Edge computing cooperative communication method for multiple heat pumps and photovoltaic system

Abstract

The invention relates to the technical field of building energy conservation, in particular to an edge computing cooperative communication method for multiple heat pumps and a photovoltaic system, which comprises an edge server, a cloud server, multiple heat pumps, a variable frequency pump and an indoor and outdoor heat supply device, and is characterized in that the edge server realizes time synchronization of equipment through an OPC UA over TCSN protocol, a multi-level energy topology network is established, and the multi-level energy topology network is connected with the cloud server; in combination with physical energy flow and information flow, based on a building heat balance equation, an edge server controls a heat pump to work at an energy-saving operation point according to indoor and outdoor temperatures and user requirements, and the flow of a variable frequency pump is adjusted. In addition, multi-mode state sensing is achieved through the distributed sensor network, a layered self-adaptive control strategy is executed, photo-thermal collaborative optimization and heat pump group collaborative control are achieved, the system collaborative control precision is remarkably improved through the method, the time synchronization error is reduced by three orders of magnitude, and the system response speed reaches the millisecond level and is 10 times faster than that of a traditional system.

Description

Technical Field

[0001] The present invention relates to the technical field of building energy conservation, specifically to an edge computing collaborative communication method for multi-heat pump and photovoltaic systems, and particularly to a method for realizing efficient collaborative control of multi-heat pump and photovoltaic systems by using edge computing technology. Background Technique

[0002] With the increasing proportion of building energy consumption in the total energy consumption, building energy conservation has become the focus of global attention. The heat pump system, as an efficient heating device, is widely used in buildings, while the photovoltaic system is a clean renewable energy power generation system. Combining the heat pump and photovoltaic systems can achieve efficient utilization of building energy.

[0003] However, there are many deficiencies in the existing collaborative control technologies for heat pump and photovoltaic systems: First, traditional control systems usually adopt a centralized architecture, with slow system response speed and unable to meet the real-time control requirements; second, the time synchronization accuracy between devices is usually in the millisecond or second level, resulting in insufficient collaborative control accuracy; third, the energy distribution lacks a global optimization perspective, and the overall energy efficiency of the system is low; fourth, the control strategy is fixed and cannot be adaptively adjusted according to environmental changes.

[0004] In addition, in the existing technologies, the heat pump and photovoltaic systems often operate independently, lacking an effective collaborative control mechanism, resulting in the inability of photovoltaic power generation to efficiently supply the heat pump and low overall energy efficiency of the system. At the same time, for multi-heat pump systems, there is a lack of control strategies for load balancing and collaborative operation, further reducing the overall efficiency of the system.

[0005] Therefore, there is an urgent need for a control method that can achieve efficient collaborative operation of multi-heat pump and photovoltaic systems to improve building energy utilization efficiency and reduce building energy consumption. Summary of the Invention

[0006] The purpose of the present invention is to provide an edge computing collaborative communication method for multi-heat pump and photovoltaic systems, and to achieve efficient collaborative control of multi-heat pump and photovoltaic systems through edge computing technology, so as to solve the problems of insufficient collaborative control accuracy of devices, lack of global optimization in energy distribution, fixed control strategies, etc. in the existing technologies.

[0007] The present invention proposes an edge computing collaborative communication method for multi-heat pump and photovoltaic systems, including:

[0008] Constructing a microsecond-level energy collaborative control mechanism based on spatio-temporal topology mapping, including an edge server, a cloud server, multiple heat pumps, multiple variable-frequency pumps, multiple indoor heating devices, and multiple outdoor heating devices;

[0009] The edge server uses the OPC UA over TCSN protocol to achieve microsecond-level time synchronization for all devices. The edge server broadcasts synchronization frames to each slave node. When each slave node receives the broadcast synchronization frame, it completes its own calibration through its own synchronization function to ensure that all devices respond collaboratively within the same clock cycle.

[0010] A multi-level energy topology network is established, which organically combines the physical energy flow and the information flow, including the thermal energy flow topology and the electrical energy flow topology. The thermal energy flow topology is variable frequency pump → heat pump → indoor heating device → building space → outdoor heating device → variable frequency pump, and the electrical energy flow topology is photovoltaic panel → battery → heat pump → photovoltaic panel.

[0011] A decision optimization layer is constructed based on the building heat balance equation. The edge server controls each heat pump to operate at a set of energy-saving operating points according to the collected indoor and outdoor temperatures and user needs, and provides heat for each heat pump by coordinately adjusting the flow rates of each variable frequency pump.

[0012] Multi-modal state perception is achieved. Environmental parameters, device status, and user needs are collected through a distributed sensor network to construct a multi-dimensional state space and track the system evolution trajectory.

[0013] A hierarchical adaptive control strategy is executed. The control parameters are dynamically adjusted according to the position in the state space to achieve collaborative optimization of light and heat and collaborative control of the heat pump group.

[0014] Preferably, the microsecond-level time synchronization specifically includes:

[0015] The edge server uses the TCSN protocol to achieve microsecond-level synchronization of all thermostats in a broadcast form.

[0016] The edge server broadcasts synchronization frames to each slave node. When each slave node receives the broadcast synchronization frame, it completes its own calibration through its own synchronization function, adds the synchronization frame with delayed transmission to the synchronization frame sent to itself, and completes time synchronization according to the synchronization algorithm.

[0017] The time interval is calculated through multiple synchronization frames, and all child nodes of the synchronization node obtain the synchronization interval.

[0018] All child nodes of the synchronization node calculate the time delay required for each child node according to the time difference between each child node and the synchronization node, and all child nodes of the synchronization node complete time synchronization according to the calculated time difference.

[0019] The edge server and the variable frequency pumps, each heat pump, each indoor heating device, and each outdoor heating device achieve synchronous response within the same clock cycle through time synchronization.

[0020] Preferably, the multi-level energy topology network specifically includes:

[0021] Physical energy layer: It includes two sub-networks of heat energy flow and electric energy flow. The heat energy flow flows among each heat pump, indoor heating device and outdoor heating device, and the electric energy flow flows among the photovoltaic system, storage battery and heat pump;

[0022] Information control layer: It includes two sub-networks of control signal flow and status information flow. The control signals flow from the edge server to each execution device, and the status information flows from each sensor to the edge server;

[0023] Decision optimization layer: A decision network constructed based on the building heat balance equation. The inputs include environmental parameters, user demands and device status, and the outputs include control parameters, operation modes and energy distribution strategies;

[0024] Implement the dynamic mapping of the networks of the physical energy layer, the information control layer and the information control layer, so that the physical energy flow and the information control flow can adapt to environmental changes in real time.

[0025] Preferably, the building heat balance equation is specifically: ,

[0026] According to the heat balance equation, by adjusting the temperature of each heat pump, coordinate the heat supply of the heat pump group to meet the heating demand and save the total energy consumption of the building;

[0027] In the formula represents the heat dissipation of the building to the indoor environment; represents the convective heat transfer amount of the indoor fan; represents the difference between the indoor heat gain value and the heat dissipation of the indoor and outdoor air to the outdoor environment through the indoor heat exchanger or the building envelope; represents the indoor heat gain; represents the solar heat gain; represents the heat supply of the heat pump group to the indoor;

[0028] The edge server takes the heat balance equation as the control algorithm of the edge computing system, sets the variable frequency pump and heat pump parameters as input parameters, and takes the control instruction sets of the variable frequency pump and heat pump as output parameters.

[0029] Preferably, the multi-modal state perception specifically includes:

[0030] Environmental state perception: A temperature sensor is installed in each outdoor heating device, a temperature sensor is installed in each heat pump, and a temperature sensor is installed in each indoor heating device; A temperature sensor and a pressure sensor are installed at the inlet of each variable frequency pump in its respective loop for collecting the flow rate and pressure data of the indoor circulation loop;

[0031] Device status perception: Real-time monitoring of the temperature, pressure, power, and efficiency of each heat pump, the power generation, conversion efficiency, and battery SOC of the photovoltaic system, and the flow rate, pressure, and power of the variable-frequency pump;

[0032] User demand perception: The edge server monitors in real-time whether the user is at home and predicts user demand based on user preferences and historical usage patterns.

[0033] Preferably, the hierarchical adaptive control strategy specifically includes:

[0034] Stable domain control: When the system state is within the stable domain of the state space, an energy efficiency priority control strategy is adopted, enabling the photovoltaic system to supply power first, the heat pump group to operate in coordination, and the flow rate of the variable-frequency pump to be precisely controlled;

[0035] Boundary domain control: When the system state is within the boundary domain of the state space, a stability priority control strategy is adopted, increasing the control frequency, expanding the control parameter adjustment range, and enabling standby equipment;

[0036] Mode switching control: According to whether the user is at home, the system operation mode is intelligently switched. When the user is at home, a comfort priority mode is adopted; when the user is not at home, an energy efficiency priority mode is adopted. During the transition period, the user's return time is predicted and the indoor environment is adjusted in advance.

[0037] Preferably, the photothermal collaborative optimization specifically includes:

[0038] The light energy conversion controller is used to control the energy transfer between the photovoltaic system and the battery in real-time. The specific steps include: obtaining the photovoltaic conversion efficiency, conversion time, and loss efficiency; determining whether the photovoltaic panel needs to be charged; if so, controlling the photovoltaic panel to charge the battery, and then determining whether it meets the preset charging requirements; if not, controlling the photovoltaic panel to discharge, and then determining whether it meets the preset discharge requirements;

[0039] During the day, the photovoltaic system directly supplies power to the heat pump, and the excess power is charged into the battery; at night, the battery supplies power to the heat pump to achieve continuous 24-hour energy supply; during the transition period, the photovoltaic electric energy is intelligently distributed to achieve the best energy efficiency ratio;

[0040] The photovoltaic conversion efficiency controller adjusts the photovoltaic conversion efficiency of the photovoltaic panel in real-time according to the power of the photovoltaic panel and provides the converted electric energy to the control device or the heat pump.

[0041] Preferably, the collaborative control of the heat pump group specifically includes:

[0042] Load balancing strategy: Intelligently distribute the load of each heat pump according to the total heat load to ensure that each heat pump operates at the best efficiency point and dynamically adjust the start-stop strategy of the heat pump;

[0043] Precise flow control: The variable-frequency pump is controlled by the edge server, and the flow of the indoor circulation loop is controlled according to the flow control requirements of each outdoor heating device to deliver heat to each indoor heating device; each heat pump inputs heat to the indoor heating device;

[0044] Fault self-healing mechanism: Monitor the operating status of the heat pump in real time to detect potential faults in advance; the faulty heat pump automatically exits the operation, and its load is shared by other heat pumps; the system automatically reconstructs the control strategy to ensure the continuity of heating.

[0045] Preferably, the edge computing system architecture specifically includes:

[0046] Edge server: Deploy a lightweight Kubernetes cluster to coordinate the energy consumption data of the heat pump group and the photovoltaic system; process local data and execute control algorithms;

[0047] Cloud server: Installed in the gateway to provide cloud storage support for the edge server; store the edge computing applications in the edge server;

[0048] Communication structure: The edge server communicates with each variable-frequency pump, each heat pump, each indoor heating device, and each outdoor heating device through the fieldbus and uses the OPC UA protocol for communication;

[0049] Data processing flow: The collected indoor and outdoor temperatures and user requirements are transmitted to the edge server through the fieldbus. The edge server calculates the heat balance equation through edge computing based on these data, and at the same time obtains the output parameters according to the heat balance equation. The output parameters include the frequency of the variable-frequency pump and the heat supply of the heat pump. The edge server transmits these output parameters to the variable-frequency pump and the heat pump to realize the adjustment of the indoor heat supply of the building.

[0050] Preferably, the edge computing collaborative communication method for the multi-heat pump and photovoltaic system specifically includes:

[0051] a) Use a wireless heat pump group, including several indoor unit nodes and several outdoor unit nodes, and data can be transmitted between adjacent indoor unit nodes, between indoor unit nodes and outdoor units, and between outdoor unit nodes; both indoor unit nodes and outdoor unit nodes are built-in with wireless communication modules, and the communication modules include Ethernet, USB, and RF;

[0052] b) Based on the wireless heat pump group, establish an indoor unit and outdoor unit communication architecture. The communication architecture includes an edge host and an edge extension. The communication methods between the edge host and the edge extension include: wifi interaction between the edge host and the edge extension, and serial communication between the edge host and the edge extension using the RS485 bus;

[0053] c) After each node collects energy consumption data, it is aggregated to the edge host through the gateway. The edge host integrates and packages the energy consumption data of each node with the current time data; using the UDP protocol within the gateway, the above-integrated and packaged data is sent to the server;

[0054] d) The edge host pushes the packaged data to the edge extension. After receiving the packaged data, the edge extension decompresses it to obtain the energy consumption data and current time data of each node, parses the data according to the communication protocol, and automatically calibrates the time after parsing the packaged data to the current time data according to the local clock, and issues control instructions to each node directly through Ethernet.

[0055] The technical solution provided by the present invention has the following beneficial effects:

[0056] 1. Through the OPC UA over TCSN protocol, microsecond-level time synchronization is achieved, significantly improving the system's collaborative control accuracy, reducing the time synchronization error by 3 orders of magnitude, and the system response speed reaching the millisecond level, which is 10 times faster than the traditional system;

[0057] 2. A multi-level energy topology network is constructed, realizing the global optimization of heat energy flow and electric energy flow, avoiding the energy island phenomenon caused by local optimization in the traditional scheme, and increasing the system's comprehensive energy efficiency ratio by 26.5%;

[0058] 3. Based on the decision-making optimization layer of the building heat balance equation, the collaborative control of the heat pump group is achieved, reducing the building heating energy consumption by more than 30%, and improving the temperature control accuracy to ±0.5°C;

[0059] 4. Through multi-modal state perception and hierarchical adaptive control strategies, the system can dynamically adjust the operation mode according to environmental changes and user needs, improving the system operation stability by 80%, and at the same time increasing the photovoltaic self-use rate to more than 85%;

[0060] 5. Adopting a wireless heat pump group architecture, combined with the collaborative work of the edge host and the edge extension, reduces the complexity of wired connections and improves the flexibility and scalability of system deployment. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 is the system architecture diagram of the edge computing collaborative communication method of the multi-heat pump and photovoltaic system of the present invention;

[0062] Figure 2 is the flowchart of the microsecond-level time synchronization of the present invention;

[0063] Figure 3 is the structure diagram of the multi-level energy topology network of the present invention;

[0064] Figure 4 is the decision-making optimization flowchart based on the building heat balance equation of the present invention;

[0065] Figure 5 is the data flow diagram of the multi-modal state perception of the present invention;

[0066] Figure 6 is the state transition diagram of the hierarchical adaptive control strategy of the present invention;

[0067] Figure 7 is the control flow diagram of the photothermal collaborative optimization of the present invention;

[0068] Figure 8 is the strategy diagram of the collaborative control of the heat pump group of the present invention;

[0069] Figure 9 is the architecture diagram of the edge computing system of the present invention;

[0070] Figure 10 is the communication architecture diagram of the wireless heat pump group of the present invention. Detailed implementation manners

[0071] Referring to Figure 1 , the present invention provides a method for edge computing collaborative communication between multiple heat pumps and a photovoltaic system. This method constructs a microsecond-level energy collaborative control mechanism based on spatio-temporal topology mapping, including an edge server 100, a cloud server 200, multiple heat pumps 300, multiple variable-frequency pumps 400, multiple indoor heating devices 500, and multiple outdoor heating devices 600.

[0072] In an embodiment of the present invention, the edge server 100 realizes microsecond-level time synchronization of all devices through the OPC UA over TCSN protocol. The system establishes a multi-level energy topology network, which organically combines the physical energy flow and the information flow, including the heat energy flow topology and the electric energy flow topology. The edge server 100 constructs a decision optimization layer based on the building heat balance equation, and controls each heat pump 300 to operate at a set of energy-saving operating points according to the collected indoor and outdoor temperatures and user requirements.

[0073] In addition, the present invention realizes multi-modal state perception. It collects environmental parameters, device states, and user requirements through a distributed sensor network, constructs a multi-dimensional state space, and tracks the system evolution trajectory. The system executes a hierarchical adaptive control strategy, dynamically adjusts control parameters according to the position in the state space, and realizes photothermal collaborative optimization and heat pump group collaborative control.

[0074] Referring to Figure 2, in a preferred embodiment of the present invention, the specific implementation of microsecond-level time synchronization is as follows: The edge server 100 adopts the TCSN protocol to achieve microsecond-level synchronization of all thermostats in a broadcast manner. The edge server 100 broadcasts a synchronization frame to each slave node. When each slave node receives the broadcast synchronization frame, it completes its own calibration through its own synchronization function, adds the synchronization frame with a delayed transmission to the synchronization frame sent to itself, and completes time synchronization according to the synchronization algorithm.

[0075] In an embodiment of the present invention, the time synchronization process is achieved through the following steps: First, calculate the time interval through multiple synchronization frames, and all child nodes of the synchronization node obtain the synchronization interval. Then, all child nodes of the synchronization node calculate the time delay required for each child node according to the time difference between each child node and the synchronization node, and all child nodes of the synchronization node complete time synchronization according to the calculated time difference. Finally, the edge server 100 and the variable-frequency pump 400, each heat pump 300, each indoor heating device 500, and each outdoor heating device 600 ensure synchronous response within the same clock cycle through time synchronization.

[0076] The time difference calculation in the synchronization algorithm can be expressed as:

[0077] ,

[0078] where, is the time difference of the i-th child node, with the unit of microsecond (μs); is the timestamp when the i-th child node receives the synchronization frame, with the unit of microsecond (μs); is the timestamp when the synchronization node sends the synchronization frame, with the unit of microsecond (μs); is the propagation delay.

[0079] Preferably, the propagation delay can be estimated by round-trip time measurement:

[0080] ,

[0081] where, is the round-trip time, that is, the time interval from when the synchronization node sends a test frame to when it receives a response frame.

[0082] In practical applications, the system usually sets the synchronization period to 10 ms, which can be shortened to 1 ms when higher-precision control is required, and can be extended to 100 ms when the system is running stably to reduce the communication burden. Through this dynamic adjustment mechanism, the system can not only ensure high-precision synchronization at critical moments but also optimize the use of network resources.

[0083] Referring to Figure 3, in an embodiment of the present invention, the multi-level energy topology network specifically includes three levels: the physical energy layer, the information control layer, and the decision optimization layer.

[0084] The physical energy layer includes two sub-networks: the heat energy flow network and the electric energy flow network. The heat energy flow circulates among the heat pumps 300, the indoor heating device 500, and the outdoor heating device 600, forming a closed-loop heat transfer path. Specifically, the heat energy flow topology is: variable-frequency pump 400 → heat pump 300 → indoor heating device 500 → building space → outdoor heating device 600 → variable-frequency pump 400. The electric energy flow circulates among the photovoltaic system 700, the storage battery 800, and the heat pump 300, forming an electric energy transfer path. Specifically, the electric energy flow topology is: photovoltaic panel 701 → storage battery 800 → heat pump 300 → photovoltaic panel 701.

[0085] The information control layer includes two sub-networks: the control signal flow network and the status information flow network. The control signals flow from the edge server 100 to each execution device, including the variable-frequency pump 400, the heat pump 300, the indoor heating device 500, and the outdoor heating device 600. The status information flows from each sensor (including temperature sensors and pressure sensors) to the edge server 100 and then to the cloud server 200.

[0086] The decision optimization layer constructs a decision network based on the building heat balance equation. The inputs include environmental parameters, user requirements, and device status, and the outputs include control parameters, operation modes, and energy distribution strategies. The edge server 100 controls the operation status of each device according to the decision optimization results to achieve the global optimization of the system.

[0087] The above three layers of networks are interconnected through a dynamic mapping mechanism, enabling the physical energy flow and the information control flow to adapt to environmental changes in real time. For example, when the user requirements change, the decision optimization layer adjusts the topology weights of the heat energy flow; when the light intensity changes, the system adjusts the topology structure of the electric energy flow; when a device fails, the system automatically reconstructs the topology to bypass the fault point.

[0088] In an embodiment of the present invention, the edge server 100 uses the building heat balance equation as the control algorithm of the edge computing system, sets the parameters of the variable-frequency pump 400 and the heat pump 300 as input parameters, and uses the control instruction sets of the variable-frequency pump 400 and the heat pump 300 as output parameters.

[0089] The building heat balance equation is specifically:

[0090] ,

[0091] where represents the heat dissipation from the building to the indoor environment, with the unit of watt (W); represents the convective heat transfer of the indoor fan, with the unit of watt (W); represents the difference between the heat indoors and the heat dissipation from the indoor and outdoor air to the outdoor environment through the indoor heat exchanger or the building envelope, with the unit of watt (W); represents the heat gain indoors, with the unit of watt (W); represents the solar heat gain, with the unit of watt (W); represents the heat supply from the heat pump unit to the indoors, with the unit of watt (W).

[0092] Convective heat transfer of the indoor fan can be expressed as:

[0093] ,

[0094] where, is the specific heat capacity of air, approximately 1005 J / (kg·°C); is the air density, approximately 1.2 kg / m³; is the air volume flow rate of the fan, with the unit of m³ / s; is the difference between the supply air temperature and the return air temperature, with the unit of °C.

[0095] Heat indoors can be expressed as:

[0096] ,

[0097] where, is the heat dissipation per person, generally about 100 W per person; is the heat dissipation of equipment; is the heat dissipation of lighting, with the unit of watt (W).

[0098] Heat from the sun can be expressed as:

[0099] ,

[0100] where, is the number of transparent envelopes; is the th area of the transparent envelope (such as a window), with the unit of square meter (m²); is the th solar radiation intensity received by the transparent envelope, with the unit of watt per square meter (W / m²); is the th solar heat gain coefficient of the transparent envelope, dimensionless, and the value range is usually 0 - 1.

[0101] Heat supply from the heat pump unit to the indoors can be expressed as:

[0102] ,

[0103] Wherein: is the number of heat pumps; is the rated heat supply of the th heat pump, with the unit of watt (W); is the actual operating efficiency of the th heat pump, dimensionless, and its value range is usually 0 - 1, affected by factors such as ambient temperature and load rate.

[0104] Preferably, the edge server 100 coordinates the heat supply of the heat pump group to meet the heating demand and save the total energy consumption of the building by adjusting the temperature of each heat pump 300 according to the building heat balance equation. Specifically, when the building heat load increases, the system increases the heat supply of the heat pump 300; when the building heat load decreases, the system reduces the heat supply of the heat pump 300; when the photovoltaic power generation is sufficient, the system preferentially uses the electric heat pump for heating; when the photovoltaic power generation is insufficient, the system optimizes the operating efficiency of the heat pump 300 to maximize the utilization of limited electric energy.

[0105] Referring to Figure 5 , in an embodiment of the present invention, the multi-modal state perception specifically includes three aspects: environmental state perception, device state perception, and user demand perception.

[0106] The environmental state perception collects multi-dimensional environmental parameters through a distributed sensor network. Specifically, a temperature sensor is installed in each outdoor heating device 600, a temperature sensor is installed in each heat pump 300, and a temperature sensor is installed in each indoor heating device 500; a temperature sensor and a pressure sensor are installed at the inlet of each variable-frequency pump 400 in its respective loop to collect the flow rate and pressure data of the indoor circulation loop. These sensors collect data at a frequency of 10Hz and transmit the data to the edge server 100 through the fieldbus.

[0107] The device state perception monitors the operating parameters and health status of each device in real time. The system monitors parameters such as the temperature, pressure, power, and efficiency of the heat pump 300, the power generation power, conversion efficiency of the photovoltaic system 700, and the SOC (state of charge) of the battery 800, and the flow rate, pressure, and power of the variable-frequency pump 400. These parameters are transmitted to the edge server 100 through the OPC UA protocol, and the sampling frequency is 1Hz.

[0108] The user demand perception monitors in real time whether the user is at home through the edge server 100 and predicts the user demand according to the user preferences and historical usage patterns. The system constructs a user preference model by analyzing the user's historical energy consumption behavior to predict the user's temperature setting demand and energy consumption time. For example, the system can predict the time when the user returns home every day and adjust the indoor temperature in advance, which not only ensures the user's comfort but also avoids energy waste.

[0109] Based on multi-modal perception data, the system constructs a multi-dimensional state space and tracks the evolution trajectory of the system in the state space. The state space includes an environmental dimension (indoor and outdoor temperature, humidity, light intensity), a device dimension (operating efficiency, energy consumption, remaining life of each device), and a user dimension (comfort requirements, at-home status, usage patterns). The system predicts the state evolution trend based on historical data and identifies stable and unstable regions in the state space to provide a decision-making basis for the hierarchical adaptive control strategy.

[0110] Refer to Figure 6 , in an embodiment of the present invention, the hierarchical adaptive control strategy specifically includes three levels: stable region control, boundary region control, and mode switching control.

[0111] Stable region control is enabled when the system state is in the stable region of the state space, and an energy efficiency priority control strategy is adopted. At this time, the system gives priority to power supply by the photovoltaic system 700, the heat pump group 300 operates in coordination, and the flow rate of the variable frequency pump 400 is precisely controlled. Specifically, when the light is sufficient, the photovoltaic system 700 directly supplies power to the heat pump 300, and the excess power is stored in the battery 800; when the loads of each heat pump 300 are balanced, the system makes each heat pump 300 operate at its highest efficiency point; the flow rate of the variable frequency pump 400 is precisely adjusted according to the demands of each indoor heating device 500 to avoid energy waste.

[0112] Boundary region control is enabled when the system state is in the boundary region of the state space, and a stability priority control strategy is adopted. At this time, the system increases the control frequency from the standard 1 Hz to 5 Hz; expands the control parameter adjustment range to allow more aggressive parameter adjustments to cope with the rapidly changing environment; enables standby devices to improve the system redundancy. For example, when the outdoor temperature drops sharply, the system starts more heat pumps 300 to ensure heating stability; when the light intensity fluctuates violently, the system adjusts the operating parameters of the photovoltaic system 700 more frequently to ensure stable power supply.

[0113] Mode switching control intelligently switches the system operating mode according to whether the user is at home. When the user is at home, the system adopts a comfort priority mode and precisely controls the indoor temperature within the comfortable range set by the user (usually 22±1°C); when the user is not at home, the system adopts an energy efficiency priority mode and allows the indoor temperature to fluctuate within a wider range (usually 18 - 26°C) to maximize energy savings; during the transition period (such as 1 hour before the user is about to return home), the system predicts the user's return time and adjusts the indoor environment in advance to ensure that the indoor temperature reaches the comfortable temperature when the user returns home.

[0114] Preferably, the system dynamically adjusts the parameters of the control strategy according to the state - space position. For example, in the core area of the stable domain, the control period can be extended to 5 minutes; in the edge area of the stable domain, the control period is shortened to 1 minute; in the boundary domain, the control period is further shortened to 10 seconds to ensure that the system can quickly respond to environmental changes.

[0115] Referring Figure 7 , in an embodiment of the present invention, the photothermal co - optimization is specifically implemented by a light - energy conversion controller, which is used to control the energy transfer between the photovoltaic system 700 and the battery 800 in real - time.

[0116] The specific steps of the light - energy conversion controller include: obtaining the photovoltaic conversion efficiency, conversion time, and loss efficiency; determining whether the photovoltaic panel 701 needs to be charged; if so, controlling the photovoltaic panel 701 to charge the battery 800, and then determining whether it meets the preset charging requirements; if not, controlling the photovoltaic panel 701 to discharge, and then determining whether it meets the preset discharge requirements.

[0117] During the day, the photovoltaic system 700 directly supplies power to the heat pump 300, and the excess power is charged into the battery 800. Preferably, when the light intensity is greater than 300 W / m², the photovoltaic system 700 can generate enough electrical energy to supply both the heat pump 300 and charge the battery 800. At night, the battery 800 supplies power to the heat pump 300 to achieve continuous 24 - hour energy supply. Preferably, the capacity of the battery 800 should be designed to support the night operation of the heat pump 300 for 8 hours, usually 10 times the rated power of the heat pump 300. During the transition period (such as dawn and dusk), the system intelligently distributes the photovoltaic electrical energy to achieve the best energy efficiency ratio.

[0118] The photovoltaic conversion efficiency controller adjusts the photovoltaic conversion efficiency of the photovoltaic panel 701 in real - time according to its power, and provides the converted electrical energy to the control device or the heat pump 300. Specifically, the controller optimizes the output power of the photovoltaic system 700 by adjusting the parameters of the maximum power point tracking (MPPT) algorithm of the inverter. The MPPT algorithm can be expressed as:

[0119] ,

[0120] Where is the output power of the photovoltaic system, in watts (W); is the output voltage, in volts (V); represents the derivative of power with respect to voltage. When this derivative is 0, it means that the system is operating at the maximum power point.

[0121] In practical applications, the photovoltaic conversion efficiency is significantly affected by temperature and can usually be expressed as:

[0122] ,

[0123] Wherein: is the actual conversion efficiency, dimensionless, and its value range is usually 0.10 - 0.25; is the conversion efficiency at the reference temperature (usually 0.15 - 0.22 at 25°C), dimensionless; is the temperature coefficient (usually 0.003 - 0.005 / °C), with the unit of 1 / °C; is the actual temperature of the photovoltaic panel, with the unit of °C; is the reference temperature (usually 25°C).

[0124] The light energy conversion controller also predicts the power generation of the photovoltaic panel 701 at the next moment according to the light intensity, indoor heat supply, and photovoltaic conversion efficiency at historical moments. The prediction model can be expressed as:

[0125] ,

[0126] Wherein: is the predicted power generation, with the unit of watt (W); is the solar radiation intensity, with the unit of watt per square meter (W / m²); is the temperature of the photovoltaic panel, with the unit of °C; is the historical conversion efficiency data set, dimensionless; is the historical power generation data set, with the unit of watt (W); is the prediction function, which can be implemented by methods such as neural networks.

[0127] Referring to Figure 8 , in an embodiment of the present invention, the cooperative control of the heat pump group specifically includes three aspects: load balancing strategy, precise flow control, and fault self-healing mechanism.

[0128] The load balancing strategy intelligently distributes the loads of each heat pump 300 according to the total heat load, ensures that each heat pump 300 operates at the best efficiency point, and dynamically adjusts the start-stop strategy of the heat pump 300. Specifically, the edge computing system controls each heat pump 300 to operate at a set of energy-saving operation points according to the collected indoor and outdoor temperatures and user requirements, and provides heat for each heat pump 300 by coordinately adjusting the flow rates of each variable-frequency pump 400.

[0129] The optimal load rate of the heat pump 300 is usually between 60% and 80%, at which time the coefficient of performance (COP) of the heat pump 300 is the highest. The system calculates the optimal combination of heat pumps 300 and the load rate of each heat pump 300 based on the total heat load and the rated capacity of each heat pump 300. For example, when the total heat load is 100 kW and there are 4 heat pumps 300 with a rated capacity of 30 kW each, the system will choose to start 4 heat pumps 300, with each load being approximately 83%, rather than starting 3 heat pumps 300, with each load being approximately 111% (overloading operation will reduce efficiency and lifespan).

[0130] Precise flow control is achieved through the variable-frequency pump 400. The variable-frequency pump 400 is controlled by the edge server 100, and controls the flow rate of the indoor circulation loop according to the flow control requirements of each outdoor heating device 600, and delivers heat supply to each indoor heating device 500. Each heat pump 300 inputs heat supply to the indoor heating device 500. The indoor heating device 500 is connected to the indoor pipeline and delivers air to each area indoors for heating.

[0131] The flow control of the variable-frequency pump 400 is based on the following relationship:

[0132] ,

[0133] where, is the flow rate, with the unit of cubic meters per hour (m3 / h); is the rotational speed of the pump, with the unit of revolutions per minute (r / min); is the impeller diameter, with the unit of meters (m).

[0134] Since the impeller diameter is fixed, the flow rate is mainly controlled by adjusting the rotational speed. The relationship between the power of the variable-frequency pump 400 and the rotational speed is:

[0135] ,

[0136] where, is the actual power, with the unit of watts (W); is the rated power, with the unit of watts (W); is the actual rotational speed, with the unit of revolutions per minute (r / min); is the rated rotational speed, with the unit of revolutions per minute (r / min); The exponent 3 represents the cubic relationship between power and rotational speed, which is one of the similarity laws of pumps.

[0137] The fault self-healing mechanism monitors the operating status of the heat pump 300 in real time to detect potential faults in advance; the faulty heat pump 300 automatically exits the operation, and its load is shared by other heat pumps 300; the system automatically reconstructs the control strategy to ensure the continuity of heat supply. Specifically, the system monitors parameters such as the compressor current, exhaust temperature, evaporation pressure, and condensation pressure of each heat pump 300. When these parameters deviate from the normal range by more than a preset threshold (such as current deviation > 20%, temperature deviation > 15°C), the system will determine that the heat pump 300 may have a fault, automatically remove it from the operation sequence, and adjust the load distribution of other heat pumps 300.

[0138] Preferably, the system adopts a predictive maintenance strategy to predict possible faults based on the operating data of the heat pump 300 and perform maintenance in advance to avoid faults. For example, when the compressor current continues to rise but the heating capacity does not increase, it may indicate a decrease in the compressor efficiency, and the system will issue a warning message and recommend a maintenance inspection.

[0139] Refer to Figure 9 , in an embodiment of the present invention, the edge computing system architecture specifically includes four aspects: an edge server 100, a cloud server 200, a communication structure, and a data processing flow.

[0140] The edge server 100 deploys a lightweight Kubernetes cluster to coordinate the energy consumption data of the heat pump group and the photovoltaic system; processes local data and executes control algorithms. The Kubernetes cluster includes a master node and multiple worker nodes. The master node is responsible for cluster management and task scheduling, and the worker nodes execute specific computing tasks. The hardware configuration of the edge server 100 is usually a multi-core CPU (such as 8 cores), a large amount of memory (such as 16GB), and sufficient storage space (such as 256GB SSD) to meet the requirements of edge computing.

[0141] The cloud server 200 is installed in the gateway to provide cloud storage support for the edge server 100; stores the edge computing applications in the edge server 100. The cloud server 200 is mainly responsible for the long-term storage and analysis of data, as well as the remote management and update of the edge server 100. The cloud server 200 communicates with the edge server 100 through an encrypted network connection to ensure the security of data transmission.

[0142] In terms of the communication structure, the edge server 100 communicates with each variable-frequency pump 400, each heat pump 300, each indoor heating device 500, and each outdoor heating device 600 through a fieldbus and uses the OPC UA protocol for communication. The fieldbus uses industrial Ethernet to support high-speed data transmission (above 100Mbps) and real-time control. The OPC UA protocol provides a unified data access interface and supports cross-platform and cross-vendor device communication.

[0143] In terms of the data processing flow, the indoor and outdoor temperatures and user requirements collected are transmitted to the edge server 100 through the fieldbus. The edge server 100 calculates the heat balance equation through edge computing based on this data, and at the same time obtains output parameters according to the heat balance equation. The output parameters include the frequency of the variable-frequency pump 400 and the heat supply of the heat pump 300. The edge server 100 transmits these output parameters to the variable-frequency pump 400 and the heat pump 300 to achieve the adjustment of the indoor heat supply of the building.

[0144] Preferably, the edge server 100 adopts a hierarchical data processing architecture: the data acquisition layer is responsible for obtaining raw data from each sensor; the data preprocessing layer performs data cleaning, filtering, and standardization; the data analysis layer executes edge computing algorithms to generate control decisions; the data storage layer caches key data locally and synchronizes it to the cloud server 200 regularly; the control execution layer converts the decisions into device control commands and monitors the execution results.

[0145] Referring to Figure 10 , in an embodiment of the present invention, the edge computing collaborative communication method between the multi-heat pump and the photovoltaic system specifically includes two core steps: using a wireless heat pump group and establishing a communication architecture.

[0146] Using a wireless heat pump group includes several indoor unit nodes and several outdoor unit nodes, and data can be transmitted between adjacent indoor unit nodes, between indoor unit nodes and outdoor units, and between outdoor unit nodes. Both the indoor unit nodes and the outdoor unit nodes are built-in with wireless communication modules, and the communication modules include Ethernet, USB, and RF. Preferably, the RF module uses the 2.4GHz frequency band, with a communication distance of up to 100 meters, supporting a mesh network topology, which improves the reliability and coverage of communication.

[0147] Based on the wireless heat pump group, an indoor unit and outdoor unit communication architecture is established. The communication architecture includes an edge host and an edge extension. The communication methods between the edge host and the edge extension include: Wi-Fi interaction between the edge host and the edge extension, and serial communication between the edge host and the edge extension using the RS485 bus. The Wi-Fi interaction adopts the 802.11n standard, with a rate of up to 300Mbps, suitable for large data volume transmission; the RS485 bus communication rate is relatively low (usually 9600bps - 115200bps), but it has strong anti-interference ability and is suitable for industrial environments.

[0148] After each node collects energy consumption data, it is aggregated to the edge host through the gateway. The edge host integrates and packages the energy consumption data of each node with the current time data. Preferably, the data packaging uses the JSON format, which includes parameters such as node ID, timestamp, temperature, pressure, power, current, and voltage. Using the UDP protocol within the gateway, the above-integrated and packaged data is sent to the server. The UDP protocol has the characteristic of low latency and is suitable for real-time control scenarios, but it does not guarantee the reliability of data transmission. Therefore, the system will implement a data confirmation mechanism at the application layer.

[0149] The edge host pushes the packaged data to the edge extension. After receiving the packaged data, the edge extension decompresses it to obtain the energy consumption data and current time data of each node, parses the data according to the communication protocol, and automatically calibrates the time after parsing the packaged data to the current time data according to the local clock, and issues control instructions to each node directly through Ethernet. The control instructions include start / stop commands for the heat pump 300, compressor frequency adjustment, electronic expansion valve opening adjustment, fan speed adjustment, etc.

[0150] Preferably, the edge extension will also perform data validity checks and filter out abnormal data. For example, data with temperature and pressure outside the normal range (such as temperature < -30°C or > 100°C, pressure < 0.1 MPa or > 4.5 MPa) is marked as invalid. At the same time, the edge extension will smooth the data to eliminate the impact of instantaneous fluctuations on control decisions.

[0151] Through the above wireless heat pump group communication architecture, the system realizes flexible communication and efficient collaborative control between devices, providing strong support for the collaborative operation of multiple heat pumps and photovoltaic systems.

[0152] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. The protection scope of the present invention is subject to the claims. Those skilled in the art can make various deformations and improvements within the scope of the substantial content of the present invention, and these deformations and improvements should also be regarded as the protection scope of the present invention.

Claims

1. Edge computing collaborative communication method for multiple heat pumps and photovoltaic systems, characterized in that: include: Construct a microsecond-level energy collaborative control mechanism based on spatiotemporal topological mapping, including edge servers, cloud servers, multiple heat pumps, multiple variable frequency pumps, multiple indoor heating devices, and multiple outdoor heating devices; The OPC UA over TCSN protocol is used by the edge server to achieve microsecond-level time synchronization of all devices. The edge server broadcasts synchronization frames to each slave node. When each slave node receives the broadcast synchronization frame, it completes its own calibration through its own synchronization function to ensure that all devices respond in coordination within the same clock cycle. Establish a multi-level energy topology network to organically combine physical energy flow with information flow, including thermal energy flow topology and electrical energy flow topology. The thermal energy flow topology is variable frequency pump → heat pump → indoor heating device → building space → outdoor heating device → variable frequency pump, and the electrical energy flow topology is photovoltaic panel → battery → heat pump → photovoltaic panel; A decision optimization layer is constructed based on the building heat balance equation, where the edge server controls each heat pump to operate at a set series of energy-saving operating points according to the collected indoor and outdoor temperatures and user needs, and provides heat for each heat pump by coordinating and adjusting the flow of each variable frequency pump; Realize multi-modal state perception, collect environmental parameters, equipment status and user needs through distributed sensor networks, build multi-dimensional state space and track system evolution trajectory; A hierarchical adaptive control strategy is implemented to dynamically adjust control parameters according to the state space position to achieve solar-thermal coordinated optimization and coordinated control of heat pump groups.

2. The edge computing collaborative communication method of multiple heat pumps and photovoltaic systems according to claim 1 is characterized in that: The microsecond-level time synchronization specifically includes: The edge server uses the TCSN protocol to achieve microsecond synchronization of all thermostats in the form of broadcasting; The edge server broadcasts synchronization frames to each slave node. When each slave node receives the broadcast synchronization frame, it completes its own calibration through its own synchronization function, adds the delayed synchronization frame to the synchronization frame sent to itself, and completes time synchronization according to the synchronization algorithm; The time interval is calculated through multiple synchronization frames, and all child nodes of the synchronization node obtain the synchronization interval; All child nodes of the synchronization node calculate the time required for each child node to be delayed based on the time difference between each child node and the synchronization node, and all child nodes of the synchronization node complete time synchronization based on the calculated time difference; The edge server and the variable frequency pump, each heat pump, each indoor heating device, and each outdoor heating device are synchronized through time synchronization to ensure synchronous response within the same clock cycle.

3. The edge computing collaborative communication method of multiple heat pumps and photovoltaic systems according to claim 1 is characterized in that: The multi-level energy topology network specifically includes: Physical energy layer: It includes two sub-networks: thermal energy flow and electrical energy flow. The thermal energy flow flows between each heat pump, indoor heating device and outdoor heating device, and the electrical energy flow flows between the photovoltaic system, battery and heat pump. Information control layer: It includes two sub-networks: control signal flow and status information flow. The control signal flows from the edge server to each execution device, and the status information flows from each sensor to the edge server. Decision optimization layer: A decision network built based on the building heat balance equation. The input includes environmental parameters, user needs and equipment status, and the output includes control parameters, operation mode and energy allocation strategy. The physical energy layer, the information control layer and the network dynamic mapping of the information control layer are realized, so that the physical energy flow and the information control flow can adapt to environmental changes in real time.

4. The edge computing collaborative communication method of multiple heat pumps and photovoltaic systems according to claim 1 is characterized in that: The building heat balance equation is specifically: , According to the heat balance equation, by adjusting the temperature of each heat pump, the heat supply of the heat pump group is coordinated to meet the heating demand and save the total energy consumption of the building; In the formula Indicates the heat dissipation from the building to the indoor environment; Indicates the convective heat transfer of the indoor fan; It indicates the difference between the indoor heat gain and the heat dissipation of indoor and outdoor air to the outdoor environment through the indoor heat exchanger or enclosure structure; Indicates indoor heat gain; Indicates the amount of heat gained by the sun; Indicates the amount of heat supplied by the heat pump unit to the room; The edge server uses the heat balance equation as the control algorithm of the edge computing system, sets the frequency conversion pump and heat pump parameters as input parameters, and uses the control instruction set of the frequency conversion pump and heat pump as output parameters.

5. The edge computing collaborative communication method of multiple heat pumps and photovoltaic systems according to claim 1, characterized in that: The multimodal state perception specifically includes: Environmental status perception: A temperature sensor is installed in each outdoor heating device, a temperature sensor is installed in each heat pump, and a temperature sensor is installed in each indoor heating device; each variable frequency pump is installed with a temperature sensor and a pressure sensor at the entrance of its own loop to collect flow and pressure data of the indoor circulation loop; Equipment status perception: real-time monitoring of the temperature, pressure, power, and efficiency of each heat pump, the power generation power, conversion efficiency, and battery SOC of the photovoltaic system, and the flow, pressure, and power of the variable frequency pump; User demand perception: The edge server monitors whether the user is at home in real time and predicts user demand based on user preferences and historical usage patterns.

6. The edge computing collaborative communication method of multiple heat pumps and photovoltaic systems according to claim 1 is characterized in that: The hierarchical adaptive control strategy specifically includes: Stable domain control: When the system state is in the stable domain of the state space, an energy efficiency priority control strategy is adopted to give priority to photovoltaic power supply, the heat pump group operates in coordination, and the variable frequency pump flow is precisely controlled; Boundary domain control: When the system state is located in the boundary domain of the state space, a stability-first control strategy is adopted to increase the control frequency, expand the control parameter adjustment range, and enable backup equipment; Mode switching control: Intelligently switch the system operation mode according to whether the user is at home. When the user is at home, the comfort priority mode is adopted. When the user is not at home, the energy efficiency priority mode is adopted. In the transition period, the user's return home time is predicted and the indoor environment is adjusted in advance.

7. The edge computing collaborative communication method of multiple heat pumps and photovoltaic systems according to claim 1, characterized in that: The photothermal synergistic optimization specifically includes: The light energy conversion controller is used to control the energy transfer between the photovoltaic system and the battery in real time. The specific steps include: obtaining the photovoltaic conversion efficiency, conversion time and consumption efficiency; judging whether the photovoltaic panel needs to be charged; if so, executing the control of the photovoltaic panel to charge the battery, and then judging whether it meets the preset charging requirements; if not, executing the control of the photovoltaic panel to discharge, and then judging whether it meets the preset discharge requirements; During the day, the photovoltaic system directly supplies power to the heat pump, and the excess power is charged into the battery; at night, the battery supplies power to the heat pump, achieving 24-hour continuous energy supply; during the transition period, the photovoltaic power is intelligently distributed to achieve the best energy efficiency ratio; The photovoltaic conversion efficiency controller adjusts the photovoltaic conversion efficiency in real time according to the power of the photovoltaic panel and provides the converted electrical energy to the control device or heat pump.

8. The edge computing collaborative communication method of multiple heat pumps and photovoltaic systems according to claim 1, characterized in that: The heat pump group coordinated control specifically includes: Load balancing strategy: intelligently distribute the load of each heat pump according to the total heat load, ensure that each heat pump works at the optimal efficiency point, and dynamically adjust the start and stop strategy of the heat pump; Precise flow control: The variable frequency pump is controlled by the edge server, and the flow of the indoor circulation loop is controlled according to the flow control requirements of each outdoor heating device to deliver heat to each indoor heating device; each heat pump inputs heat to the indoor heating device; Fault self-healing mechanism: Real-time monitoring of the operating status of the heat pump to detect potential faults in advance; the faulty heat pump automatically exits operation and its load is shared by other heat pumps; the system automatically reconstructs the control strategy to ensure continuity of heating.

9. The edge computing collaborative communication method of multiple heat pumps and photovoltaic systems according to claim 4, characterized in that: The architecture of the edge computing system specifically includes: Edge server: deploys lightweight Kubernetes clusters to coordinate energy consumption data of heat pump clusters and photovoltaic systems; processes local data and executes control algorithms; Cloud server: installed in the gateway to provide cloud storage support for edge servers; stores edge computing applications in edge servers; Communication structure: The edge server communicates with each variable frequency pump, each heat pump, each indoor heating device, and each outdoor heating device through the field bus and adopts the OPC UA protocol; Data processing flow: The collected indoor and outdoor temperatures and user demands are transmitted to the edge server through the field bus. The edge server calculates the heat balance equation based on these data through edge computing, and derives the output parameters based on the heat balance equation. The output parameters include the frequency of the variable frequency pump and the heating amount of the heat pump. The edge server transmits these output parameters to the variable frequency pump and the heat pump to adjust the indoor heating amount of the building.

10. The edge computing collaborative communication method of multiple heat pumps and photovoltaic systems according to claim 1, characterized in that: The edge computing collaborative communication method of the multi-heat pump and photovoltaic system specifically includes: a) Use a wireless heat pump group, including several indoor unit nodes and several outdoor unit nodes. Data can be transmitted between adjacent indoor unit nodes, indoor unit nodes and outdoor unit nodes, and outdoor unit nodes. Both indoor unit nodes and outdoor unit nodes are equipped with built-in wireless communication modules, including Ethernet, USB, and RF. b) Based on the wireless heat pump group, a communication architecture between the indoor unit and the outdoor unit is established, wherein the communication architecture includes an edge host and an edge extension, and the communication methods between the edge host and the edge extension include: WiFi interaction between the edge host and the edge extension, and serial communication between the edge host and the edge extension using an RS485 bus; c) After collecting energy consumption data from each node, it is collected and sent to the edge host through the gateway. The edge host integrates and packages the energy consumption data of each node with the current time data; and uses the UDP protocol in the gateway to send the above integrated and packaged data to the server; d) The edge host pushes the packaged data to the edge extension. After receiving the packaged data, the edge extension decompresses it to obtain the energy consumption data and current time data of each node, parses the data according to the communication protocol, and automatically calibrates the time after parsing the packaged data to the current time data according to the local clock, and directly issues control instructions to each node through Ethernet.

Images