A control method and system for an energy-saving ground-source heat pump

Abstract

The present invention discloses a control method and system for an energy-saving ground-source heat pump, comprising: establishing a dynamic model of a ground-source heat pump system including an indoor heat balance, a ground-source heat pump unit, and a soil heat exchange model; obtaining its topological network through coupling; designing a distributed control and fault-tolerant operation strategy based on the topological network structure; locally controlling each node based on its location and information about adjacent nodes; constructing a state-space model; finding an optimal control action sequence through quadratic programming to balance energy consumption with an objective function of indoor and outdoor temperature deviation to achieve an optimal operating state; learning the mapping relationship between the optimal operating state and environmental parameters through a neural network; designing a ground-source circulation flow control algorithm to achieve automatic adjustment of the circulation pump flow distribution and the unit operating parameters. This method can effectively improve the control accuracy and stability of the ground-source heat pump system, ensure its reliable operation under complex working conditions, significantly improve energy utilization efficiency, achieve energy conservation and consumption reduction, and have good interpretability.

Description

Technical Field

[0001] The present invention relates to the field of energy system optimization, and in particular to a control method and system for an energy-saving ground source heat pump. Background Art

[0002] With the continued growth of global energy demand and the increasing emphasis on environmental protection, improving energy efficiency has become a key research direction in many fields. As an efficient and environmentally friendly heating and cooling technology, ground-source heat pumps (GSHPs) have been widely used in building environmental control. However, traditional GSHP systems still face numerous challenges in their actual operation.

[0003] Geothermal heat pump systems involve multiple complex subsystems: the indoor environment, the heat pump unit, and the soil. These systems experience intense thermal interaction and energy transfer. Existing control methods often struggle to accurately describe and coordinate the dynamic relationships between these subsystems, resulting in suboptimal system operation. Furthermore, the reliability and stability of geothermal heat pump systems are crucial for their long-term, efficient operation. In practical applications, various fault scenarios may arise. The lack of effective fault-tolerance mechanisms results in slow and inefficient system recovery, further increasing energy consumption and operating costs.

[0004] A control method and system for an energy-saving ground-source heat pump can comprehensively consider the complexity, reliability, and intelligent requirements of the ground-source heat pump system, design reasonable control strategies and intelligent algorithms, and achieve efficient energy utilization, a comfortable and stable indoor environment, and reliable operation of the system. Summary of the Invention

[0005] The object of the present invention is to provide a control method and system for an energy-saving ground source heat pump.

[0006] To achieve the above object, the present invention is implemented according to the following technical solutions:

[0007] A first aspect of the present invention provides a control method for an energy-saving ground-source heat pump, comprising:

[0008] S100 establishes a dynamic model of the ground source heat pump system, including the indoor heat balance model, the ground source heat pump unit model and the soil heat exchange model;

[0009] S200 couples the dynamic model of the ground source heat pump system to obtain a topological network of the ground source heat pump system;

[0010] S300 designs a distributed control strategy and a fault-tolerant operation strategy according to the structure of the topological network, and each node performs local control according to its position in the topological network and information about adjacent nodes;

[0011] S400 constructs a state space model of state variables and control variables in the local control, solves the future control action sequence through quadratic programming, optimizes the balance of the objective function of energy consumption and indoor and outdoor temperature deviation, and obtains the optimal operating state of the ground source heat pump system;

[0012] S500 uses a neural network to learn the mapping relationship between the optimal operating state and the corresponding environmental parameters;

[0013] S600 designs a ground source circulation flow control algorithm based on the mapping relationship and the predicted environmental data, and automatically adjusts the flow distribution of the circulation pump and the working parameters of the ground source heat pump unit.

[0014] As a further method, the method for establishing a dynamic model of a ground source heat pump system includes:

[0015] The indoor heat balance model is expressed as:

[0016]

[0017] Among them, C in is the total indoor heat capacity, T in is the indoor temperature, t is the time, n is the number of indoor surfaces, h i is the convective heat transfer coefficient of the i-th indoor surface, A i is the area of ​​the i-th indoor surface, T s,i is the temperature of the i-th indoor surface, ∈ i is the reflectivity of the i-th indoor surface, σ is the Stefan-Boltzmann constant, q hp is the heat / cooling provided by the ground source heat pump, m is the number of indoor and outdoor heat exchange paths, A j is the area of ​​the jth heat exchange path, h j is the convective heat transfer coefficient of the jth heat exchange path, d j is the material thickness of the jth heat exchange path, λ j is the thermal conductivity of the material in the jth heat exchange path;

[0018] The ground source heat pump unit model is expressed as:

[0019]

[0020] Where Q is the heating / cooling capacity of the ground source heat pump unit, m ref is the refrigerant mass flow rate, h1, h2, h3 and h4 are the refrigerant enthalpy at the evaporator inlet, the refrigerant enthalpy at the condenser inlet, the refrigerant enthalpy at the condenser outlet and the refrigerant enthalpy at the evaporator outlet, respectively, k is the polytropic index of the refrigerator in the compressor, p1 and p2 are the compressor inlet pressure and outlet pressure, V is the refrigerant volume flow rate at the compressor inlet, η is the compressor efficiency, Tevap and T cond They are the evaporator heat transfer and condenser heat transfer, a0 is the basic performance coefficient, a1 is the evaporator temperature influence coefficient, and a2 is the condenser temperature influence coefficient;

[0021] The soil heat exchange model is expressed as:

[0022]

[0023] Among them, ρ s is the soil density, C s is the specific heat capacity of soil, T s (r, z, t) is the soil temperature at radius r, depth z and time t, λ s is the thermal conductivity of soil, p is the number of buried pipes, q k (t) is the heat flux of the kth buried pipe at time t, and δ is the Dirac function.

[0024] As a further method, the method of coupling the dynamic model of the ground source heat pump system to obtain the topological network of the ground source heat pump system includes:

[0025] For the indoor heat balance model, computational fluid dynamics methods are used to simulate the flow and heat transfer process of indoor air to obtain indoor temperature field and velocity field information. For the ground source heat pump unit model, heat transfer equations are used to obtain the state changes and heat transfer information of the refrigerant in the evaporator, condenser and compressor. For the soil heat exchange model, the finite element method is used to obtain the heat conduction and heat convection process information in the soil.

[0026] Integrate the information of each model and identify the coupling variables, including temperature variables and flow variables, and establish coupling equations between the models based on energy conservation;

[0027] The indoor heat balance model, ground source heat pump unit model and soil heat exchange model are used as nodes in the topological network. The coupling relationship between the models is represented by directed edges, and the direction of the edge represents the flow of energy.

[0028] The indoor temperature set value is used as an external input node and connected to the ground source heat pump unit model node to control the operation of the unit and obtain the ground source heat pump system topology network.

[0029] As a further method, the method of designing a distributed control strategy and a fault-tolerant operation strategy according to the structure of the topological network includes:

[0030] S4.1 Distributed control strategy, specifically:

[0031] A central coordination hub is set up to collect data from indoor heat balance, ground source heat pump units and soil heat exchange nodes, and divide the indoor temperature control task into subtasks and send them to the corresponding nodes;

[0032] The indoor heat balance node estimates heating and cooling demand based on personnel, equipment, and climate, and sends this information to the ground-source heat pump node. The ground-source heat pump node adjusts the compressor speed and expansion valve opening according to subtask and demand information, combined with its own evaporator and condenser status and refrigerant parameters, and also feeds back operating parameters. The soil node controls the circulation pump flow and starts and stops according to the ground-source heat pump unit's heat demand and soil conditions.

[0033] Each node shares and interacts with the indoor and outdoor temperature and humidity, soil temperature change, and ground source heat pump unit operating parameter information;

[0034] S4.2 Fault-tolerant operation strategy, specifically:

[0035] Redundant configuration is adopted for the equipment and components in the ground source heat pump system. In case of failure, redundant equipment is automatically switched and the operating parameters are adjusted. If normal functions cannot be fully restored through redundant switching, the system enters a derated operation state and the indoor temperature set value, ground source heat pump unit operating parameters and soil heat exchange strategy are adjusted according to the performance of the remaining equipment or components.

[0036] As a further method, the method of performing local control of each node according to its position in the topological network and information about adjacent nodes includes:

[0037] S5.1 Each node obtains information about adjacent nodes based on the topological network connection relationship. Specifically, the indoor heat balance node obtains the circulating water pump information of the ground-source heat pump unit node, the heat exchange information of the adjacent indoor area, and the heat transfer information of the outdoor environment. The ground-source heat pump unit node receives the heat demand information of the indoor heat balance node and the soil and fluid information of the soil heat exchange node, and collects its own component status information. The soil heat exchange node receives the heat demand information of the ground-source heat pump unit node and obtains the ground pipe heat exchanger status information and soil temperature information.

[0038] S5.2 calculates heat parameters based on the acquired information. Specifically, the indoor heat balance node calculates the ground source heat pump heat supply, adjacent area heat exchange, and outdoor ambient heat transfer. The ground source heat pump unit node calculates the total heat demand, soil-side heat exchange, and evaporator and condenser heat exchange. The soil heat exchange node calculates the heat exchange efficiency.

[0039] S5.3 Each node establishes a heat balance or energy conservation equation based on the calculated heat parameters, specifically:

[0040] S5.3.1 Indoor heat balance node Establish the indoor heat balance equation, which is expressed as:

[0041]

[0042] Among them, c r is the equivalent specific heat capacity of indoor air and objects, ρ r is the equivalent density of indoor air and objects, V r is the volume of the interior space, q int is the initial indoor heat, q adj is the heat exchanged between the interior and adjacent areas through the enclosure structure, q env It is the heat exchanged between the interior and the outdoor environment through the enclosing structure;

[0043] If the temperature change trend deviates from the preset threshold range, the adjustment amount is calculated and an adjustment request is sent to the connected ground source heat pump unit node;

[0044] S5.3.2 Establish the energy conservation equation for the ground source heat pump unit node, which is expressed as:

[0045] T evap =T cond

[0046] Adjust the refrigerant flow and expansion valve opening and feedback the operating status information. The adjustment expression is:

[0047]

[0048] Among them, N comp is the compressor speed, k1 is the influence coefficient of the compressor speed on the refrigerant flow, k2 is the refrigerant basic flow coefficient, x exp is the expansion valve opening, p evap is the evaporator pressure, p cond is the condenser pressure, k3, k4 and k5 are the influence coefficients of expansion valve opening, evaporator pressure and condenser pressure on refrigerant flow respectively, and k6 is the comprehensive compensation coefficient of refrigerant;

[0049] S5.3.3 The soil heat exchange node calculates the soil temperature field change trend based on the heat conduction equation, where the heat conduction equation is expressed as:

[0050] Q demand =η1c p,fluid m pump (T fluid,in -T fluid,out )

[0051] Among them, Q demand is the heat demand of the ground source heat pump unit for soil heat exchange, η1 is the heat exchange efficiency of the buried pipe heat exchanger, c p,fluid is the specific heat capacity of the fluid on the soil side, m pump is the circulation pump flow rate, T fluid,in and Tfluid,out are the temperatures of the fluid entering and leaving the ground heat exchanger, respectively;

[0052] Adjust the circulation pump flow, control the circulation pump start and stop time and flow rate according to the soil temperature threshold, and feed back the soil temperature information and circulation pump operation status information to the connected ground source heat pump unit node.

[0053] As a further method, the method of constructing a state space model of state variables and control variables in the local control includes:

[0054] Set the state variable vector to x = [x1, x2, x3, x4, x5], which are the indoor temperature, the average temperature of the soil temperature in the heat exchanger, the refrigerant temperature of the ground source heat pump unit, the pressure of the circulation pump, and the power output of the ground source heat pump unit, respectively. The control variable vector is u = [u1, u2, u3], which are the flow adjustment amount of the circulation pump, the compressor frequency of the ground source heat pump unit, and the opening of the heat exchanger valve;

[0055] Construct the state space model, the expression is:

[0056]

[0057] Among them, K hx is the heat transfer coefficient of the indoor heat exchanger, K ghx is the heat transfer coefficient between the ground source heat pump unit and the soil heat exchanger, K soil is the heat transfer coefficient of the soil heat exchanger, T soil,far is the soil temperature of the heat exchanger, Q comp Compressor power function, Q cond is the heat transfer function of the condenser, Q evap is the evaporator heat transfer function, C pump is the equivalent heat capacity of the circulating pump, K pump is the circulation pump flow regulation function, R pipe is the pipeline resistance function, w1, w2, w3, w4 and w5 are the process noises of x1, x2, x3, x4 and x5 respectively.

[0058] As a further method, the method of solving the future control action sequence by quadratic programming to optimize the balance of the objective function of energy consumption and indoor and outdoor temperature deviation includes:

[0059] The objective function that comprehensively considers energy consumption and indoor and outdoor temperature deviation is expressed as:

[0060]

[0061] Among them, t0 and t fare the start and end time of the control period, α is the weight coefficient of power consumption, P total (t) is the total power consumption of the ground source heat pump system at time t, T in (t) is the actual indoor temperature at time t, T set (t) is the indoor set temperature at time t, θθ is the number of soil monitoring points, T g,i (t) is the soil temperature at the i-th monitoring point of the buried pipe at time t, T g,avg (t) is the average temperature of all soil monitoring points at time t;

[0062] Obtain the constraints from the state-space model, organize the objective function and constraints into the standard form of quadratic programming, and input them into the quadratic programming solver to obtain the future control action sequence to minimize the objective function and obtain the optimal operating state of the ground-source heat pump system.

[0063] As a further method, the method of using a neural network to learn the mapping relationship between the optimal operating state and the corresponding environmental parameters includes:

[0064] A multi-layer feedforward neural network is used to build the model. The input layer is the environmental parameters, and the output layer is the operating status of the ground source heat pump system. The expression of the neural network model output is:

[0065]

[0066] Among them, w0 is the bias term of the output layer, m0 is the number of neurons in the middle layer, and w j is the connection weight from the jth neuron in the middle layer to the output layer, f(·) is the Sigmoid activation function, θ j is the bias term of the jth neuron in the middle layer, n0 is the number of neurons in the input layer, and w ij is the connection weight from the i-th neuron in the input layer to the j-th neuron in the middle layer, x i is the environmental parameter of the i-th neuron in the input layer;

[0067] The optimal operating state of the ground source heat pump and the corresponding environmental parameters are combined into a data set, and the data set is divided into training set, test set and validation set in a ratio of 7:2:1 to train the neural network model.

[0068] As a further method, the method for designing a ground source circulation flow control algorithm based on the mapping relationship and the predicted environmental data includes:

[0069] The predicted environmental data is input into the neural network model, the expected operating state is output, and the proportional, integral and differential coefficients are determined according to the current operating state and the expected operating state;

[0070] The ground source circulation flow control algorithm is constructed using the PID controller, and the expression is:

[0071]

[0072] Among them, u(t) is the control signal of the PID controller at time t, K p , K i and K d are proportional, integral and differential coefficients respectively, Q s (t) is the circulating pump flow rate and the working parameter status of the unit at time t, Q a (t) is the circulation pump flow rate and the working parameter status of the unit at time t.

[0073] A second aspect of the present invention provides an energy-saving ground source heat pump control system, comprising:

[0074] Dynamic model building module, used to build the dynamic model of the ground source heat pump system, including indoor heat balance model, ground source heat pump unit model and soil heat exchange model;

[0075] A topology network generation module is used to couple the dynamic model of the ground source heat pump system to obtain a topology network of the ground source heat pump system;

[0076] A distributed fault-tolerant control module is used to design a distributed control strategy and a fault-tolerant operation strategy according to the structure of the topological network, and each node performs local control according to its position in the topological network and information about adjacent nodes;

[0077] A state space optimization module is used to construct a state space model of the state variables and control variables in the local control, and solve the future control action sequence through quadratic programming to optimize the balance of the objective functions of energy consumption and indoor and outdoor temperature deviation, thereby obtaining the optimal operating state of the ground source heat pump system;

[0078] A neural network learning module, configured to learn a mapping relationship between the optimal operating state and corresponding environmental parameters using a neural network;

[0079] The flow control algorithm module is used to design a ground source circulation flow control algorithm based on the mapping relationship and the predicted environmental data, and automatically adjust the flow distribution of the circulation pump and the working parameters of the ground source heat pump unit.

[0080] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0081] (1) The present invention establishes a topological network and a distributed control strategy based on a dynamic model to realize intelligent local control of each node. The information exchange between nodes is rapid, and local changes can be responded to quickly, which reduces control delays and enhances the system's adaptability to dynamic loads. It ensures that the system always operates in an efficient state, improves the overall energy utilization efficiency and warms the indoor environment.

[0082] (2) The present invention can realize intelligent control by using a neural network to learn the mapping relationship between the optimal operating state and environmental parameters. The ground source circulation flow control algorithm designed based on the predicted environmental data can automatically adjust the circulation pump flow distribution and the working parameters of the ground source heat pump unit, and adjust the system operation mode in advance according to the changes in environmental parameters, so that the system is always in an efficient operating state without the need for frequent manual intervention, saving labor costs and improving user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] Figure 1 This is a flowchart of the steps of a control method for an energy-saving ground source heat pump in an embodiment of the present invention. DETAILED DESCRIPTION

[0084] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0085] Reference Figure 1 As shown, the present invention provides an energy-saving ground source heat pump control method, comprising:

[0086] S100 establishes a dynamic model of the ground source heat pump system, including the indoor heat balance model, the ground source heat pump unit model and the soil heat exchange model;

[0087] In the actual evaluation, the system was applied to an office building with a building area of ​​4000 square meters. For the indoor thermal balance model, the total indoor heat capacity was calculated to be 1.5×10^6 J / ℃. There were four surfaces with different orientations and materials in the room, and the convection heat transfer coefficients were 6W / (m 2 ·℃)、8W / (m 2 ·℃)、7W / (m 2 ·℃)、9W / (m 2 ℃), the area is 300m 2 , 400m 2 、350m 2 、450m 2The main indoor and outdoor heat exchange paths are the exterior wall and glass curtain wall. The thickness of the exterior wall material is 0.35 meters, the thermal conductivity is 0.7W / (m·℃), the area of ​​the glass curtain wall is 500 square meters, and the convection heat transfer coefficient is 22W / (m 2 ·℃), the refrigerant mass flow rate of the ground source heat pump unit is between 0.6-2.2kg / s, the refrigerant enthalpy at the evaporator inlet is 390kJ / kg, and the refrigerant enthalpy at the condenser inlet is about 240kJ / kg. Professional unit performance tests show that the soil density is 2100kg / m 3 , specific heat capacity 1100 J / (kg·℃), number of buried pipes 40, heat flow of each buried pipe between 3-9 kW, and further obtain the dynamic model of the ground source heat pump system.

[0088] S200 couples the dynamic model of the ground source heat pump system to obtain a topological network of the ground source heat pump system;

[0089] In actual evaluations, computational fluid dynamics (CFD) simulations of indoor air flow revealed that wind speeds near air conditioning outlets can reach 0.8 m / s, with relatively rapid temperature fluctuations. Heat transfer equations were used to analyze the ground-source heat pump unit, indicating that the condenser heat exchange can reach up to 120 kW during summer cooling. Finite element simulations of soil heat exchange revealed that the temperature at 4 meters below ground level is less affected by seasonal fluctuations. After determining the coupling variables, coupling equations were established based on the principle of energy conservation, and a topological network was constructed. The indoor temperature setpoint was connected as an external input node to the ground-source heat pump unit model node, enabling precise control of unit operation.

[0090] S300 designs a distributed control strategy and a fault-tolerant operation strategy according to the structure of the topological network, and each node performs local control according to its position in the topological network and information about adjacent nodes;

[0091] In the actual evaluation, the central coordination hub is responsible for data collection and task allocation. In one scenario in the summer afternoon, the indoor heat balance node estimated the cooling demand to be 80kW based on the heating conditions of indoor personnel and office equipment, and sent it to the ground source heat pump unit node. The unit node adjusts the compressor speed to 1800r / min and the expansion valve opening to 35% based on its own evaporator temperature of 8°C, condenser temperature of 42°C and refrigerant parameters, and promptly feeds back the operating parameters. The soil node controls the circulation pump flow to 35m / min based on the heat demand of the ground source heat pump unit and the soil temperature of 20°C. 3 If the ground source heat pump unit's compressor suddenly fails, the redundant compressor will automatically switch immediately. If the redundant equipment cannot fully restore normal function, the system will enter derated operation, the indoor temperature setpoint will be adjusted to 28°C, the ground source heat pump unit's operating parameters will be adjusted accordingly, the cooling output will be reduced, and the soil exchange strategy will also be optimized to reduce the soil heat load.

[0092] In the actual evaluation, the indoor heat balance node obtains the circulating water pump information of the ground source heat pump unit and calculates that the ground source heat pump heat supply is 75kW, the adjacent area heat exchange is 5kW and the outdoor environment heat transfer is 3kW. The ground source heat pump unit node calculates the total heat demand of 83kW, the soil side heat exchange capacity of 25kW, the evaporator and condenser heat exchange capacity of 58kW and 25kW respectively, adjusts the refrigerant flow and expansion valve opening, and the soil heat exchange node calculates the heat exchange efficiency of 0.62. Based on the soil temperature threshold (upper limit 25℃ in summer, lower limit 8℃ in winter), the start and stop time and flow rate of the circulating pump are controlled. When the soil temperature reaches 23℃, the circulating pump flow rate is increased to 38m 3 / h.

[0093] S400 constructs a state space model of state variables and control variables in the local control, solves the future control action sequence through quadratic programming, optimizes the balance of the objective function of energy consumption and indoor and outdoor temperature deviation, and obtains the optimal operating state of the ground source heat pump system;

[0094] It's important to explain that the goal of these steps is to find the optimal operating strategy for the ground-source heat pump system. By integrating key system states and control variables into a model, their internal relationships are clearly visualized. This approach utilizes quadratic programming to balance energy and temperature deviations based on the objective function. This determines the control action sequence within constraints such as equipment operation, ensuring optimal system operation that meets indoor comfort requirements while minimizing energy waste and improving the overall system's economic efficiency and stability.

[0095] In the actual evaluation, the indoor temperature is 27℃, the soil temperature is 21℃ at the average temperature of the heat exchanger, the refrigerant temperature of the ground source heat pump unit is 32℃, the pressure of the circulation pump is 0.32MPa, the power output of the ground source heat pump unit is 60kW, and the flow adjustment volume of the circulation pump is 6m 3 / h, the compressor frequency of the ground source heat pump unit is 1600Hz and the opening of the heat exchanger valve is 30%. Based on these actual data, a state space model is constructed, and the power consumption weight coefficient is set to 0.65. On a certain day in autumn, the future control action sequence is obtained through quadratic programming to minimize the objective function. The optimal operating state is determined as follows: the indoor temperature is stable at 22°C, the soil temperature is maintained at an average temperature of 18°C ​​at the heat exchanger, the refrigerant temperature of the ground source heat pump unit is maintained at 26°C, the pressure of the circulating pump is stable at 0.28MPa, and the power output of the ground source heat pump unit is adjusted to 45kW. At this time, the compressor frequency of the ground source heat pump unit is optimized to 1300Hz, the opening of the heat exchanger valve is adjusted to 22%, and the flow adjustment amount of the circulating pump is 4m 3 / h. This state can meet indoor comfort requirements while minimizing energy consumption.

[0096] S500 uses a neural network to learn the mapping relationship between the optimal operating state and the corresponding environmental parameters;

[0097] In the actual evaluation, a multi-layer feedforward neural network was used. The input layer was outdoor temperature, humidity, and solar radiation intensity, and the output layer was the operating status of the ground-source heat pump. The optimal operating status of the ground-source heat pump and the corresponding environmental parameters were combined into a data set, which was divided into training set, test set, and validation set in a ratio of 7:2:1 to train the neural network model.

[0098] S600 designs a ground source circulation flow control algorithm based on the mapping relationship and the predicted environmental data, and automatically adjusts the flow distribution of the circulation pump and the working parameters of the ground source heat pump unit.

[0099] In the actual evaluation, the predicted environmental data for the next 4 hours (outdoor temperature gradually decreases to 12°C and humidity increases to 70%) is input into the neural network model, and the expected operating state is output. The proportional, integral, and differential coefficients are determined based on the current operating state and the expected operating state. The PID controller is used to construct the ground source circulation flow control algorithm, and it is calculated that the circulation pump flow should be adjusted to 28m 3 / h, the unit's operating parameters are also optimized accordingly, thereby automatically adjusting the flow distribution of the circulation pump and the operating parameters of the ground-source heat pump unit to ensure continuous and efficient operation of the system;

[0100] Specifically, for a certain day's weather forecast, from 6 a.m. to 12 noon, the outdoor temperature slowly rises from 10°C to 18°C, the humidity drops from 60% to 40%, and the wind speed is maintained at about 3 m / s. The control result is that the circulation pump flow rate is reduced to 60% of the rated flow rate, and the operating frequency of the ground source heat pump unit is reduced to 70% of the rated frequency; from 12 noon to 6 p.m., the temperature rises from 18°C ​​to 25°C, the humidity is maintained at about 40%, and the wind speed increases to 5 m / s. The control result is that the circulation pump flow rate is increased to 80% of the rated flow rate, and the operating frequency of the ground source heat pump unit is increased to 90% of the rated frequency; from 6 p.m. to 12 p.m., the temperature drops from 25°C to 15°C, the humidity rises to 60%, and the wind speed drops to 2 m / s. The control result is that the circulation pump flow rate is maintained at 70% of the rated flow rate, the heating operation parameters of the ground source heat pump unit are increased, and the operating frequency is increased to 85% of the rated frequency;

[0101] Using this method for a complete heating and cooling season can save about 10.2% energy compared with the traditional control method. Under the traditional control method, the indoor temperature fluctuates greatly. During cooling in summer, the indoor temperature fluctuates within ±2°C, and during heating in winter, the indoor temperature fluctuates within ±1.5°C. After using the flow control algorithm, the indoor temperature fluctuation range in summer is controlled within ±0.5°C, and the indoor temperature fluctuation range in winter is controlled within ±0.8°C. Under the traditional control method, when parameters such as indoor ambient temperature, humidity and wind speed change, the response time of the ground source heat pump system usually takes 30 minutes or more to adjust to the appropriate operating state. After adopting the new flow control algorithm, the system can complete the adjustment of the circulation pump flow and the working parameters of the ground source heat pump unit within 10 minutes, and the system response is significantly shortened.

[0102] In this embodiment, the method for establishing a dynamic model of a ground-source heat pump system includes:

[0103] The indoor heat balance model is expressed as:

[0104]

[0105] Among them, C in is the total indoor heat capacity, T in is the indoor temperature, t is the time, n is the number of indoor surfaces, h i is the convective heat transfer coefficient of the i-th indoor surface, A i is the area of ​​the i-th indoor surface, T s,i is the temperature of the i-th indoor surface, ∈ i is the reflectivity of the i-th indoor surface, σ is the Stefan-Boltzmann constant, q hp is the heat / cooling provided by the ground source heat pump, m is the number of indoor and outdoor heat exchange paths, A j is the area of ​​the jth heat exchange path, h j is the convective heat transfer coefficient of the jth heat exchange path, d j is the material thickness of the jth heat exchange path, λ j is the thermal conductivity of the material in the jth heat exchange path;

[0106] The ground source heat pump unit model is expressed as:

[0107]

[0108] Where Q is the heating / cooling capacity of the ground source heat pump unit, m refis the refrigerant mass flow rate, h1, h2, h3 and h4 are the refrigerant enthalpy at the evaporator inlet, the refrigerant enthalpy at the condenser inlet, the refrigerant enthalpy at the condenser outlet and the refrigerant enthalpy at the evaporator outlet, respectively, k is the polytropic index of the refrigerator in the compressor, p1 and p2 are the compressor inlet pressure and outlet pressure, V is the refrigerant volume flow rate at the compressor inlet, η is the compressor efficiency, T evap and T cond They are the evaporator heat transfer and condenser heat transfer, a0 is the basic performance coefficient, a1 is the evaporator temperature influence coefficient, and a2 is the condenser temperature influence coefficient;

[0109] The soil heat exchange model is expressed as:

[0110]

[0111] Among them, ρ s is the soil density, C s is the specific heat capacity of soil, T s (r, z, t) is the soil temperature at radius r, depth z and time t, λ s is the thermal conductivity of soil, p is the number of buried pipes, q k (t) is the heat flux of the kth buried pipe at time t, and δ is the Dirac function.

[0112] In this embodiment, the method of coupling the dynamic model of the ground-source heat pump system to obtain the topological network of the ground-source heat pump system includes:

[0113] For the indoor heat balance model, computational fluid dynamics methods are used to simulate the flow and heat transfer process of indoor air to obtain indoor temperature field and velocity field information. For the ground source heat pump unit model, heat transfer equations are used to obtain the state changes and heat transfer information of the refrigerant in the evaporator, condenser and compressor. For the soil heat exchange model, the finite element method is used to obtain the heat conduction and heat convection process information in the soil.

[0114] Integrate the information of each model and identify the coupling variables, including temperature variables and flow variables, and establish coupling equations between the models based on energy conservation;

[0115] The indoor heat balance model, ground source heat pump unit model and soil heat exchange model are used as nodes in the topological network. The coupling relationship between the models is represented by directed edges, and the direction of the edge represents the flow of energy.

[0116] The indoor temperature set value is used as an external input node and connected to the ground source heat pump unit model node to control the operation of the unit and obtain the ground source heat pump system topology network.

[0117] In this embodiment, the method for designing a distributed control strategy and a fault-tolerant operation strategy based on the structure of the topological network includes:

[0118] S4.1 Distributed control strategy, specifically:

[0119] A central coordination hub is set up to collect data from indoor heat balance, ground source heat pump units and soil heat exchange nodes, and divide the indoor temperature control task into subtasks and send them to the corresponding nodes;

[0120] The indoor heat balance node estimates heating and cooling demand based on personnel, equipment, and climate, and sends this information to the ground-source heat pump node. The ground-source heat pump node adjusts the compressor speed and expansion valve opening according to subtask and demand information, combined with its own evaporator and condenser status and refrigerant parameters, and also feeds back operating parameters. The soil node controls the circulation pump flow and starts and stops according to the ground-source heat pump unit's heat demand and soil conditions.

[0121] Each node shares and interacts with the indoor and outdoor temperature and humidity, soil temperature change, and ground source heat pump unit operating parameter information;

[0122] S4.2 Fault-tolerant operation strategy, specifically:

[0123] Redundant configuration is adopted for the equipment and components in the ground source heat pump system. In case of failure, redundant equipment is automatically switched and the operating parameters are adjusted. If normal functions cannot be fully restored through redundant switching, the system enters a derated operation state and the indoor temperature set value, ground source heat pump unit operating parameters and soil heat exchange strategy are adjusted according to the performance of the remaining equipment or components.

[0124] In this embodiment, the method for each node to perform local control based on its position in the topological network and information about adjacent nodes includes:

[0125] S5.1 Each node obtains information about adjacent nodes based on the topological network connection relationship. Specifically, the indoor heat balance node obtains the circulating water pump information of the ground-source heat pump unit node, the heat exchange information of the adjacent indoor area, and the heat transfer information of the outdoor environment. The ground-source heat pump unit node receives the heat demand information of the indoor heat balance node and the soil and fluid information of the soil heat exchange node, and collects its own component status information. The soil heat exchange node receives the heat demand information of the ground-source heat pump unit node and obtains the ground pipe heat exchanger status information and soil temperature information.

[0126] S5.2 calculates heat parameters based on the acquired information. Specifically, the indoor heat balance node calculates the ground source heat pump heat supply, adjacent area heat exchange, and outdoor ambient heat transfer. The ground source heat pump unit node calculates the total heat demand, soil-side heat exchange, and evaporator and condenser heat exchange. The soil heat exchange node calculates the heat exchange efficiency.

[0127] S5.3 Each node establishes a heat balance or energy conservation equation based on the calculated heat parameters, specifically:

[0128] S5.3.1 Indoor heat balance node Establish the indoor heat balance equation, which is expressed as:

[0129]

[0130] Among them, c r is the equivalent specific heat capacity of indoor air and objects, ρ r is the equivalent density of indoor air and objects, V r is the volume of the interior space, q int is the initial indoor heat, q adj is the heat exchanged between the interior and adjacent areas through the enclosure structure, q env It is the heat exchanged between the interior and the outdoor environment through the enclosing structure;

[0131] If the temperature change trend deviates from the preset threshold range, the adjustment amount is calculated and an adjustment request is sent to the connected ground source heat pump unit node;

[0132] S5.3.2 Establish the energy conservation equation for the ground source heat pump unit node, which is expressed as:

[0133] T evap =T cond

[0134] Adjust the refrigerant flow and expansion valve opening and feedback the operating status information. The adjustment expression is:

[0135]

[0136] Among them, N comp is the compressor speed, k1 is the influence coefficient of the compressor speed on the refrigerant flow, k2 is the refrigerant basic flow coefficient, x exp is the expansion valve opening, p evap is the evaporator pressure, p cond is the condenser pressure, k3, k4 and k5 are the influence coefficients of expansion valve opening, evaporator pressure and condenser pressure on refrigerant flow respectively, and k6 is the comprehensive compensation coefficient of refrigerant;

[0137] S5.3.3 The soil heat exchange node calculates the soil temperature field change trend based on the heat conduction equation, where the heat conduction equation is expressed as:

[0138] Q demand =η1c p,fluid m pump (T fluid,in -T fluid,out )

[0139] Among them, Qdemand is the heat demand of the ground source heat pump unit for soil heat exchange, η1 is the heat exchange efficiency of the buried pipe heat exchanger, c p,fluid is the specific heat capacity of the fluid on the soil side, m pump is the circulation pump flow rate, T fluid,in and T fluid,out are the temperatures of the fluid entering and leaving the ground heat exchanger, respectively;

[0140] Adjust the circulation pump flow, control the circulation pump start and stop time and flow rate according to the soil temperature threshold, and feed back the soil temperature information and circulation pump operation status information to the connected ground source heat pump unit node.

[0141] In this embodiment, the method for constructing a state space model of state variables and control variables in the local control includes:

[0142] Set the state variable vector to x = [x1, x2, x3, x4, x5], which are the indoor temperature, the average temperature of the soil temperature in the heat exchanger, the refrigerant temperature of the ground source heat pump unit, the pressure of the circulation pump, and the power output of the ground source heat pump unit, respectively. The control variable vector is u = [u1, u2, u3], which are the flow adjustment amount of the circulation pump, the compressor frequency of the ground source heat pump unit, and the opening of the heat exchanger valve;

[0143] Construct the state space model, the expression is:

[0144]

[0145] Among them, K hx is the heat transfer coefficient of the indoor heat exchanger, K ghx is the heat transfer coefficient between the ground source heat pump unit and the soil heat exchanger, K soil is the heat transfer coefficient of the soil heat exchanger, T soil,far is the soil temperature of the heat exchanger, Q comp Compressor power function, Q cond is the heat transfer function of the condenser, Q evap is the evaporator heat transfer function, C pump is the equivalent heat capacity of the circulating pump, K pump is the circulation pump flow regulation function, R pipe is the pipeline resistance function, w1, w2, w3, w4 and w5 are the process noises of x1, x2, x3, x4 and x5 respectively.

[0146] In this embodiment, the method for solving the future control action sequence through quadratic programming to optimize the balance of the objective functions of energy consumption and indoor and outdoor temperature deviation includes:

[0147] The objective function that comprehensively considers energy consumption and indoor and outdoor temperature deviation is expressed as:

[0148]

[0149] Among them, t0 and t f are the start and end time of the control period, α is the weight coefficient of power consumption, P total (t) is the total power consumption of the ground source heat pump system at time t, T in (t) is the actual indoor temperature at time t, T set (t) is the indoor set temperature at time t, θθ is the number of soil monitoring points, T g,i (t) is the soil temperature at the i-th monitoring point of the buried pipe at time t, T g,avg (t) is the average temperature of all soil monitoring points at time t;

[0150] Obtain the constraints from the state-space model, organize the objective function and constraints into the standard form of quadratic programming, and input them into the quadratic programming solver to obtain the future control action sequence to minimize the objective function and obtain the optimal operating state of the ground-source heat pump system.

[0151] In this embodiment, the method of using a neural network to learn the mapping relationship between the optimal operating state and the corresponding environmental parameters includes:

[0152] A multi-layer feedforward neural network is used to build the model. The input layer is the environmental parameters, and the output layer is the operating status of the ground source heat pump system. The expression of the neural network model output is:

[0153]

[0154] Among them, w0 is the bias term of the output layer, m0 is the number of neurons in the middle layer, and w j is the connection weight from the jth neuron in the middle layer to the output layer, f(·) is the Sigmoid activation function, θ j is the bias term of the jth neuron in the middle layer, n0 is the number of neurons in the input layer, and w ij is the connection weight from the i-th neuron in the input layer to the j-th neuron in the middle layer, x i is the environmental parameter of the i-th neuron in the input layer;

[0155] The optimal operating state of the ground source heat pump and the corresponding environmental parameters are combined into a data set, and the data set is divided into training set, test set and validation set in a ratio of 7:2:1 to train the neural network model.

[0156] In this embodiment, the method for designing a ground source circulation flow control algorithm based on the mapping relationship and the predicted environmental data includes:

[0157] The predicted environmental data is input into the neural network model, the expected operating state is output, and the proportional, integral and differential coefficients are determined according to the current operating state and the expected operating state;

[0158] The ground source circulation flow control algorithm is constructed using the PID controller, and the expression is:

[0159]

[0160] Among them, u(t) is the control signal of the PID controller at time t, K p , K i and K d are proportional, integral and differential coefficients respectively, Q s (t) is the circulating pump flow rate and the working parameter status of the unit at time t, Q a (t) is the circulation pump flow rate and the working parameter status of the unit at time t.

[0161] A second aspect of the present invention provides an energy-saving ground source heat pump control system, comprising:

[0162] Dynamic model building module, used to build the dynamic model of the ground source heat pump system, including indoor heat balance model, ground source heat pump unit model and soil heat exchange model;

[0163] A topology network generation module is used to couple the dynamic model of the ground source heat pump system to obtain a topology network of the ground source heat pump system;

[0164] A distributed fault-tolerant control module is used to design a distributed control strategy and a fault-tolerant operation strategy according to the structure of the topological network, and each node performs local control according to its position in the topological network and information about adjacent nodes;

[0165] A state space optimization module is used to construct a state space model of the state variables and control variables in the local control, and solve the future control action sequence through quadratic programming to optimize the balance of the objective functions of energy consumption and indoor and outdoor temperature deviation, thereby obtaining the optimal operating state of the ground source heat pump system;

[0166] A neural network learning module, configured to learn a mapping relationship between the optimal operating state and corresponding environmental parameters using a neural network;

[0167] The flow control algorithm module is used to design a ground source circulation flow control algorithm based on the mapping relationship and the predicted environmental data, and automatically adjust the flow distribution of the circulation pump and the working parameters of the ground source heat pump unit.

[0168] The above content is merely an example and explanation of the structure of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, they should all fall within the scope of protection of the present invention.

Claims

1. A method for controlling an energy-saving ground source heat pump, characterized in that: The following steps are involved: Establish a dynamic model of the ground source heat pump system, including the indoor heat balance model, ground source heat pump unit model and soil heat exchange model; coupling the dynamic model of the ground source heat pump system to obtain a topological network of the ground source heat pump system; Designing a distributed control strategy and a fault-tolerant operation strategy based on the structure of the topological network, wherein each node performs local control based on its position in the topological network and information about adjacent nodes; Constructing a state space model of state variables and control variables in the local control, solving the future control action sequence through quadratic programming, optimizing the balance of the objective function of energy consumption and indoor and outdoor temperature deviation, and obtaining the optimal operating state of the ground source heat pump system; Using a neural network to learn the mapping relationship between the optimal operating state and the corresponding environmental parameters; Based on the mapping relationship and the predicted environmental data, a ground source circulation flow control algorithm is designed to automatically adjust the flow distribution of the circulation pump and the working parameters of the ground source heat pump unit.

2. The control method of an energy-saving ground source heat pump according to claim 1, characterized in that: The method for establishing a dynamic model of a ground-source heat pump system comprises: The indoor heat balance model is expressed as: in, is the total indoor heat capacity, is the indoor temperature, For time, is the number of indoor surfaces, For the The convective heat transfer coefficient of each indoor surface is For the The area of ​​the interior surface, For the The temperature of the indoor surface, For the The reflectivity of the indoor surface, is the Stefan-Boltzmann constant, The heat / cooling provided by the ground source heat pump, is the number of indoor and outdoor heat exchange paths, For the The area of ​​the heat exchange path, For the The convective heat transfer coefficient of each heat exchange path is For the The material thickness of each heat exchange path, For the Thermal conductivity of the materials in each heat exchange path; The ground source heat pump unit model is expressed as: in, is the heating / cooling capacity of the ground source heat pump unit, is the refrigerant mass flow rate, 、 、 and are the refrigerant enthalpy at the evaporator inlet, the refrigerant enthalpy at the condenser inlet, the refrigerant enthalpy at the condenser outlet, and the refrigerant enthalpy at the evaporator outlet. is the variable index of the refrigerator in the compressor, and are the compressor inlet pressure and outlet pressure, is the refrigerant volume flow rate at the compressor inlet, is the compressor efficiency, and are the evaporator heat transfer and the condenser heat transfer, is the basic performance coefficient, is the evaporator temperature influence coefficient, is the condenser temperature influence coefficient; The soil heat exchange model is expressed as: in, is the soil density, is the specific heat capacity of soil, For soil in radius ,depth and time temperature, is the thermal conductivity of soil, is the number of buried pipes, For the buried pipes in time The heat flow, is the Dirac function.

3. The control method of an energy-saving ground source heat pump according to claim 1, characterized in that: The method of coupling the dynamic model of the ground source heat pump system to obtain the topological network of the ground source heat pump system includes: For the indoor heat balance model, computational fluid dynamics methods are used to simulate the flow and heat transfer process of indoor air to obtain indoor temperature field and velocity field information. For the ground source heat pump unit model, heat transfer equations are used to obtain the state changes and heat transfer information of the refrigerant in the evaporator, condenser and compressor. For the soil heat exchange model, the finite element method is used to obtain the heat conduction and heat convection process information in the soil. Integrate the information of each model and identify the coupling variables, including temperature variables and flow variables, and establish coupling equations between the models based on energy conservation; The indoor heat balance model, ground source heat pump unit model and soil heat exchange model are used as nodes in the topological network. The coupling relationship between the models is represented by directed edges, and the direction of the edge represents the flow of energy. The indoor temperature set value is used as an external input node and connected to the ground source heat pump unit model node to control the operation of the unit and obtain the ground source heat pump system topology network.

4. The control method of an energy-saving ground source heat pump according to claim 1, characterized in that: The method for designing a distributed control strategy and a fault-tolerant operation strategy according to the structure of the topological network includes: S4.1 Distributed control strategy, specifically: A central coordination hub is set up to collect data from indoor heat balance, ground source heat pump units and soil heat exchange nodes, and divide the indoor temperature control task into subtasks and send them to the corresponding nodes; The indoor heat balance node estimates heating and cooling demand based on personnel, equipment, and climate, and sends this information to the ground-source heat pump node. The ground-source heat pump node adjusts the compressor speed and expansion valve opening according to subtask and demand information, combined with its own evaporator and condenser status and refrigerant parameters, and also feeds back operating parameters. The soil node controls the circulation pump flow and starts and stops according to the ground-source heat pump unit's heat demand and soil conditions. Each node shares and interacts with the indoor and outdoor temperature and humidity, soil temperature change, and ground source heat pump unit operating parameter information; S4.2 Fault-tolerant operation strategy, specifically: Redundant configuration is adopted for the equipment and components in the ground source heat pump system. In case of failure, redundant equipment is automatically switched and the operating parameters are adjusted. If normal functions cannot be fully restored through redundant switching, the system enters a derated operation state and the indoor temperature set value, ground source heat pump unit operating parameters and soil heat exchange strategy are adjusted according to the performance of the remaining equipment or components.

5. The control method of an energy-saving ground source heat pump according to claim 1, characterized in that: The method of performing local control on each node according to its position in the topological network and information about adjacent nodes includes: S5.1 Each node obtains information about adjacent nodes based on the topological network connection relationship. Specifically, the indoor heat balance node obtains the circulating water pump information of the ground-source heat pump unit node, the heat exchange information of the adjacent indoor area, and the heat transfer information of the outdoor environment. The ground-source heat pump unit node receives the heat demand information of the indoor heat balance node and the soil and fluid information of the soil heat exchange node, and collects its own component status information. The soil heat exchange node receives the heat demand information of the ground-source heat pump unit node and obtains the ground pipe heat exchanger status information and soil temperature information. S5.2 calculates heat parameters based on the acquired information. Specifically, the indoor heat balance node calculates the ground source heat pump heat supply, adjacent area heat exchange, and outdoor ambient heat transfer. The ground source heat pump unit node calculates the total heat demand, soil-side heat exchange, and evaporator and condenser heat exchange. The soil heat exchange node calculates the heat exchange efficiency. S5.3 Each node establishes a heat balance or energy conservation equation based on the calculated heat parameters, specifically: S5.3.1 Indoor heat balance node Establish the indoor heat balance equation, which is expressed as: in, is the equivalent specific heat capacity of indoor air and objects, is the equivalent density of indoor air and objects, is the volume of the interior space, is the initial indoor heat, The heat exchanged between the interior and adjacent areas through the enclosure structure. It is the heat exchanged between the interior and the outdoor environment through the enclosing structure; Indicates the amount of heat or cooling provided by the ground source heat pump unit to the room. If the temperature change trend deviates from the preset threshold range, the adjustment amount is calculated and an adjustment request is sent to the connected ground source heat pump unit node; S5.3.2 Establish the energy conservation equation for the ground source heat pump unit node, which is expressed as: Adjust the refrigerant flow and expansion valve opening and feedback the operating status information. The adjustment expression is: in, is the compressor speed, is the influence coefficient of compressor speed on refrigerant flow rate, is the basic flow coefficient of refrigerant, is the expansion valve opening, is the evaporator pressure, is the condenser pressure, 、 and are the influence coefficients of expansion valve opening, evaporator pressure and condenser pressure on refrigerant flow, is the comprehensive compensation coefficient of refrigerant, Indicates the mass flow rate of refrigerant; S5.3.3 The soil heat exchange node calculates the soil temperature field change trend based on the heat conduction equation, where the heat conduction equation is expressed as: in, is the heat demand of the ground source heat pump unit for soil heat exchange, is the heat exchange efficiency of the ground heat exchanger, is the specific heat capacity of the fluid on the soil side, is the circulation pump flow rate, and are the temperatures of the fluid entering and leaving the ground heat exchanger, respectively; Adjust the circulation pump flow, control the circulation pump start and stop time and flow rate according to the soil temperature threshold, and feed back the soil temperature information and circulation pump operation status information to the connected ground source heat pump unit node.

6. The control method of an energy-saving ground source heat pump according to claim 1, characterized in that: The method for constructing a state space model of state variables and control variables in the local control includes: Set the state variable vector to , are respectively the indoor temperature, the average temperature of the soil temperature in the heat exchanger, the refrigerant temperature of the ground source heat pump unit, the pressure of the circulation pump and the power output of the ground source heat pump unit. The control variable vector is , are the flow regulation of the circulation pump, the compressor frequency of the ground source heat pump unit and the opening of the heat exchanger valve respectively; Construct the state space model, the expression is: in, is the heat transfer coefficient of the indoor heat exchanger, is the heat transfer coefficient between the ground source heat pump unit and the soil heat exchanger, is the heat transfer coefficient of the soil heat exchanger, is the soil temperature of the heat exchanger, Compressor power function, is the condenser heat transfer function, is the evaporator heat transfer function, is the equivalent heat capacity of the circulating pump, is the circulation pump flow regulation function, is the pipeline resistance function, 、 、 、 and They are 、 、 、 and The process noise, represents the total indoor heat capacity, q int Indicates the initial indoor heat. It represents the heat exchanged between the indoor and adjacent areas through the enclosure structure. It represents the heat exchanged between the indoor environment and the outdoor environment through the enclosure structure. That is, the refrigerant mass flow rate.

7. The control method of an energy-saving ground source heat pump according to claim 1, characterized in that: The method for solving the future control action sequence through quadratic programming to optimize the balance of the objective functions of energy consumption and indoor and outdoor temperature deviation includes: The objective function that comprehensively considers energy consumption and indoor and outdoor temperature deviation is expressed as: in, and are the start and end time of the control time period, is the weight coefficient of power consumption, Ground source heat pump system The total power consumption at the moment, for The actual indoor temperature at the moment, for Set the indoor temperature at all times. is the number of soil monitoring points, for Time buried pipe The soil temperature at each monitoring point, for The average temperature of all soil monitoring points at the moment; Obtain the constraints from the state-space model, organize the objective function and constraints into the standard form of quadratic programming, and input them into the quadratic programming solver to obtain the future control action sequence to minimize the objective function and obtain the optimal operating state of the ground-source heat pump system.

8. The control method of an energy-saving ground source heat pump according to claim 1, characterized in that: The method of using a neural network to learn the mapping relationship between the optimal operating state and the corresponding environmental parameters includes: A multi-layer feedforward neural network is used to build the model. The input layer is the environmental parameters, and the output layer is the operating status of the ground source heat pump system. The expression of the neural network model output is: in, is the bias term of the output layer, is the number of neurons in the middle layer, For the middle layer The connection weights of neurons to the output layer, is the Sigmoid activation function, For the middle layer The bias term of a neuron, is the number of neurons in the input layer, The input layer neurons to the middle layer The connection weights of neurons, The input layer Environmental parameters of each neuron; The optimal operating state of the ground source heat pump and the corresponding environmental parameters are combined into a data set, and the data set is divided into training set, test set and validation set in a ratio of 7:2:1 to train the neural network model.

9. The energy-saving ground source heat pump control method according to claim 1, characterized in that: The method for designing a ground-source circulation flow control algorithm based on the mapping relationship and the predicted environmental data includes: The predicted environmental data is input into the neural network model, the expected operating state is output, and the proportional, integral and differential coefficients are determined according to the current operating state and the expected operating state; The ground source circulation flow control algorithm is constructed using the PID controller, and the expression is: in, for The control signal of the PID controller at this moment, 、 and are the proportional, integral and differential coefficients respectively, for The circulating pump flow rate and the working parameter status of the unit at all times, for The circulating pump flow rate and the working parameter status of the unit at each moment.

10. A control system for an energy-saving ground-source heat pump, for executing the control method for an energy-saving ground-source heat pump according to any one of claims 1 to 9, characterized in that: The system comprises: Dynamic model building module, used to build the dynamic model of the ground source heat pump system, including indoor heat balance model, ground source heat pump unit model and soil heat exchange model; A topology network generation module is used to couple the dynamic model of the ground source heat pump system to obtain a topology network of the ground source heat pump system; A distributed fault-tolerant control module is used to design a distributed control strategy and a fault-tolerant operation strategy according to the structure of the topological network, and each node performs local control according to its position in the topological network and information about adjacent nodes; A state space optimization module is used to construct a state space model of the state variables and control variables in the local control, and solve the future control action sequence through quadratic programming to optimize the balance of the objective functions of energy consumption and indoor and outdoor temperature deviation, thereby obtaining the optimal operating state of the ground source heat pump system; A neural network learning module, configured to learn a mapping relationship between the optimal operating state and corresponding environmental parameters using a neural network; The flow control algorithm module is used to design a ground source circulation flow control algorithm based on the mapping relationship and the predicted environmental data, and automatically adjust the flow distribution of the circulation pump and the working parameters of the ground source heat pump unit.

Images