a low anomaly; if the non-uniformity coefficient is between 30% and 60% of the threshold, it is judged as a medium anomaly; and if the non-uniformity coefficient exceeds 60% of the threshold, it is judged as a high anomaly.

[0108] Step S260: Based on the above-mentioned risk assessment results, the weight coefficients of each risk dimension are determined using a preset risk weight allocation table, and the comprehensive risk value is calculated using a weighted algorithm. The risk dimensions include load shock, load accumulation, and abnormal temperature distribution.

[0109] Among them, the comprehensive risk value is a comprehensive indicator reflecting the risk of soil thermal imbalance, calculated through a weighted algorithm. The risk weight allocation table is a pre-set table used to determine the weight coefficients of each risk dimension (such as load shock, load accumulation, and abnormal temperature distribution).

[0110] The acquisition method and process are described below:

[0111] In step S260, based on the load impact level, load accumulation level, and temperature distribution anomaly level assessed in step S250, the system determines the weight coefficients for each risk dimension using a preset risk weight allocation table, and calculates the comprehensive risk value using a weighted algorithm. The specific process is as follows:

[0112] 1. Determine weighting coefficients: The system assigns weighting coefficients to each risk factor according to a preset risk weighting allocation table. For example, the weighting coefficient for load shock level is 0.4, the weighting coefficient for load accumulation level is 0.3, and the weighting coefficient for abnormal temperature distribution level is 0.3.

[0113] 2. Calculate the comprehensive risk value: The system multiplies the level value of each risk dimension by its corresponding weight coefficient, and then adds the results to obtain the comprehensive risk value. For example: Comprehensive risk value = (load impact level value × 0.4) + (load accumulation level value × 0.3) + (temperature distribution anomaly level value × 0.3).

[0114] Assuming a load shock level of 2, a load accumulation level of 1, and a temperature distribution anomaly level of 3, the comprehensive risk value is calculated as follows:

[0115] Overall risk value = (2×0.4)+(1×0.3)+(3×0.3)=0.8+0.3+0.9=2.0.

[0116] Step S270: Determine the final soil thermal imbalance risk level by comparing the comprehensive risk value with the preset risk level threshold range.

[0117] Among them, the risk level threshold range is a preset threshold range used to map the comprehensive risk value to a specific soil thermal imbalance risk level.

[0118] The specific process is as follows:

[0119] 1. Preset Risk Level Threshold Range: The system presets the mapping relationship between the comprehensive risk value and the risk level, which is usually divided into three levels: low risk, medium risk, and high risk. For example: 1. Low risk: comprehensive risk value between 0 and 1; 2. Medium risk: comprehensive risk value between 1 and 2; 3. High risk: comprehensive risk value greater than 2.

[0120] 2. Risk Level Determination: The system compares the calculated overall risk value with a preset threshold range to determine the final risk level. For example, if the overall risk value is 2.7, it is determined to be high risk based on the preset threshold range.

[0121] Based on the determined operating mode and the assessed soil thermal imbalance risk level, the corresponding target control parameter set is retrieved from the preset control parameter library to generate control commands, including:

[0122] Step S310: Based on the determined operating mode and the assessed soil thermal imbalance risk level, the corresponding target control parameter set is called from the preset control parameter library.

[0123] Among them, the target control parameter set is a specific set of control parameters called from the control parameter library according to the current operating mode and risk level.

[0124] In step S310, the system retrieves the corresponding target control parameter set from the preset control parameter library based on the determined operating mode (heating or cooling) and the assessed soil thermal imbalance risk level (low, medium, high). The specific process is as follows:

[0125] 1. Operation Mode and Risk Level Determination: The system obtains the current operation mode and soil thermal imbalance risk level from previous steps. For example, the system may determine that the current mode is heating and the risk level is medium risk.

[0126] 2. Target Control Parameter Set Invocation: The system selects the corresponding target control parameter set from the control parameter library based on the operating mode and risk level. The control parameter library is pre-defined based on historical data and simulation analysis, and contains the optimal control parameters for different operating modes and risk levels.

[0127] Step S320: If the soil thermal imbalance risk level is only assessed based on soil temperature data, then a fuzzy control algorithm is used to generate control commands based on the temperature setpoint and control threshold in the target control parameter set.

[0128] Fuzzy control algorithm: A control method based on fuzzy logic, used to handle uncertainty and fuzziness, and suitable for complex system control.

[0129] The specific process of step S320 can be referred to steps S321 to S325, and will not be repeated here.

[0130] Step S330: If the soil thermal imbalance risk level is obtained by jointly assessing the building load and soil temperature distribution data, then based on the target control parameter set called, the predictive model control algorithm and the fuzzy control algorithm are used to collaboratively generate control commands.

[0131] Among them, Predictive Model Control (MPC) is a control method based on a system model that predicts future system behavior and optimizes control actions to achieve long-term performance goals. Cooperative control combines the advantages of multiple control algorithms to improve the system's control performance and adaptability.

[0132] The specific process of step S330 can be referred to steps S331 to S335, and will not be repeated here.

[0133] Step S340: Perform a safety check on the generated control commands to ensure that each parameter is within the safety boundary range defined by the target control parameter set.

[0134] Among these, security verification involves checking the generated control commands to ensure they remain within safety boundaries and prevent potential damage to the system. Safety boundaries are preset parameter ranges that ensure the system operates in a safe and stable state.

[0135] In step S340, the system performs a safety check on the generated control commands to ensure that each parameter is within the safety boundary range defined by the target control parameter set. The specific process is as follows:

[0136] 1. Obtain control commands: The system obtains the generated control commands from step S320 or S330. These commands include heat pump power, circulating water pump flow rate, heat pipe start / stop status, etc.

[0137] 2. Safety Boundary Check: The system checks whether each control command is within the allowable range based on the safety boundaries in the target control parameter set. For example, whether the heat pump power is between the minimum and maximum power, and whether the circulating water pump flow rate is within the design flow rate range.

[0138] 3. Adjust control commands: If any control command exceeds the safety limits, the system will automatically adjust the command to bring it back within the safe range. For example, if the heat pump power exceeds the maximum value, the system will adjust the power to the maximum allowable value.

[0139] 4. Confirm safety verification passed: The adjusted control commands undergo another safety verification to ensure all parameters are within safety boundaries. Only control commands that pass the safety verification will be sent to the actuator.

[0140] Step S350: The control command that passes the security check is used as the final generated control command.

[0141] Based on the temperature setpoint and control threshold in the target control parameter set, a fuzzy control algorithm is used to generate control commands, including:

[0142] Step S321: Extract real-time input variables from the system operating status feature set, including soil temperature change rate and heat pump unit operating status data.

[0143] The general process is described below:

[0144] 1. Extracting Soil Temperature Change Rate: The system acquires soil temperature data from a pre-set sensor matrix and calculates the soil temperature change rate. The soil temperature change rate reflects the change in soil temperature over time and is an important indicator for assessing the risk of soil thermal imbalance.

[0145] 2. Extracting heat pump unit operating status data: The system obtains heat pump operating status data from the heat pump unit's operation monitoring system, including heat pump operating power, operating time, etc.

[0146] Step S322: Based on the temperature setpoint and control threshold in the target control parameter set, perform standardized preprocessing on the input variables.

[0147] The general process is as follows:

[0148] 1. Obtain the target control parameter set: The system extracts the temperature setpoint and control threshold from the target control parameter set. These parameters are preset based on the current operating mode and the soil thermal imbalance risk level.

[0149] 2. Standardize input variables: Convert soil temperature change rate and heat pump unit operating status data into standardized values. For example, the following formula can be used for standardization: Standardized value = (Actual value - Setpoint) / Threshold.

[0150] For the operating power of a heat pump, the following formula can be used: Standardized power = (Actual power − Minimum power) / (Maximum power − Minimum power).

[0151] Step S323: Based on the membership function parameters stored in the target control parameters centrally, establish the mapping relationship between input variables and fuzzy sets, and convert the preprocessed input variables into corresponding fuzzy sets.

[0152] In fuzzy logic, a fuzzy set represents the degree to which an element belongs to a certain set, rather than the "belongs to" or "does not belong to" in traditional binary logic. The membership function defines the degree to which an element belongs to a fuzzy set; it is typically a value between 0 and 1.

[0153] In step S323, the system converts the standardized input variables (such as soil temperature change rate and heat pump unit operating status data) into corresponding fuzzy sets based on the membership function parameters in the target control parameter set. The specific process is as follows:

[0154] 1. Obtaining Membership Function Parameters: The system extracts membership function parameters from the target control parameter set. These parameters define how input variables are mapped to fuzzy sets, such as the "low," "medium," and "high" fuzzy sets for the rate of temperature change.

[0155] 2. Mapping Input Variables to Fuzzy Sets: Use membership functions to convert standardized input variables into fuzzy sets. For example, for the rate of change of soil temperature, three fuzzy sets can be defined: "low", "medium", and "high", and membership functions can be used to calculate the degree to which the input variable belongs to each fuzzy set.

[0156] Step S324: Based on the established fuzzy set, apply the preset fuzzy rule base to perform inference calculations and generate a fuzzy set of output variables. The preset fuzzy rule base contains the output rules corresponding to all combinations of the fuzzy sets of input variables.

[0157] The system comprises: a fuzzy rule base (a set of preset fuzzy rules defining the output rules for all combinations of fuzzy sets of input variables); fuzzy inference (a reasoning process based on fuzzy logic that maps the fuzzy sets of input variables to the fuzzy sets of output variables using the fuzzy rule base); and output variables (the result of fuzzy inference, representing the control actions the system needs to take, such as adjusting the heat pump power or the circulating water pump flow rate).

[0158] In step S324, the system performs inference calculations based on the established fuzzy set and a preset fuzzy rule base to generate a fuzzy set of output variables. The specific process is as follows:

[0159] 1. Obtaining the Fuzzy Rule Base: The system extracts a pre-set fuzzy rule base from the target control parameter set. These rule bases are pre-defined based on the system's operating experience and historical data, defining the output rules corresponding to all combinations of the fuzzy sets of input variables.

[0160] 2. Fuzzy Inference Calculation: The system performs inference calculations based on the fuzzy sets of input variables using a fuzzy rule base. For each combination of fuzzy sets of input variables, the system finds the corresponding output rule and calculates the fuzzy set of the output variable. For example, if the fuzzy sets of input variables are "low temperature change rate" and "high heat pump operating power", then according to the preset fuzzy rule "if the temperature change rate is low and the heat pump operating power is high, then output reduce heat pump power", the system generates the fuzzy set of output variable "reduce heat pump power".

[0161] Step S325: The centroid method is used to defuzzify the fuzzy set of the output variables, converting the fuzzy quantities into precise control quantity values.

[0162] Defuzzification involves converting the fuzzy set obtained from fuzzy inference into specific numerical values ​​to execute specific control actions. The centroid method is a commonly used defuzzification method that determines the specific output value by calculating the centroid position of the fuzzy set.

[0163] In step S325, the system uses the centroid method to defuzzify the fuzzy set of the output variables, converting the fuzzy quantities into precise control values. The specific process is as follows:

[0164] 1. Obtain the fuzzy set of output variables: The system obtains the fuzzy set of output variables from step S324. These fuzzy sets are obtained based on fuzzy rule base reasoning. For example, output variables may include "reduce heat pump power" and "maintain heat pump power", and each output variable has a corresponding membership value.

[0165] 2. Defuzzification using the centroid method: The centroid method determines the specific output value by calculating the centroid position of the fuzzy set. The specific formula is: Output value = ∑(Membership degree × Output value) / ∑Membership degree.

[0166] This formula calculates the weighted average of the fuzzy sets, where the membership degree serves as the weight and the output value is the specific numerical value. For example, if the membership degree of "reduce heat pump power" is 0.9, the corresponding output value is -10kW; and the membership degree of "maintain heat pump power" is 0.1, the corresponding output value is 0kW. The calculation result is: Output value = [(0.9 × −10) + (0.1 × 0)] / (0.9 + 0.1) = −9kW. This means the system needs to reduce the heat pump power by 9kW.

[0167] The generation of control commands using a combination of predictive model control algorithms and fuzzy control algorithms includes:

[0168] Step S331: Extract raw input data from the system operating status feature set, and perform standardized preprocessing on the input data based on the target control parameter set to form a collaborative algorithm input dataset. The raw input data includes operating data and environmental data.

[0169] In step S331, the system extracts raw input data from the system operating state feature set, including operating data and environmental data, and performs standardized preprocessing on this data based on the target control parameter set to form the collaborative algorithm input dataset. The specific process is as follows:

[0170] 1. Extracting raw input data: The system extracts operational and environmental data from the operational status feature set. Operational data includes the operating power of the heat pump unit, the flow rate of the circulating water pump, etc.; environmental data includes soil temperature, outdoor meteorological parameters, etc.

[0171] 2. Standardization Preprocessing: The system standardizes the extracted raw input data based on parameters from the target control parameter set. For example, the heat pump operating power is standardized to a range of 0 to 1, and the soil temperature change rate is standardized to a preset threshold range.

[0172] Step S332: The MPC algorithm is used to process the input dataset of the collaborative algorithm. The prediction time domain, constraint conditions and system thermal characteristic parameters stored in the target control parameter centrally are called to predict the building load change trend and soil temperature field evolution within the future preset time period. A system dynamic response model is established, and a baseline operation strategy is generated based on the model using a rolling time domain optimization strategy. The baseline operation strategy includes the optimal operating point of the heat pump unit, the start-up and shutdown sequence of the heat pipe array and the set values ​​of system operation parameters.

[0173] The MPC algorithm (Model Predictive Control) is a model-based control algorithm that achieves long-term performance goals by predicting future system behavior and optimizing control actions. Prediction time domain: The time range used in the MPC algorithm to predict the future behavior of the system. Constraints: The restrictions that the system must meet during operation, such as the operating range of the equipment and safety boundaries. System thermal characteristic parameters: Parameters describing the thermal behavior of the system, such as heat capacity and thermal resistance. Baseline operating strategy: The optimized control strategy generated by the MPC algorithm, including the optimal operating point of the equipment and the setpoints of the operating parameters.

[0174] In step S332, the system uses the MPC algorithm to process the input dataset of the collaborative algorithm, calls the prediction time domain, constraint conditions, and system thermal characteristic parameters stored in the target control parameter central storage, predicts the building load change trend and soil temperature field evolution within a preset time period, establishes a system dynamic response model, and generates a baseline operating strategy based on this model using a rolling time domain optimization strategy. The specific process is as follows:

[0175] 1. Obtain prediction time domain, constraints and system thermal characteristic parameters: The system extracts prediction time domain (e.g., 24 hours), constraints (e.g., heat pump power range 50-70kW) and system thermal characteristic parameters (e.g., heat capacity, thermal resistance) from the target control parameter set.

[0176] 2. Establish a system dynamic response model: Using the extracted parameters and the input dataset of the collaborative algorithm, a system dynamic response model is established. This model can predict the trend of building load changes and the evolution of the soil temperature field within a preset time period. For example, the model may predict that the soil temperature will rise during high-load periods, requiring advance adjustment of the heat pump power.

[0177] 3. Predicting Future System Behavior: Using the established dynamic response model, predict the trend of building load changes and the evolution of soil temperature field within a preset time period. For example, the model predicts that in the next 24 hours, the building load will reach 70kW during the midday peak, and the soil temperature may rise by 2℃.

[0178] 4. Rolling Time-Domain Optimization Strategy: Based on the prediction results, a baseline operating strategy is generated using a rolling time-domain optimization strategy. This strategy includes the optimal operating point of the heat pump unit (e.g., increasing power to 65kW in advance during high-load periods), the start-up and shutdown sequence of the heat pipe array (e.g., shutting down some heat pipes during low-load periods), and the setpoints of system operating parameters (e.g., adjusting the circulating water pump flow rate to 12m³ / h). The rolling time-domain optimization strategy continuously updates predictions and optimizes control actions to ensure that the system always operates in an optimal state under dynamically changing conditions.

[0179] Step S333: The real-time input data is processed using a fuzzy control algorithm. The membership function parameters and rule base thresholds in the target control parameter set are called to perform fuzzy inference and output dynamic adjustment instructions. The real-time input data is extracted from the system operation status feature set, including the current soil temperature change rate calculated from the soil temperature distribution time series data, and the heat pump unit load rate extracted from the real-time monitoring data of the heat pump unit operating power.

[0180] In step S333, the system uses a fuzzy control algorithm to process real-time input data, calls the membership function parameters and rule base thresholds from the target control parameter set, performs fuzzy inference, and outputs dynamic adjustment instructions. The specific process is as follows:

[0181] 1. Acquire real-time input data: The system extracts real-time input data from the operating status feature set, including the current soil temperature change rate calculated from the soil temperature distribution time series data, and the heat pump unit load rate extracted from the real-time monitoring data of the heat pump unit operating power.

[0182] 2. Invoking Membership Function Parameters and Rule Base Thresholds: The system extracts membership function parameters and rule base thresholds from the target control parameter set. These parameters define how input variables are mapped to fuzzy sets and the triggering conditions for fuzzy rules.

[0183] 3. Fuzzy Reasoning: The system converts real-time input data into fuzzy sets based on membership function parameters. For example, the soil temperature change rate may be mapped to the fuzzy sets "low," "medium," and "high," and the heat pump unit load rate may also be mapped to the fuzzy sets "low," "medium," and "high." The system applies fuzzy rules for reasoning based on rule base thresholds. For example, if the soil temperature change rate is "low" and the heat pump unit load rate is "high," the output will be "reduce heat pump power."

[0184] 4. Output dynamic adjustment commands: Based on the results of fuzzy inference, the system outputs dynamic adjustment commands. These commands include heat pump power adjustment and heat pipe start / stop commands, used to adjust the system's operating status in real time.

[0185] Step S334: The baseline operation strategy output by the MPC algorithm and the dynamic adjustment command output by the fuzzy control algorithm are integrated through a weight allocation mechanism. The weight allocation mechanism is dynamically adjusted based on the system operation mode, the soil thermal imbalance risk level, and the real-time stability index calculated from the system operation state feature set. The real-time stability index includes the load fluctuation rate calculated based on the building load information, the temperature gradient change rate calculated based on the soil temperature distribution, and the equipment operation state index calculated based on the operation data.

[0186] The system includes the following components: a weight allocation mechanism (a method for dynamically adjusting the weights of different control algorithm outputs to optimize the overall control performance of the system), a baseline operating strategy (an optimized control strategy generated by the MPC algorithm, including the optimal operating point of the equipment and the setpoints for operating parameters), dynamic adjustment commands (real-time control commands generated by the fuzzy control algorithm for rapid response to real-time changes in the system), and real-time stability indicators (indices calculated based on the system's operating state feature set, including load fluctuation rate, temperature gradient change rate, and equipment operating state index).

[0187] In step S334, the system fuses the baseline operating strategy output by the MPC algorithm and the dynamic adjustment instructions output by the fuzzy control algorithm through a weight allocation mechanism. The specific process is as follows:

[0188] 1. Obtaining the baseline operating strategy and dynamic adjustment instructions: The system obtains the baseline operating strategy generated by the MPC algorithm from step S332. For example, the MPC algorithm suggests setting the heat pump power to 65kW, turning on the heat pipe array during high-load periods, and adjusting the circulating water pump flow rate to 12m³ / h. The system obtains the dynamic adjustment instructions generated by the fuzzy control algorithm from step S333. For example, the fuzzy control algorithm suggests reducing the heat pump power by 5kW when the current soil temperature change rate is low and the heat pump load rate is high.

[0189] 2. Call real-time stability indicators: The system extracts real-time stability indicators from the set of operating status features. For example, the current load fluctuation rate is 10%, the temperature gradient change rate is 0.2℃ / h, and the equipment operating status index is 0.8 (close to 1 indicates that the equipment is operating well).

[0190] Load volatility: Calculated based on building load information, load volatility reflects the degree of load variation per unit time. The calculation formula is: Load volatility = Where L(t) is the load at time t, and T is the total time.

[0191] Temperature gradient change rate: The temperature gradient change rate calculated based on the soil temperature distribution reflects the degree of change of the temperature gradient per unit time.

[0192] The calculation formula is: Temperature gradient change rate = Where ∇T(t) is the temperature gradient at time t, and T is the total time.

[0193] Equipment Operating Status Index: Calculated based on operating data, this index reflects the operating status of the equipment. The calculation formula is: Equipment Operating Status Index = Where Pi is the actual power of the i-th device. N is the rated power of the equipment, and N is the total number of equipment.

[0194] 3. Weight Allocation: The system dynamically adjusts the weight allocation based on the system operation mode (e.g., high-load operation mode), soil thermal imbalance risk level (e.g., medium risk), and real-time stability indicators. For example, if the system is in a high-load operation mode and the soil thermal imbalance risk level is medium risk, the system may set the weight of the MPC algorithm to 0.7 and the weight of the fuzzy control algorithm to 0.3 to ensure long-term stability and rapid response to real-time changes.

[0195] 4. Integrated Control Commands: Based on the weight allocation results, the system integrates the baseline operating strategy and dynamic adjustment commands to generate the final coordinated control commands. For example, the integrated heat pump power is: Integrated heat pump power = (0.7 × 65) + (0.3 × (65 − 5)) = 45.5 + 18 = 63.5 kW. The system sets the heat pump power to 63.5 kW, the heat pipe array is turned on during high-load periods, and the circulating water pump flow rate is adjusted to 12 m³ / h.

[0196] Step S335: Generate collaborative control commands based on the fusion results, including the heat pump unit frequency setpoint, heat pipe start / stop command, and heat exchanger medium flow adjustment value.

[0197] The baseline operating strategy output by the MPC algorithm and the dynamic adjustment instructions output by the fuzzy control algorithm are integrated through a weight allocation mechanism, including:

[0198] Step S334.1: Based on historical operation mode data, timestamps and external meteorological conditions, identify the current operation scenario label through a preset scenario classification model, input the generated scenario label into a preset priority evaluation model, and output the priority score value of the scenario label.

[0199] The system comprises the following components: **Scenario Classification Model:** A pre-defined model used to identify the current operating scenario label based on historical operating mode data, timestamps, and external weather conditions. **Scenario Label:** A label describing the system's current operating state, such as "high load operation," "low load operation," or "large temperature fluctuation." **Priority Evaluation Model:** A pre-defined model used to evaluate the priority score of the scenario label to determine the control priority of different scenarios. **Priority Score:** A numerical value representing the priority of the scenario label, used for subsequent weight allocation.

[0200] The general process is described below:

[0201] 1. Acquisition of Historical Operation Mode Data, Timestamps, and External Meteorological Conditions: The system extracts operation mode data from historical data, including past load conditions and equipment operating status. For example, historical data shows that the building load typically reaches 70kW during the midday peak on weekdays. The system obtains the current timestamp to identify the current time period (e.g., weekday, weekend, holiday). For example, the current timestamp is 12:00 on a weekday. The system obtains current meteorological conditions from external weather stations, such as temperature, humidity, and wind speed. For example, the current external temperature is 30℃, humidity is 60%, and wind speed is 5m / s.

[0202] 2. Identify Current Operating Scenario Labels: The system uses a pre-defined scenario classification model, combined with the above data, to identify the current operating scenario labels. For example, the model may identify the current scenario as "high load operation" and "high temperature environment." The scenario classification model may be based on machine learning algorithms, such as decision trees or random forests, classifying the current operating status according to input features (such as timestamps, load data, and weather conditions).

[0203] 3. Priority Score Evaluation of Scene Tags: The system inputs the identified scene tags into a preset priority evaluation model. The model outputs a priority score for the scene tags based on preset rules. For example, if the current scene is "high load operation," the model might output a priority score of 80 (out of 100), indicating that control strategies for this scene should be prioritized; if the current scene is "high temperature environment," the model might output a priority score of 70. The priority evaluation model may be based on a rule engine, calculating priority scores according to preset rules (such as high load having higher priority than high temperature environment).

[0204] Step S334.2: Combine the priority score of the scene label with the real-time stability index, and generate a comprehensive scene weight through the preset comprehensive weight calculation rules.

[0205] Among them, the comprehensive scene weight is a weight calculated according to preset rules, combining the priority score of scene labels and real-time stability indicators, and is used to adjust the output of the MPC algorithm and fuzzy control algorithm. Real-time stability indicators are metrics reflecting the current operating status of the system, such as load fluctuation rate, temperature gradient change rate, and equipment operating status index.

[0206] The general process is described below:

[0207] 1. Obtain the priority score of the scene tag: The system obtains the priority score of the scene tag from step S334.1. For example, if the current scene is "high load operation", the priority score is 80; if the current scene is "high temperature environment", the priority score is 70.

[0208] 2. Obtain real-time stability indicators: The system extracts real-time stability indicators from the set of operating status features, such as load fluctuation rate, temperature gradient change rate, and equipment operating status index. For example, the current load fluctuation rate is 10%, the temperature gradient change rate is 0.2℃ / h, and the equipment operating status index is 0.8 (close to 1 indicates that the equipment is operating well).

[0209] 3. Preset Comprehensive Weight Calculation Rules: The system has a set of preset comprehensive weight calculation rules, which adjust the priority score based on real-time stability indicators. For example:

[0210] If the load volatility is higher than 15%, the weight is increased by 10%.

[0211] If the rate of change of the temperature gradient is higher than 0.3℃ / h, the weight is increased by 5%.

[0212] If the equipment operating status index is below 0.7, the weight is reduced by 5%.

[0213] 4. Calculate the overall scene weight: The system calculates the overall scene weight based on preset rules, priority scores, and real-time stability indicators. The specific calculation process is as follows:

[0214] Load volatility: The current load volatility is 10%, which is lower than 15%, so no weighting is added.

[0215] Temperature gradient change rate: The current temperature gradient change rate is 0.2℃ / h, which is lower than 0.3℃ / h, so no weight is added.

[0216] Equipment operating status index: The current equipment operating status index is 0.8, which is higher than 0.7, so the weight will not be reduced.

[0217] The formula for calculating the overall scene weight is: Overall scene weight = Priority score value × Adjustment coefficient.

[0218] For the "high load operation" scenario: the overall scenario weight = 80 × 1.0 = 80.

[0219] For the "high temperature environment" scenario: the overall scenario weight = 70 × 1.0 = 70.

[0220] Weight Adjustment: If real-time stability indicators show potential instability factors in the system, the system will adjust the weights according to preset rules. For example, if the load fluctuation rate exceeds 15%, the system will increase the weight of "high load operation" from 80 to 88 (an increase of 10%) to prioritize handling high load situations. If the temperature gradient change rate exceeds 0.3℃ / h, the system will increase the weight of "high temperature environment" from 70 to 73.5 (an increase of 5%) to prioritize handling temperature changes.

[0221] Step S334.3: Input the comprehensive scene weights into the preset weight mapping model to determine the initial weight ratio between the MPC algorithm and the fuzzy control algorithm.

[0222] Initial weight ratio: The initial weight ratio set when fusing the outputs of the MPC algorithm and the fuzzy control algorithm, used to balance the contributions of the two control algorithms. Weight mapping model: A preset model used to determine the initial weight ratio of the MPC algorithm and the fuzzy control algorithm based on the overall scene weights.

[0223] The general process is described below:

[0224] 1. Obtain the overall scenario weight: The system obtains the overall scenario weight from step S334.2. For example, for the "high load operation" scenario, the overall scenario weight is 80; for the "high temperature environment" scenario, the overall scenario weight is 70.

[0225] 2. Input Weight Mapping Model: The system inputs the comprehensive scene weights into a preset weight mapping model. This model determines the initial weight ratios of the MPC algorithm and the fuzzy control algorithm based on the comprehensive scene weights and preset rules. For example, the weight mapping model might preset the following rules:

[0226] If the overall scene weight is higher than 75, the weight of the MPC algorithm is 0.7, and the weight of the fuzzy control algorithm is 0.3.

[0227] If the overall scene weight is between 60 and 75, the weight of the MPC algorithm is 0.6, and the weight of the fuzzy control algorithm is 0.4.

[0228] If the overall scene weight is less than 60, the weight of the MPC algorithm is 0.5, and the weight of the fuzzy control algorithm is 0.5.

[0229] 3. Determine the initial weight ratio: Based on the overall scenario weights, the weight mapping model outputs the initial weight ratios for the MPC algorithm and the fuzzy control algorithm. For example:

[0230] For the "high load operation" scenario (the overall scenario weight is 80), the weight mapping model outputs a weight of 0.7 for the MPC algorithm and a weight of 0.3 for the fuzzy control algorithm.

[0231] For the "high temperature environment" scenario (with a comprehensive scenario weight of 70), the weight mapping model outputs a weight of 0.6 for the MPC algorithm and a weight of 0.4 for the fuzzy control algorithm.

[0232] Step S334.4: Based on the initial weight ratio and combined with the real-time system response data within the preset optimization period, the weight ratio is dynamically adjusted through a preset reinforcement learning algorithm to obtain the optimized weight ratio. The real-time system response data includes the operating status of the heat pump unit, the temperature of the heat exchanger medium, the start-up and shutdown status of the heat pipe, and the rate of change of soil temperature.

[0233] The algorithm includes: Reinforcement learning algorithm: a machine learning method that learns the optimal behavioral strategy through trial and error to maximize cumulative reward. Real-time system response data: data collected in real time during system operation, including the operating status of the heat pump unit, heat exchanger medium temperature, heat pipe start / stop status, and soil temperature change rate, used to evaluate the current system response. Optimized weight ratio: a weight ratio dynamically adjusted through the reinforcement learning algorithm, used to more accurately fuse the outputs of the MPC algorithm and the fuzzy control algorithm. Initial weight ratio: the initial weight ratio set when fusing the outputs of the MPC algorithm and the fuzzy control algorithm, used to balance the contributions of the two control algorithms.

[0234] The general process is as follows:

[0235] 1. Determine the optimization cycle and collect real-time system response data, as follows:

[0236] 1.1 Set a preset optimization cycle: Based on the system response sensitivity (such as the lag time of heat pump load adjustment and the soil temperature change cycle), preset the weight optimization cycle (e.g., 5-15 minutes / cycle) to ensure that the algorithm can capture changes in operating conditions in a timely manner and avoid frequent adjustments that cause system fluctuations.

[0237] 1.2 Real-time data acquisition and preprocessing: During each optimization cycle, the operating status of the heat pump unit (such as compressor frequency, load rate, COP value), heat exchanger medium temperature (temperature difference between inlet and outlet on the ground source side / load side, medium flow rate), heat pipe start-up and shutdown status (number of heat pipes started, heat exchange capacity of a single heat pipe) and soil temperature change rate (hourly change value of soil temperature at a depth of 2m / 4m around the pile) are continuously collected; outlier removal (such as removing out-of-range data caused by sensor failure) and normalization processing (mapping each indicator to the [0,1] interval) are performed on the collected data to form a standardized real-time system response dataset.

[0238] 2. Construct the state space and action space for reinforcement learning, as follows:

[0239] 2.1 Defining the State Space: The state features of reinforcement learning are defined as "real-time system response dataset + current initial weight ratio," specifically including:

[0240] System operating status characteristics: heat pump load rate (e.g., 0.3-1.0), heat exchanger medium temperature difference (e.g., 2-10℃), percentage of activated heat pipes (e.g., 0.2-1.0), and soil temperature change rate (e.g., -0.5-0.5℃ / h).

[0241] Weight state characteristics: current weight ratio of MPC algorithm (e.g., 0.3-0.7), weight ratio of fuzzy control algorithm (1-MPC weight);

[0242] 2.2 Define the action space: Take the "weight ratio adjustment range" as the algorithm action, set the adjustment step size (e.g., ±0.05 / time), and limit the action range to "MPC weight 0.3-0.7, fuzzy control weight 0.3-0.7". For example, the actions include "MPC weight +0.05, fuzzy control weight -0.05", "MPC weight -0.05, fuzzy control weight +0.05", and "weights remain unchanged".

[0243] 3. Design the reward function for reinforcement learning, as follows:

[0244] To guide the algorithm to adjust weights toward the goals of "optimal system energy efficiency, stable operation, and soil thermal balance", a multi-dimensional reward function is preset. The formula for calculating the total reward value R is as follows: R = α × R1 + β × R2 + γ × R3, where α, β, and γ are preset weight coefficients (e.g., α = 0.4, β = 0.3, γ = 0.3), ensuring coordinated optimization of each goal.

[0245] R1 (Energy Efficiency Bonus): Based on the heat pump COP value, if COP ≥ preset high efficiency threshold (e.g., 4.0), then R1 = 1.0. For every 0.1 below the threshold, R1 decreases by 0.1 (minimum -0.5). The incentive algorithm prioritizes high energy efficiency.

[0246] R2 (Stability Bonus): Based on load fluctuation rate (≤10% is considered stable) and heat exchanger medium temperature difference fluctuation rate (≤5% is considered stable). If both are stable, R2 = 0.8. If either exceeds the threshold, R2 decreases by 0.2 (minimum -0.3) to avoid system fluctuations caused by weight adjustments.

[0247] R3 (Soil Thermal Balance Bonus): Based on the rate of change in soil temperature, if the rate of change is within ±0.2℃ / h (no risk of thermal imbalance), then R3=0.8; if it exceeds the range, R3 decreases by 0.3 (minimum is -0.4) to prevent long-term operation from causing soil temperature imbalance.

[0248] 4. Implement the weight adjustment strategy and system feedback evaluation, as detailed below:

[0249] 4.1 Initial policy execution: The "initial weight ratio" obtained in step S334.3 is used as the initial action of reinforcement learning to drive the MPC algorithm to output the baseline policy and the fuzzy control algorithm to output dynamic instructions. After fusion according to the initial weights, control instructions (such as adjusting the heat pump frequency and the number of heat pipes to start and stop) are generated.

[0250] 4.2 System Feedback Observation: During the current optimization cycle, the system performance after the execution of the fusion command is observed by collecting "real-time system response data" (such as whether COP increases, whether load fluctuation decreases, and whether soil temperature changes stabilize), and the total reward value R for the current cycle is calculated based on the above reward function.

[0251] 5. Update the weight adjustment strategy based on the reward value. Use a preset reinforcement learning algorithm (such as Q-learning algorithm) to update the strategy, as follows:

[0252] 5.1 Strategy Update Rules: If the current total reward value R ≥ the preset reward threshold (e.g., 0.5), it means that the current weight ratio is suitable for the working conditions. The weight will be retained or only slightly adjusted (e.g., adjustment step size ±0.02) in the next cycle. If R < the reward threshold, the Q table will be updated according to the "status-action-reward" data. Actions that can increase the reward value will be selected first (e.g., if the current R is low due to large load fluctuations, the action priority of the fuzzy control weight will be increased to take advantage of its anti-fluctuation advantage).

[0253] 5.2 Iterative optimization: Repeat the process of "strategy execution → feedback evaluation → strategy update" until the total reward value R of a certain optimization cycle is greater than or equal to the reward threshold for 3 consecutive times, and the system response indicators (COP, load fluctuation rate, soil temperature change rate) are all stable within the preset reasonable range. Stop the iteration and output the current weight ratio as the "optimized weight ratio".

[0254] 6. Output the optimization results and closed-loop application, as detailed below:

[0255] The optimized weight ratio is fed back to the "MPC-Fuzzy Control Fusion Module" in real time for the algorithm output fusion in the next optimization cycle; at the same time, the "optimized weight ratio + corresponding real-time system response data + reward value" is stored in the historical database to provide data support for the parameter iteration of subsequent scene classification models and priority evaluation models, forming a closed loop of "weight optimization - data accumulation - model iteration".

[0256] Step S334.5: Using the optimized weight ratio, the baseline running strategy output by the MPC algorithm and the dynamic adjustment instructions output by the fuzzy control algorithm are integrated through a preset weighted fusion algorithm.

[0257] Based on the optimized weight ratio obtained in step S334.4 (the proportion of MPC algorithm and fuzzy control algorithm has been determined, such as MPC accounting for 60% and fuzzy control accounting for 40%, MPC accounting for 30% and fuzzy control accounting for 70%), the output is integrated through a preset weighted fusion algorithm, as follows:

[0258] 1. Extract core parameters: Take the MPC's baseline operating strategy, including the compressor's baseline frequency (e.g., 45Hz), the heat exchanger's medium baseline flow rate (e.g., 70m³ / h), and the heat pipe's baseline start-up count (e.g., 3 units); take the fuzzy control's dynamic adjustment commands, including the compressor frequency fine-tuning amount (e.g., +4Hz), the medium flow rate correction amount (e.g., +8m³ / h), and the heat pipe start / stop adjustment number (e.g., +1 unit).

[0259] 2. Integration according to weighting rules: If the MPC weight is ≥50%, then the fuzzy control adjustment is added proportionally based on its benchmark parameter (e.g., when the weight is 60%, the compressor frequency mainly refers to 45Hz, and only 40% of the +4Hz correction is included to prioritize stability); if the MPC weight is <50%, then the fuzzy control adjustment is strengthened (e.g., when the weight is 30%, the compressor frequency mainly refers to the +4Hz correction to quickly adapt to the operating conditions).

[0260] 3. Parameter verification: Strictly check whether the parameters after fusion are within the safe range of the equipment (such as compressor frequency 25-60Hz, medium flow rate 30-100m³ / h). If they exceed the range, automatically correct them to the critical value (e.g., if the calculated frequency is 62Hz, then take 60Hz).

[0261] 4. Issue commands: Integrate the verified parameters (such as compressor 47.4Hz, flow rate 73.2m³ / h, 3 heat pipes) into commands and send them to the compressor inverter, circulating pump driver and heat pipe controller to achieve coordinated operation.

[0262] Considering that the heat pipe can be directly coupled to the building floor structure, i.e., the evaporation section of the heat pipe is buried underground and in contact with the soil around the pile or the concrete of the pile foundation, and the condensation section extends into the interior of the building floor or is close to the bottom surface of the floor to form direct thermal coupling, and adjustable-angle heat-conducting fins are installed on the outer wall of the heat pipe, the operation and management method of a heat pipe ground source heat pump system also includes the following steps:

[0263] Step SA00 involves real-time acquisition of system operation data via a sensor matrix. This data includes: equipment operating parameters (temperature of the heat pipe evaporation and condensation sections, heat pipe operating pressure, inlet and outlet temperatures of the heat exchanger medium, operating power and load rate of the heat pump unit), structural status parameters (floor surface temperature, temperature gradient between the heat pipe and the floor), and environmental data (soil moisture, indoor building temperature, building load information, outdoor meteorological parameters, and soil temperature distribution). The acquisition process employs a preset sampling frequency and preset accuracy.

[0264] Step SB00: Process the data collected in step SA00: Remove out-of-range data caused by sensor malfunctions (such as temperature > 80℃ or < -10℃, pressure > 2MPa); integrate equipment, structural and environmental data using timestamp synchronization (accuracy ≤ 1s); standardize each parameter (mapped to the [0,1] interval) to generate a standardized dataset for subsequent analysis.

[0265] Step SC00, based on a standardized dataset, uses an MPC prediction model to integrate system characteristic parameters such as the thermal conductivity of the heat pipe-floor slab and the coefficient of linear expansion of the material to conduct a quantitative risk assessment.

[0266] Winter risk assessment: Set thermal stress thresholds (low risk ≤1.5MPa, medium risk 1.5-2.0MPa, high risk >2.0MPa) and expansion difference thresholds (low risk ≤0.1mm / m, medium risk 0.1-0.2mm / m, high risk >0.2mm / m).

[0267] When the temperature difference gradient between the heat pipe and the floor slab exceeds the preset threshold (e.g., 3℃ / cm), the thermal stress of the floor slab concrete and the expansion difference between the heat pipe and the floor slab are calculated in combination with the heat pipe operating pressure. The comprehensive risk value (0-3 points) is calculated according to "thermal stress level × 0.6 + expansion difference level × 0.4", corresponding to low risk (0-1 point), medium risk (1-2 points), and high risk (2-3 points).

[0268] Summer risk assessment: Dew point temperature is calculated based on indoor temperature and humidity of the building, and thresholds for condensation thermal resistance increment are set (low risk ≤0.03(m²·K) / W, medium risk 0.03-0.05(m²·K) / W, high risk >0.05(m²·K) / W); when the floor surface temperature is lower than the dew point and continues for more than the preset duration (low risk 5-10min, medium risk 10-15min, high risk >15min), the actual thermal resistance increment is calculated in combination with soil moisture (the thermal resistance increment coefficient increases by 0.1 for every 10% RH increase in humidity), and low, medium and high risk levels are correspondingly classified.

[0269] Step SD00 generates differentiated control instructions based on the risk level results of step SC00, as follows:

[0270] Winter control: Under low risk conditions, the heat pump unit power adjustment command is generated based on the inlet and outlet temperatures of the heat exchanger medium (preheating rate 1.5-2℃ / h), and the opening angle of the heat-conducting fins is controlled to 15°-20° in conjunction with the soil temperature distribution; under medium risk conditions, the preheating rate is reduced to 1-1.5℃ / h, and the fin angle is adjusted to 20°-25°; under high risk conditions, the preheating rate is ≤1℃ / h, the fin angle is adjusted to 25°-30°, and the upper limit of the heat pump unit load rate is triggered (≤80%).

[0271] Summer control: Under low risk, generate a fin half-open command (angle 60°-90°), and adjust the ventilation fan speed (800-1000r / min) in conjunction with outdoor meteorological parameters; under medium risk, the fin opening angle is 90°-120°, and the fan speed is 1000-1200r / min; under high risk, the fins are fully open (180°), and the fan speed is 1200-1500r / min, ensuring that the floor surface temperature is 3-5℃ higher than the dew point temperature.

[0272] Step SE00 involves refining the control commands of step SD00 using a fuzzy control system.

[0273] Fin angle optimization: Based on the real-time temperature difference between the heat pipe and the floor (for every 1° increase from the preset value, the fin angle increases by different increments according to the risk level: 5° for low risk, 6° for medium risk, and 8° for high risk) and the operating pressure deviation of the heat pipe (when it exceeds the rated value by ±5%, the adjustment range increases by 20%), the fin servo motor is driven to dynamically correct the angle.

[0274] Ventilation and heat pump parameter optimization: Combining the building's real-time load (adjusting fan speed / heat pump power when load fluctuation exceeds 10%) and soil moisture (enhancing ventilation intensity when >70%RH), the ventilation fan frequency converter and heat pump unit controller are driven to fine-tune parameters, forming optimized instructions adapted to real-time operating conditions.

[0275] In step SF00, the optimized control command from step SE00 is sent to the corresponding actuator (heat pump unit controller, fin drive motor, and ventilation fan frequency converter).

[0276] The embodiments described in this specific implementation are preferred embodiments of this application and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A method for operating and managing a heat-pipe ground-source heat pump system, characterized by, The heat pipe ground source heat pump system comprises: a pile foundation; at least one heat pipe combined with the pile foundation, the heat pipe being a closed pipe body with a capillary core structure and internally evacuated and filled with a phase change working medium; the heat pipe comprising an evaporation section at a lower portion and / or a side portion of the pile foundation and a condensation section at an upper portion of the pile foundation; a heat exchanger in fluid communication with the condensation section of the heat pipe through a sealed pipeline; a heat pump unit in fluid communication with the heat exchanger at a ground source side; wherein the evaporation section of the heat pipe is used for heat exchange with soil around the pile, and the condensation section of the heat pipe transmits heat to or releases heat from the heat pump unit to the soil through the heat exchanger; the capillary core structure is configured to drive the phase change working medium in liquid state to autonomously return between the evaporation section and the condensation section by capillary force; the operation management method comprises: real-time collection of operation data and environmental data of the heat pipe ground source heat pump system through a preset sensor matrix, data preprocessing, and formation of a system operation state feature set, wherein the operation data comprises heat pipe operation pressure and temperature, inlet and outlet temperatures of a medium in the heat exchanger, operation power and load rate of the heat pump unit, the environmental data comprises building indoor temperature, building load information, outdoor meteorological parameters, and soil temperature distribution; comparison of the building indoor temperature and the outdoor meteorological parameters in the system operation state feature set with preset mode determination thresholds to determine a current operation mode of the system, the operation mode comprising a heating mode and a cooling mode; meanwhile, based on time series data of the soil temperature distribution, a soil temperature field change in a future set period is simulated through a heat conduction model, and according to a rising and falling amplitude and a rate of the soil temperature in the simulation result, comparison is made with preset safety thresholds, and according to a degree of exceeding the thresholds, a risk level of soil heat imbalance is quantified; according to the determined operation mode and the evaluated risk level of soil heat imbalance, a corresponding target control parameter set is called from a preset control parameter library to generate a control instruction; the generated control instruction is sent to corresponding execution mechanisms to drive the heat pump unit, a device for adjusting medium flow of the heat exchanger, and the heat pipe array to operate according to the instruction.

2. The operation management method of a heat pipe ground source heat pump system according to claim 1, characterized by, quantifying the risk level of soil heat imbalance comprises: based on the soil temperature distribution data, a current soil heat state characteristic value is calculated by using a preset soil heat state characteristic value calculation method, and comparison is made with a preset primary warning threshold to determine whether there is a risk of soil heat imbalance; if not, comparison is made between a rising and falling amplitude and a rate of the soil temperature in the simulation result and preset safety thresholds, and according to a degree of exceeding the thresholds, a risk level of soil heat imbalance in a future set period is quantified; if yes, time series characteristic data of building load, including a load intensity change curve, a load duration distribution, and a load fluctuation frequency feature, are extracted from the system operation state feature set; spatial gradient analysis is performed on the soil temperature distribution data to obtain a temperature gradient distribution feature; data fusion is performed on the building load time series characteristic data and the soil temperature field simulation result to form a load-temperature coupling feature set; Perform multi-dimensional risk assessment based on load-temperature coupling characteristics: based on the load intensity change curve, calculate the load intensity change rate, and compare the load intensity change rate with the preset load impact threshold to determine the load impact level; based on the load duration distribution, compare the load duration with the preset cumulative effect threshold to determine the load cumulative level; based on the temperature gradient distribution characteristics, calculate the non-uniformity coefficient of the temperature gradient distribution characteristics, and compare the non-uniformity coefficient with the preset distribution uniformity threshold to determine the temperature distribution abnormality level; Based on the above level evaluation results, use the preset risk weight distribution table to determine the weight coefficient of each risk dimension, and calculate the comprehensive risk value through the weighting algorithm, the risk dimensions include load impact, load accumulation and temperature distribution abnormality; According to the comparison between the comprehensive risk value and the preset risk level threshold interval, the final soil thermal imbalance risk level is determined.

3. The operation management method of a heat pipe ground source heat pump system according to claim 2, characterized by, According to the determined operation mode and the evaluated soil thermal imbalance risk level, the corresponding target control parameter set is called from the preset control parameter library to generate control instructions, including: Based on the determined operation mode and the evaluated soil thermal imbalance risk level, the corresponding target control parameter set is called from the preset control parameter library; If the soil thermal imbalance risk level is evaluated based on only the soil temperature distribution data, then based on the temperature set value and the control threshold in the called target control parameter set, a fuzzy control algorithm is used to generate control instructions; If the soil thermal imbalance risk level is evaluated based on both building load and soil temperature distribution data, then based on the called target control parameter set, a predictive model control algorithm and a fuzzy control algorithm are used to generate control instructions cooperatively; Safety check is performed on the generated control instructions to ensure that each parameter is within the safety boundary range defined in the target control parameter set; The control instructions that pass the safety check are used as the final generated control instructions.

4. The operation management method of a heat pipe ground source heat pump system according to claim 3, characterized by, Based on the temperature set value and the control threshold in the called target control parameter set, a fuzzy control algorithm is used to generate control instructions, including: Extract real-time input variables from the system operation state characteristics set, including soil temperature change rate and heat pump unit operation state data; Based on the temperature set value and the control threshold in the target control parameter set, standardize the input variables for preprocessing; Based on the membership function parameters stored in the target control parameter set, establish the mapping relationship between the input variables and the fuzzy sets, and convert the preprocessed input variables to the corresponding fuzzy sets; Based on the established fuzzy sets, apply the preset fuzzy rule library for inference calculation to generate the fuzzy set of output variables, and the preset fuzzy rule library includes the output rules corresponding to all combinations of input variable fuzzy sets; Use the barycentric method to de-fuzzify the fuzzy set of output variables, and convert the fuzzy quantity to an accurate control quantity value.

5. The method of operating and managing a heat pipe ground source heat pump system according to claim 3, wherein, The predictive model control algorithm and the fuzzy control algorithm are used to generate control instructions cooperatively, including: Extract the original input data from the system operation state characteristics set, and standardize the input data based on the target control parameter set to form a cooperative algorithm input data set, the original input data includes operation data and environment data; The MPC algorithm is used to process the collaborative algorithm input data set, and the prediction time domain, constraint conditions and system thermal characteristic parameters stored in the target control parameter set are called to predict the building load change trend and soil temperature field evolution in the future preset period, establish a system dynamic response model, and generate a benchmark operation strategy based on the model using a rolling time domain optimization strategy. The benchmark operation strategy includes the best operation point of the heat pump unit, the start-stop timing of the heat pipe array and the system operation parameter set value; The fuzzy control algorithm is used to process real-time input data, and the membership function parameters and rule base threshold values in the target control parameter set are called to perform fuzzy reasoning and output dynamic adjustment instructions. The real-time input data is extracted from the system operation state feature set, including the current soil temperature change rate calculated from the soil temperature distribution time series data and the heat pump unit load rate extracted from the real-time monitoring data of the heat pump unit operation power; The benchmark operation strategy output by the MPC algorithm and the dynamic adjustment instructions output by the fuzzy control algorithm are fused through a weight distribution mechanism. The weight distribution mechanism is dynamically adjusted based on the system operation mode, soil thermal imbalance risk level and real-time stability index calculated from the system operation state feature set. The real-time stability index includes the load fluctuation rate calculated based on the building load information, the temperature gradient change rate calculated based on the soil temperature distribution, and the equipment operation state index calculated based on the operation data; Based on the fusion result, the collaborative control instructions are generated, including the heat pump unit frequency set value, the heat pipe start-stop command and the heat exchanger medium flow adjustment value.

6. The method of operating and managing a heat pipe ground source heat pump system according to claim 5, wherein, The benchmark operation strategy output by the MPC algorithm and the dynamic adjustment instructions output by the fuzzy control algorithm are fused through a weight distribution mechanism, including: Based on historical operation mode data, timestamps and external weather conditions, the current operation scene label is identified through a preset scene classification model. The generated scene label is input into a preset priority evaluation model to output a priority score value of the scene label; The priority score value of the scene label is combined with the real-time stability index to generate a comprehensive scene weight through a preset comprehensive weight calculation rule; The comprehensive scene weight is input into a preset weight mapping model to determine the initial weight proportion of the MPC algorithm and the fuzzy control algorithm; Based on the initial weight proportion, combined with real-time system response data within a preset optimization period, the weight proportion is dynamically adjusted through a preset reinforcement learning algorithm to obtain an optimized weight proportion. The real-time system response data includes the heat pump unit operation state, the heat exchanger medium temperature, the heat pipe start-stop state and the soil temperature change rate; The optimized weight proportion is used to integrate the benchmark operation strategy output by the MPC algorithm and the dynamic adjustment instructions output by the fuzzy control algorithm through a preset weighted fusion algorithm.

Images