Fan-coil control method and system based on LoRa communication

Abstract

The application discloses a fan-coil control method and system based on LoRa communication, and the method comprises the following steps: collecting temperature data according to a predetermined period, and pre-processing the temperature data; based on the pre-processed temperature data, a distributed control algorithm is used to calculate a cooperative control strategy; historical operation data of the fan-coil are acquired, a machine learning model is constructed, a reinforcement learning algorithm is used to optimize the control parameters of the machine learning model in a simulation environment; and based on the cooperative control strategy and the optimized control parameters, the operation state of the fan-coil is adjusted. The application applies LoRa communication to the fan-coil control system, realizes remote control of the fan-coil control system, and improves energy utilization.

Description

Technical Field

[0001] This invention belongs to the field of industrial control and discloses a fan coil control method and system based on LoRa communication. Background Technology

[0002] The main application scenarios for fan coil unit central water-cooled air conditioners are large buildings, such as office buildings and shopping malls, which have relatively large power and electricity consumption. Traditional fan coil unit controllers are operated manually and cannot be centrally controlled remotely, resulting in relatively low energy utilization. Summary of the Invention

[0003] Purpose of the invention: To provide a fan coil control method and system based on LoRa communication to solve the above-mentioned problems existing in the prior art.

[0004] Technical solution: A fan coil control method based on LoRa communication, comprising the following steps:

[0005] S1. Collect temperature data according to the predetermined cycle and preprocess the temperature data;

[0006] S2. Based on the preprocessed temperature data, a distributed control algorithm is used to calculate the cooperative control strategy;

[0007] S3. Obtain historical operating data of fan coil units, construct a machine learning model, and use reinforcement learning algorithms to optimize the control parameters of the machine learning model in a simulation environment;

[0008] S4. Adjust the operating status of the fan coil unit based on the collaborative control strategy and optimized control parameters.

[0009] According to one aspect of this application, step S1 further comprises:

[0010] S11 The fan coil unit intelligent controller collects temperature data according to a predetermined cycle and sends the temperature data to the edge gateway through the LoRa module.

[0011] S12. The edge gateway forwards the received temperature data to the server;

[0012] S13. The server preprocesses the received temperature data, including removing noise, correcting deviations, and filling in missing values.

[0013] According to one aspect of this application, step S2 further comprises:

[0014] S21. Using a clustering algorithm, the preprocessed temperature data is grouped to determine the area where each fan coil unit is located;

[0015] S22. In each area, select a fan coil unit as a representative fan coil unit, and the representative fan coil unit collects the status information of other fan coil units in the area.

[0016] S23. Based on the collected status information, the fan coil unit adopts a distributed control algorithm to calculate the collaborative control strategy;

[0017] S24 indicates that the fan coil unit distributes the collaborative control strategy to other fan coil units in the area via the LoRa network.

[0018] According to one aspect of this application, step S3 further comprises:

[0019] S31. The server collects historical operating data of fan coil units in various regions and constructs machine learning samples based on the historical operating data;

[0020] S32. Obtain environmental feature data of various typical shopping malls, and construct a reinforcement learning simulation environment set based on the environmental feature data;

[0021] S33. Construct a fan coil unit control strategy network. For each environment in the reinforcement learning simulation environment set, use the DQN algorithm and machine learning samples to train the fan coil unit control strategy network.

[0022] S34. Collect the fan coil control strategy networks trained in each environment to form a strategy pool;

[0023] S35. Construct a meta-learning model and train it based on the network parameters in the policy pool.

[0024] S36. Obtain the new environmental feature data of the current shopping mall. Based on the new environmental feature data, use the trained meta-learning model to generate the optimal fan coil control strategy network and obtain the optimal control parameters.

[0025] S37. The server sends the optimal control parameters to the representative fan coil units in each region, and the representative fan coil units then send the optimal control parameters to other fan coil units in the region.

[0026] According to one aspect of this application, step S4 further comprises:

[0027] S41. Based on the collaborative control strategy and optimal control parameters, the fan coil unit intelligent controller adjusts the operating status of the fan coil unit to achieve the preset target temperature.

[0028] S42. Obtain the real-time temperature after adjusting the operating status. Based on the difference between the real-time temperature and the target temperature, use the PID algorithm to further adjust the operating status of the fan coil unit to achieve precise temperature control.

[0029] S43. Periodically feed back the effect of temperature control to the representative fan coil unit for the next round of optimization iteration.

[0030] According to one aspect of this application, step S23 further comprises:

[0031] S231. Based on the collected status information, construct the optimization target for each representative fan coil unit;

[0032] S232. Based on the optimization objective of each representative fan coil unit, the concept of Nash equilibrium in game theory is used to form a joint optimization problem;

[0033] S233. Solve the joint optimization problem through iterative game theory to obtain the optimal collaborative control strategy.

[0034] A fan coil control system based on LoRa communication, using the fan coil control method based on LoRa communication described in any of the above technical solutions, includes a power supply unit, a human-machine interaction unit, a temperature and humidity detection unit, a processing unit, a control unit, and a LoRa communication unit;

[0035] The power supply unit includes an AC-DC power conversion circuit that converts AC voltage to DC voltage and provides power to the processing unit, human-machine interaction unit, temperature and humidity detection unit and LoRa communication unit.

[0036] The human-computer interaction unit is used to receive user control commands and send the control commands to the processing unit, and can also display system status information;

[0037] The temperature and humidity detection unit is used to monitor the temperature and humidity data of the environment and send it to the processing unit.

[0038] The processing unit receives temperature and humidity data from the temperature and humidity detection unit and control commands from the human-machine interaction unit, and sends control commands to the control unit.

[0039] The control unit receives control commands from the processing unit, controls the operating status of the fan coil unit, and feeds back the operating status of the fan coil unit to the processing unit.

[0040] The LoRa communication unit sends the temperature and humidity data and the operating status of the fan coil unit from the processing unit to the server, and the server sends control commands to the processing unit through the LoRa communication unit.

[0041] According to one aspect of this application, the AC-CDC power conversion circuit uses an isolated flyback power converter, wherein the AC input voltage range is 176V-264V; the DC output voltage is 5V, the output current is 250mA; and the AC-CDC primary and secondary isolation voltage is 2500V.

[0042] According to one aspect of this application, the control unit includes at least five control relays.

[0043] According to one aspect of this application, the processing unit includes a microcontroller and its peripheral circuitry, and the LoRa communication unit includes an S78F transceiver and its peripheral circuitry, wherein the S78F transceiver employs a LoRa long-range modem.

[0044] Beneficial effects: By applying LoRa communication to the fan coil control system, this invention enables remote control of the fan coil control system and improves energy utilization. Attached Figure Description

[0045] Figure 1 This is a flowchart of the present invention.

[0046] Figure 2 This is a flowchart of step S1 of the present invention.

[0047] Figure 3 This is a flowchart of step S2 of the present invention.

[0048] Figure 4 This is a flowchart of step S3 of the present invention.

[0049] Figure 5 This is a flowchart of step S4 of the present invention.

[0050] Figure 6 This is a schematic diagram of a fan coil control system based on LoRa communication.

[0051] Figure 7 This is a diagram showing the relationship between the processing unit and the LoRa communication unit.

[0052] Figure 8 This is a diagram showing the relationship between the processing unit and the control unit.

[0053] Figure 9 This is a diagram showing the relationship between the processing unit and the temperature and humidity detection unit. Detailed Implementation

[0054] like Figure 1 As shown, this application proposes a fan coil control method based on LoRa communication, the specific steps of which are as follows:

[0055] S1. Collect temperature data according to the predetermined cycle and preprocess the temperature data;

[0056] S2. Based on the preprocessed temperature data, a distributed control algorithm is used to calculate the cooperative control strategy;

[0057] S3. Obtain historical operating data of fan coil units, construct a machine learning model, and use reinforcement learning algorithms to optimize the control parameters of the machine learning model in a simulation environment;

[0058] S4. Adjust the operating status of the fan coil unit based on the collaborative control strategy and optimized control parameters.

[0059] In this embodiment, the system automatically collects ambient temperature data within a predetermined period. This data forms the basis for the control system's decisions. The collected data may contain noise or be incomplete; preprocessing improves data quality, ensuring accuracy and reliability. Using the preprocessed data, a collaborative control strategy is calculated through a distributed control algorithm. This algorithm can allocate tasks among multiple control nodes, enabling collaborative work. Historical operating data of the fan coil units is collected to build a machine learning model. Reinforcement learning, a method that learns optimal strategies through rewards and penalties, enables the control system to learn and adapt, improving the intelligence of the control strategy and the overall performance of the system. The optimized control parameters and collaborative control strategy are then applied to actual control, adjusting the fan coil unit's operating state in real time to achieve the desired temperature control effect. This ensures the system can make optimal control decisions based on actual conditions, improving energy efficiency and user comfort.

[0060] This embodiment utilizes advanced data processing and machine learning technologies to achieve intelligent control of the fan coil system, improving system efficiency and reliability while reducing energy consumption.

[0061] like Figure 2 As shown, according to one aspect of this application, step S1 further comprises:

[0062] S11 The fan coil unit intelligent controller collects temperature data according to a predetermined cycle and sends the temperature data to the edge gateway through the LoRa module.

[0063] S12. The edge gateway forwards the received temperature data to the server;

[0064] S13. The server preprocesses the received temperature data, including removing noise, correcting deviations, and filling in missing values.

[0065] LoRa wireless communication technology is a low-power local area network (LAN) wireless standard. Its biggest advantage is that it can propagate farther than other wireless methods under the same power consumption conditions, achieving a balance between low power consumption and long distance. In this embodiment, the collected temperature data is sent to the edge gateway through the LoRa module, realizing low-power and long-distance data transmission. This allows the system to be deployed over a wide area, suitable for centralized control of large buildings or multiple buildings. The edge gateway acts as a relay station, improving the efficiency and reliability of data transmission while reducing the workload of the central server, allowing it to handle more data from different edge gateways. The server uses algorithms to remove noise from the data, improving data quality; it corrects the data to eliminate system errors; and it fills in missing data points using interpolation or other methods, ensuring the accuracy and completeness of the data and providing a reliable foundation for subsequent data analysis and decision-making.

[0066] In a further embodiment, multiple intelligent fan coil unit controllers are installed in a large office building. Each controller periodically collects temperature data for its respective area and transmits it to the floor's edge gateway via a LoRa module. The edge gateway on each floor then sends the collected data to a central server located in the building's basement. The server preprocesses all the data and then adjusts the building's temperature control system accordingly to ensure that the temperature in each area remains within a set comfort range. Such a system not only improves energy efficiency but also enhances employee comfort and work efficiency.

[0067] like Figure 3 As shown, according to one aspect of this application, step S2 further comprises:

[0068] S21. Using a clustering algorithm, the preprocessed temperature data is grouped to determine the area where each fan coil unit is located;

[0069] S22. In each area, select a fan coil unit as a representative fan coil unit, and the representative fan coil unit collects the status information of other fan coil units in the area.

[0070] S23. Based on the collected status information, the fan coil unit adopts a distributed control algorithm to calculate the collaborative control strategy;

[0071] S24 indicates that the fan coil unit distributes the collaborative control strategy to other fan coil units in the area via the LoRa network.

[0072] In this embodiment, a clustering algorithm is used to group the preprocessed temperature data to determine the regions where different fan coil units are located. The clustering algorithm can be K-means, DBSCAN, etc. Clustering identifies regions with similar temperature characteristics, providing a foundation for distributed control. A representative fan coil unit collects status information including temperature, humidity, and wind speed. This not only reduces data transmission volume but also lowers network load because only the representative fan coil unit needs to communicate with the server. A distributed control algorithm is used to calculate a collaborative control strategy. This strategy considers the status of the entire region and can more accurately adjust the operation of each fan coil unit, achieving collaborative operation of fan coil units within the region and improving overall control efficiency and comfort. The representative fan coil unit distributes the collaborative control strategy to other fan coil units within the region via a LoRa network. Each fan coil unit adjusts its operating status according to the received strategy, ensuring that all fan coil units within the region can operate efficiently according to a unified strategy, optimizing energy use.

[0073] In a further embodiment, each floor of a large office building is divided into multiple zones, each equipped with multiple fan coil units. The system first uses a clustering algorithm to analyze the floor's temperature data to determine the boundaries of each zone. Then, a representative fan coil unit is selected at the center of each zone, and status information of other fan coil units within that zone is collected. The representative fan coil unit uses a distributed control algorithm to calculate the optimal collaborative control strategy and distributes this strategy to other fan coil units within the same zone via a LoRa network. In this way, the temperature control system for the entire commercial center can automatically adjust the temperature of each zone based on real-time data and a unified strategy to achieve optimal energy efficiency and comfort. This method not only improves the system's intelligence level but also helps save energy and reduce costs.

[0074] Step S23 further includes:

[0075] S231. Based on the collected status information, construct the optimization target for each representative fan coil unit;

[0076] S232. Based on the optimization objective of each representative fan coil unit, the concept of Nash equilibrium in game theory is used to form a joint optimization problem;

[0077] S233. Solve the joint optimization problem through iterative game theory to obtain the optimal collaborative control strategy.

[0078] In a further embodiment, the status information of all fan coil units (FCUs) within the area is collected on behalf of the fan coil unit (FCU), including current temperature, set temperature, energy consumption, etc.; each FCU is modeled as an agent whose goal is to minimize its own energy consumption while meeting temperature requirements; using the Nash equilibrium concept from game theory, the optimization objectives of multiple FCUs are formalized into a joint optimization problem; the game G = {N, {S} is defined. i}, {U i}}, where N is the set of intelligent agents (fan coil units), S i U is the policy space of the i-th agent, which contains the selectable temperature setpoint and airflow rate. i It is the utility function of the i-th agent, taking into account factors such as energy consumption, temperature deviation, and degree of cooperation. Through iterative game theory, each agent continuously adjusts its own strategy until Nash equilibrium is reached, at which point the control strategy of each FCU is optimal. After the game converges, it represents the fan coil unit distributing the calculated control strategy to other fan coil units in the region.

[0079] like Figure 4 As shown, according to one aspect of this application, step S3 further comprises:

[0080] S31. The server collects historical operating data of fan coil units in various regions and constructs machine learning samples based on the historical operating data;

[0081] S32. Obtain environmental feature data of various typical shopping malls, and construct a reinforcement learning simulation environment set based on the environmental feature data;

[0082] S33. Construct a fan coil unit control strategy network. For each environment in the reinforcement learning simulation environment set, use the DQN algorithm and machine learning samples to train the fan coil unit control strategy network.

[0083] S34. Collect the fan coil control strategy networks trained in each environment to form a strategy pool;

[0084] S35. Construct a meta-learning model and train it based on the network parameters in the policy pool.

[0085] S36. Obtain the new environmental feature data of the current shopping mall. Based on the new environmental feature data, use the trained meta-learning model to generate the optimal fan coil control strategy network and obtain the optimal control parameters.

[0086] S37. The server sends the optimal control parameters to the representative fan coil units in each region, and the representative fan coil units then send the optimal control parameters to other fan coil units in the region.

[0087] In this embodiment, firstly, machine learning samples are constructed to provide rich training data for the machine learning model, helping the model to better understand the system's operating rules. Environmental characteristic data from various typical shopping malls are collected, including malls with different layouts and foot traffic. Based on this data, a reinforcement learning simulation environment set is constructed to simulate different shopping mall environments, enabling the control policy network to be trained in diverse environments and improving the model's generalization ability. For each environment, a dedicated FCU control policy network is trained using the DQN algorithm. The state space includes temperature, humidity, and number of people in each area; the action space is a combination of selectable FCU temperature setpoints; and the reward function comprehensively considers factors such as energy consumption, temperature comfort, and load balance. Through the DQN algorithm, the model can learn how to make optimal decisions in different environments. The control policy networks trained in different environments are then combined to form a... A policy pool contains control policies for various environments, providing diverse learning materials for the meta-learning model. Network parameters from the policy pool are used as input to train a meta-policy network. This network takes simulation environment characteristics as input and outputs the optimal policy network parameters for each environment. The meta-learning model can quickly adapt to new environments, improving its adaptability and flexibility. In practical applications, based on the characteristics of the current shopping mall environment, a dedicated optimal control policy network is generated using the meta-policy network, ensuring that the control policy is optimized for the specific environment of the mall. The generated policy network is used to calculate the optimal control parameters, which are then distributed to representative fan coil units. During actual operation, feedback data is continuously collected for offline updates to the meta-policy network, achieving adaptive optimization of the policy. All fan coil units can operate according to the optimal control parameters, improving the efficiency and consistency of the entire system.

[0088] In this embodiment, adaptive generation of control policies is achieved by learning the optimal policies under different environments. The combination of offline training and online generation reduces the exploration cost of real-world systems and improves the generalization ability of the policies.

[0089] In a further embodiment, the server first collects and analyzes the operating data of fan coil units in various office buildings over the past year. Then, it constructs a simulation environment set by combining environmental characteristic data from different seasons and special events. Next, it trains a control policy network using the DQN algorithm and aggregates these networks into a policy pool. Through training the meta-learning model, the system can quickly adapt to new environmental characteristics such as the Christmas shopping season. Finally, the server generates the optimal control policy network based on the latest environmental data and distributes the control parameters to the fan coil units on each floor, ensuring a comfortable shopping environment even during periods of high foot traffic and achieving efficient energy utilization.

[0090] like Figure 5 As shown, according to one aspect of this application, step S4 further comprises:

[0091] S41. Based on the collaborative control strategy and optimal control parameters, the fan coil unit intelligent controller adjusts the operating status of the fan coil unit to achieve the preset target temperature.

[0092] S42. Obtain the real-time temperature after adjusting the operating status. Based on the difference between the real-time temperature and the target temperature, use the PID algorithm to further adjust the operating status of the fan coil unit to achieve precise temperature control.

[0093] S43. Periodically feed back the effect of temperature control to the representative fan coil unit for the next round of optimization iteration.

[0094] In this embodiment, the intelligent controller of the fan coil unit adjusts its operating state according to the collaborative control strategy and optimal control parameters, including adjusting the fan speed and changing the cooling or heating mode, to achieve the preset target temperature. This ensures that the fan coil unit system can operate efficiently according to the overall optimal strategy, achieving energy saving and improved comfort. After adjusting the operating state, the system acquires real-time temperature data. Based on the difference between the real-time temperature and the target temperature, the system uses a PID algorithm to further adjust the operating state of the fan coil unit. The PID algorithm can achieve precise temperature control, quickly reduce temperature deviation, and improve the system's response speed and stability. Finally, the effect of temperature control is periodically fed back to the representative fan coil unit, which performs the next round of optimization iteration based on the feedback information. Through periodic feedback and iterative optimization, the system can continuously improve itself, adapt to environmental changes, and improve the effectiveness of the overall control strategy.

[0095] In a further embodiment, within the office building, the fan coil units in each area first undergo initial adjustments based on the collaborative control strategy and optimal control parameters issued by the server. Then, the system monitors the indoor temperature in real time using sensors and fine-tunes the fan coil unit's operation using a PID algorithm to ensure the temperature reaches the set value. Finally, each fan coil unit feeds back the control effect to its area representative, who then sends this information back to the server. The server analyzes this data and optimizes the control parameters, completing one iteration. This process is repeated periodically to ensure the office building's temperature remains at its most comfortable level while also achieving efficient energy utilization. This method not only improves employee comfort but also helps reduce energy consumption, achieving green office practices.

[0096] This application combines LoRa technology with edge computing to achieve distributed collaborative control of fan coil units, improving control flexibility and scalability. An adaptive optimization mechanism is introduced, utilizing reinforcement learning to continuously optimize control system parameters, enhancing the system's intelligence level. A hierarchical control architecture achieves global collaboration while ensuring real-time control through local control. This innovative solution enables highly efficient energy-saving control of fan coil units, reducing energy consumption and improving user comfort. Furthermore, thanks to LoRa's low-power, long-distance communication characteristics, this solution can significantly reduce system deployment and maintenance costs.

[0097] like Figure 6 As shown, a fan coil control system based on LoRa communication includes a power supply unit, a human-machine interaction unit, a temperature and humidity detection unit, a processing unit, a control unit, and a LoRa communication unit.

[0098] The power supply unit includes an AC-DC power conversion circuit that converts AC voltage to DC voltage and provides power to the processing unit, human-machine interaction unit, temperature and humidity detection unit and LoRa communication unit.

[0099] The human-computer interaction unit is used to receive user control commands and send the control commands to the processing unit, and can also display system status information;

[0100] The temperature and humidity detection unit is used to monitor the temperature and humidity data of the environment and send it to the processing unit.

[0101] The processing unit receives temperature and humidity data from the temperature and humidity detection unit and control commands from the human-machine interaction unit, and sends control commands to the control unit.

[0102] The control unit receives control commands from the processing unit, controls the operating status of the fan coil unit, and feeds back the operating status of the fan coil unit to the processing unit.

[0103] The LoRa communication unit sends the temperature and humidity data and the operating status of the fan coil unit from the processing unit to the server, and the server sends control commands to the processing unit through the LoRa communication unit.

[0104] The first output terminal of the power supply unit is connected to the power input terminal of the processing unit, the second output terminal of the power supply unit is connected to the power input terminal of the human-machine interaction unit, the third output terminal of the power supply unit is connected to the temperature and humidity detection unit, and the fourth output terminal of the power supply unit is connected to the LoRa communication unit; the first signal input / output terminal of the processing unit is connected to the human-machine interaction unit, the second signal input / output terminal of the processing unit is connected to the LoRa communication module, the third signal input / output terminal of the processing unit is connected to the temperature and humidity detection unit, and the signal output terminal of the processing unit is connected to the control unit.

[0105] The AC-CDC power conversion circuit of the power supply unit adopts isolated flyback power conversion, with an AC input voltage range of 176V-264V; DC output voltage of 5V; output current of 250mA; and AC-CDC primary and secondary isolation voltage of 2500V.

[0106] The control unit includes HT, CL, HF, MF, and LF input terminals. These input terminals are connected to the HT, CL, HF, MF, and LF output terminals of the processing unit, respectively. The HT, CL, HF, MF, and LF input terminals are also connected to five control relays on the control unit. The opening and closing of these five relays are controlled in real-time by the processing unit. The output states controlled by these relays include cooling, heating, high-speed fan, medium-speed fan, and low-speed fan, which are reflected in the fan coil unit controller's output states. Specifically, the HT input terminal is for high-temperature operation, the CL input terminal is for closed-loop operation, the HF input terminal is for high-frequency operation, the MF input terminal is for medium-frequency operation, and the LF input terminal is for low-frequency operation.

[0107] The control unit also includes drive circuits for five control relays.

[0108] The processing unit includes the main microcontroller and its peripheral circuits, such as... Figure 7 , Figure 8 and Figure 9 As shown, the input / output signals of the processing unit are connected to the input / output signals of the LoRa communication unit; the output signal of the processing unit is connected to the input signal of the control unit; the processing unit is connected to the IIC interface of the temperature and humidity acquisition unit through the IIC interface; after acquiring temperature and humidity parameters, the processing unit controls the airflow of the fan coil unit through the corresponding relay of the control unit, thereby controlling the indoor temperature within the set range. Alternatively, the current control parameters can be uploaded to the main control software via the LoRa communication unit for remote monitoring, and the main control software can send the control parameters back to the processing unit via the LoRa communication unit for remote control functionality. The IIC (Inter-Integrated Circuit) interface is a serial bus used for communication between a microcontroller and peripheral devices.

[0109] The human-computer interaction unit includes an LCD display panel, buttons, and other human-computer interaction components. The LCD display panel allows users to view parameters such as current temperature, wind speed, and mode in real time, while the buttons ensure that users can also set some lower-priority parameters on the LCD display panel.

[0110] The temperature and humidity detection unit includes a temperature and humidity sensor and its peripheral circuitry. The temperature and humidity sensor, model NSHT30, integrates a capacitive relative humidity sensor, a CMOS (Complementary Metal-Oxide-Semiconductor) temperature sensor, a signal processor, and a high-speed IIC interface onto a single chip, packaged in a small dual-plane leadless package (small open-cavity DFN package). Humidity detection accuracy reaches 3%RH, and temperature detection accuracy reaches 0.3℃. The collected temperature and humidity data is uploaded to the processing unit via the IIC interface, enabling the processing unit to precisely control the ambient temperature, achieving energy conservation and emission reduction.

[0111] The LoRa communication unit includes an S78F transceiver and its peripheral circuitry. The transceiver includes GPIO_DIO3 input / output ports, GPIO_DIO2 input / output ports, GPIO_DIO0 input / output ports, GPIO_RF input / output ports, SDN_RF input / output ports, RF_MOSI input ports, RF_MISO output ports, RF_CLK input / output ports, and RF_NSS input / output ports. The GPIO_DIO3 input / output port is connected to the GPIO_DIO3 input / output port of the processing unit, the GPIO_DIO2 input / output port is connected to the GPIO_DIO2 input / output port of the processing unit, and the GPIO_DIO0 input / output port is connected to the GPIO_DIO0 input / output port of the processing unit. The S78F transceiver is connected to the processing unit's GPIO_RF input / output port, SDN_RF input / output port, RF_MOSI input port, RF_MISO output port, RF_CLK input / output port, and RF_NSS input / output port. The transceiver uses a LoRa long-range modem, providing ultra-long-range spread spectrum communication and high interference immunity while minimizing current consumption, achieving a sensitivity exceeding -138dBm. Compared to conventional modulation techniques, LoRa offers significant advantages in terms of blocking and selectivity, solving the problems of distance, interference immunity, and power consumption inherent in traditional designs. It has significant advantages in environments requiring long-distance communication, such as large buildings.

[0112] In a further embodiment, a system fan coil control system is provided, with the following system architecture: each fan coil unit is equipped with an intelligent controller integrating a LoRa communication module and a temperature sensor; multiple LoRa edge gateways are deployed in the shopping mall to be responsible for data aggregation and forwarding; and FCU control and optimization algorithms are deployed on a local server to achieve real-time control and policy updates.

[0113] This application utilizes LoRa wireless communication technology to enable remote control of fan coil unit controllers. LoRa communication features ultra-long-distance spread spectrum communication and high anti-interference capabilities, allowing fan coil unit controllers to achieve true remote control within large buildings, thereby achieving energy conservation and emission reduction.

[0114] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all fall within the protection scope of the present invention.

Claims

1. A fan coil unit control method based on LoRa communication, characterized in that, Includes the following steps: S1. Collect temperature data according to the predetermined cycle and preprocess the temperature data; S2. Based on the preprocessed temperature data, a distributed control algorithm is used to calculate the cooperative control strategy; S3. Obtain historical operating data of fan coil units, construct a machine learning model, and use reinforcement learning algorithms to optimize the control parameters of the machine learning model in a simulation environment; S4. Adjust the operating status of the fan coil unit based on the collaborative control strategy and optimized control parameters; Step S2 further comprises: S21. Using a clustering algorithm, the preprocessed temperature data is grouped to determine the area where each fan coil unit is located; S22. In each area, select a fan coil unit as a representative fan coil unit, and the representative fan coil unit collects the status information of other fan coil units in the area. S23. Based on the collected status information, the fan coil unit adopts a distributed control algorithm to calculate the collaborative control strategy; S24 indicates that the fan coil unit sends the collaborative control strategy to other fan coil units in the area through the LoRa network; Step S3 further includes: S31. The server collects historical operating data of fan coil units in various regions and constructs machine learning samples based on the historical operating data; S32. Obtain environmental feature data of various typical shopping malls, and construct a reinforcement learning simulation environment set based on the environmental feature data; S33. Construct a fan coil unit control strategy network. For each environment in the reinforcement learning simulation environment set, use the DQN algorithm and machine learning samples to train the fan coil unit control strategy network. S34. Collect the fan coil control strategy networks trained in each environment to form a strategy pool; S35. Construct a meta-learning model and train it based on the network parameters in the policy pool. S36. Obtain the new environmental feature data of the current shopping mall. Based on the new environmental feature data, use the trained meta-learning model to generate the optimal fan coil control strategy network and obtain the optimal control parameters. S37. The server sends the optimal control parameters to the representative fan coil units in each region, and the representative fan coil units then send the optimal control parameters to other fan coil units in the region. Specifically, for each environment, a dedicated FCU control policy network is trained using the DQN algorithm. The state space consists of the temperature, humidity, and number of people in each area; the action space consists of the combination of FCU temperature setpoints; and the reward function is a comprehensive consideration of energy consumption, temperature comfort, and load balance. Through the DQN algorithm, the model can learn how to make optimal decisions in different environments. The control policy networks trained in different environments are aggregated to form a policy pool, which contains control policies in various environments, providing diverse learning materials for the meta-learning model. The network parameters in the policy pool are used as input to the meta-learning model to train a meta-policy network. The meta-policy network takes the characteristics of the simulation environment as input and outputs the parameters of the optimal policy network in each environment.

2. The fan coil unit control method based on LoRa communication according to claim 1, characterized in that, Step S1 further comprises: S11 The fan coil unit intelligent controller collects temperature data according to a predetermined cycle and sends the temperature data to the edge gateway through the LoRa module. S12. The edge gateway forwards the received temperature data to the server; S13. The server preprocesses the received temperature data, including removing noise, correcting deviations, and filling in missing values.

3. The fan coil unit control method based on LoRa communication according to claim 1, characterized in that, Step S4 further includes: S41. Based on the collaborative control strategy and optimal control parameters, the fan coil unit intelligent controller adjusts the operating status of the fan coil unit to achieve the preset target temperature. S42. Obtain the real-time temperature after adjusting the operating status. Based on the difference between the real-time temperature and the target temperature, use the PID algorithm to further adjust the operating status of the fan coil unit to achieve precise temperature control. S43. Periodically feed back the effect of temperature control to the representative fan coil unit for the next round of optimization iteration.

4. The fan coil unit control method based on LoRa communication according to claim 1, characterized in that, Step S23 further includes: S231. Based on the collected status information, construct the optimization target for each representative fan coil unit; S232. Based on the optimization objective of each representative fan coil unit, the concept of Nash equilibrium in game theory is used to form a joint optimization problem; S233. Solve the joint optimization problem through iterative game theory to obtain the optimal collaborative control strategy.

5. A fan coil unit control system based on LoRa communication, characterized in that, The fan coil control method based on LoRa communication according to any one of claims 1 to 4 includes a power supply unit, a human-machine interaction unit, a temperature and humidity detection unit, a processing unit, a control unit, and a LoRa communication unit. The power supply unit includes an AC-DC power conversion circuit that converts AC voltage to DC voltage and provides power to the processing unit, human-machine interaction unit, temperature and humidity detection unit and LoRa communication unit. The human-computer interaction unit is used to receive user control commands and send the control commands to the processing unit, and can also display system status information; The temperature and humidity detection unit is used to monitor the temperature and humidity data of the environment and send it to the processing unit. The processing unit receives temperature and humidity data from the temperature and humidity detection unit and control commands from the human-machine interaction unit, and sends control commands to the control unit. The control unit receives control commands from the processing unit, controls the operating status of the fan coil unit, and feeds back the operating status of the fan coil unit to the processing unit. The LoRa communication unit sends the temperature and humidity data and the operating status of the fan coil unit from the processing unit to the server, and the server sends control commands to the processing unit through the LoRa communication unit.

6. The fan coil unit control system based on LoRa communication according to claim 5, characterized in that, The AC-CDC power conversion circuit uses an isolated flyback power converter, with an AC input voltage range of 176V-264V; a DC output voltage of 5V and an output current of 250mA; and an AC-CDC primary and secondary isolation voltage of 2500V.

7. The fan coil unit control system based on LoRa communication according to claim 5, characterized in that, The control unit includes at least five control relays.

8. The fan coil unit control system based on LoRa communication according to claim 5, characterized in that, The processing unit includes a microcontroller and its peripheral circuits, and the LoRa communication unit includes an S78F transceiver and its peripheral circuits. The S78F transceiver uses a LoRa long-range modem.

Images