A control method, apparatus, equipment, and medium for a refrigerant circulation system.

Abstract

This invention relates to the field of refrigeration and dehumidification technology, and discloses a control method, device, equipment, and medium for a refrigerant circulation system. The method includes: using the superheat setpoint output by a fuzzy neural network as the target value, predicting the system output using a model predictive control algorithm, and performing rolling optimization to solve for the optimal control action; predicting the probability that the exhaust temperature exceeds the alarm threshold using a long short-term memory network; executing actions according to the step-by-step protection logic of the variable frequency compressor; dynamically adjusting the frequency change step size of the variable frequency compressor according to the frequency of the variable frequency compressor; and entering the next control cycle if the shutdown conditions are not met. This invention achieves intelligent control from passive response to active prediction by combining look-ahead collaborative control of the model predictive control algorithm, dynamic superheat adaptive setting of the fuzzy neural network, fault prediction of the long short-term memory network, and step-by-step protection logic of the variable frequency compressor, thereby improving the adaptability, operational continuity, and reliability of the refrigerant circulation system.

Description

Technical Field

[0001] This invention relates to the field of refrigeration and dehumidification technology, specifically to a control method, device, equipment, and medium for a refrigerant circulation system. Background Technology

[0002] Refrigerant circulation systems, based on thermodynamic principles, control and regulate temperature through the phase change and circulation of the refrigerant, and are widely used in refrigeration and heating equipment. Current control strategies for refrigerant circulation systems typically treat the compressor, electronic expansion valve, fan, and other actuators as individual units, resulting in simplistic protection strategies. Therefore, achieving intelligent control of refrigerant circulation systems has become a crucial issue. Summary of the Invention

[0003] This invention provides a control method, apparatus, equipment, and medium for a refrigerant circulation system to solve the problem of how to achieve intelligent control of a refrigerant circulation system.

[0004] In a first aspect, the present invention provides a control method for a refrigerant circulation system, the refrigerant circulation system including a variable frequency compressor, an electronic expansion valve, and a main controller, the method comprising: When the main controller receives the power-on command, it reads the high-pressure sensor value and the low-pressure sensor value. When the pressure difference between the high-pressure sensor value and the low-pressure sensor value is greater than the start-up pressure difference threshold, it controls the electronic expansion valve to open to the pressure difference opening degree until the pressure difference falls back to the safe range. Before the variable frequency compressor starts, the electronic expansion valve is controlled to open to the initial opening degree and maintain the preset opening time. After the inverter compressor starts, it makes a working mode decision and switches to the corresponding working mode. The working modes include forced dehumidification mode, cooling mode and dehumidification mode. Within the control cycle, the overheating setpoint output by the fuzzy neural network is used as the overheating tracking target value. The system output is predicted using the model predictive control algorithm, and the optimal control action is solved by rolling optimization with the goal of minimizing the cost function. The probability of exhaust temperature exceeding the alarm threshold is predicted using a long short-term memory network, and the system state is determined based on the probability of exhaust temperature exceeding the alarm threshold. The system state includes frequency limiting preparation state and frequency reduction preparation state. When any of the exhaust temperature, high pressure, or variable frequency compressor current is detected to reach the corresponding threshold, the corresponding actions are executed according to the pre-set variable frequency compressor step protection logic. Forced intervention at high and low evaporation temperatures using electronic expansion valves; The frequency change step size of the variable frequency compressor is dynamically adjusted according to the frequency of the variable frequency compressor. If the shutdown conditions are not met, the system enters the next control cycle and returns to the steps of using the model predictive control algorithm to predict the system output and to continuously optimize and solve for the optimal control action with the goal of minimizing the cost function.

[0005] This invention achieves intelligent control from passive response to active prediction by combining forward-looking collaborative control of model predictive control algorithm, dynamic superheat adaptive setting of fuzzy neural network, fault prediction of long short-term memory network and step-by-step protection logic of variable frequency compressor, thereby improving the adaptability, operation continuity and reliability of refrigerant circulation system.

[0006] In one optional implementation, a working mode decision is made, and the system switches to the corresponding working mode, including: If the forced dehumidification input signal is valid, switch to forced dehumidification mode; If the forced dehumidification input signal is invalid and the indoor temperature is greater than or equal to the temperature threshold, then switch to cooling mode; After exiting the cooling mode, if the indoor humidity is greater than or equal to the humidity threshold, it will switch to the dehumidification mode.

[0007] This invention achieves response to dehumidification needs and adaptive adjustment of temperature and humidity through an automatic working mode switching mechanism. When the temperature exceeds the standard, it prioritizes cooling to avoid excessive dehumidification, and when the humidity exceeds the standard, it prioritizes dehumidification to maintain a dry environment and avoid energy efficiency loss caused by frequent start-stop of the variable frequency compressor.

[0008] In one optional implementation, the system output is predicted using a model predictive control algorithm, and the optimal control action is solved through rolling optimization with the goal of minimizing the cost function, including: Read system status parameters, including variable frequency compressor frequency, electronic expansion valve opening, exhaust temperature, suction superheat, and evaporation temperature; Using system state parameters as initial conditions, a lightweight system response model is used to predict the system output. Substitute the predicted sequence into the cost function, and take the control sequence that minimizes the cost function as the optimal control sequence. Extract the first control action as the optimal control action.

[0009] This invention uses a model predictive control algorithm to predict the system output based on system state parameters, and then uses rolling optimization to solve for the optimal control action, thereby achieving look-ahead control and shortening the system response delay.

[0010] In one alternative implementation, the fuzzy neural network includes an input layer, a fuzzification layer, a fuzzy inference layer, and a defuzzification layer, and outputs an overheating setpoint in the following manner: The input layer receives node variables, which include evaporation temperature, ambient temperature, ambient humidity, and variable frequency compressor frequency. The node variables received by the input layer are blurred using a blurring layer; The fuzzy inference layer performs fuzzy inference based on the fuzzification results and stored fuzzy rules, and outputs the rule trigger strength. The output of the fuzzy inference layer is converted into an overheat setting value using a defuzzification layer.

[0011] This invention uses a fuzzy neural network to dynamically output a superheat setpoint, replacing the traditional fixed superheat setpoint. It dynamically adjusts the superheat setpoint based on operating conditions such as evaporation temperature, ambient temperature and humidity, and variable frequency compressor frequency, transforming from static to dynamic adaptive, thus taking into account energy efficiency, heat exchange efficiency, and anti-liquid slugging requirements.

[0012] In one optional implementation, the system state is determined based on the probability that the exhaust temperature exceeds an alarm threshold, including: If the probability that the exhaust temperature exceeds the alarm threshold is greater than or equal to the first probability threshold, the control system will enter the frequency limiting preparation state in advance. If the probability that the exhaust temperature exceeds the alarm threshold is greater than or equal to the second probability threshold, the control system will perform frequency reduction action in advance.

[0013] This invention achieves intelligent control with advance prediction and tiered intervention by advancing protection decisions based on the probability of exhaust temperature exceeding the alarm threshold predicted by the long short-term memory network. It enables the control system to enter the frequency limiting preparation state in advance to prevent the frequency from rising further, and to execute frequency reduction actions in advance to reduce the load.

[0014] In one alternative implementation, the electronic expansion valve is subjected to forced intervention at high and low evaporation temperatures, including: If the evaporation temperature is less than or equal to the low evaporation temperature setting value and remains so for the first preset duration, the opening of the electronic expansion valve will be increased. If the evaporation temperature is greater than or equal to the high evaporation temperature setting value and continues for a second preset duration, the opening of the electronic expansion valve will be reduced.

[0015] This invention adds a high and low evaporation temperature forced intervention logic to the electronic expansion valve, increasing or decreasing the opening of the electronic expansion valve to achieve active intervention, maintain the evaporation temperature within a safe range, and ensure the safe and reliable operation of the system.

[0016] In one alternative implementation, the frequency change step size of the variable frequency compressor is calculated according to the following formula: S_f=S_min+(S_max-S_min)×(f-f_min) / (f_max-f_min) Where S_f is the frequency change step size of the variable frequency compressor, S_min is the lower limit of the step size ratio, S_max is the upper limit of the step size ratio, f is the frequency of the variable frequency compressor, f_min is the minimum frequency of the variable frequency compressor, and f_max is the maximum frequency of the variable frequency compressor.

[0017] This invention solves the inherent defects of traditional fixed step sizes by setting the frequency change step size of the variable frequency compressor to adaptively adjust with the frequency of the variable frequency compressor.

[0018] In a second aspect, the present invention provides a control device for a refrigerant circulation system, the refrigerant circulation system including a variable frequency compressor, an electronic expansion valve, and a main controller, the device comprising: The first control module is used to read the high-pressure sensor value and the low-pressure sensor value when the main controller receives the power-on command. When the pressure difference between the high-pressure sensor value and the low-pressure sensor value is greater than the start-up pressure difference threshold, it controls the electronic expansion valve to open to the pressure difference opening degree until the pressure difference falls back to the safe range. The second control module is used to control the electronic expansion valve to open to the initial opening degree before the variable frequency compressor starts, and to maintain the preset opening time. The switching module is used to make a working mode decision after the inverter compressor starts and switch to the corresponding working mode. The working modes include forced dehumidification mode, cooling mode and dehumidification mode. The solution module is used to predict the system output using the overheat setpoint output by the fuzzy neural network as the overheat tracking target value within the control cycle, and to solve for the optimal control action by rolling optimization with the goal of minimizing the cost function. The decision module is used to predict the probability of exhaust temperature exceeding the alarm threshold using a long short-term memory network, and to decide the system state based on the probability of exhaust temperature exceeding the alarm threshold. The system state includes frequency limiting preparation state and frequency reduction preparation state. The execution module is used to execute corresponding actions according to the pre-set step protection logic of the variable frequency compressor when any one of the exhaust temperature, high pressure, or variable frequency compressor current reaches the corresponding threshold. The forced intervention module is used to force the electronic expansion valve to operate at high and low evaporation temperatures. The dynamic adjustment module is used to dynamically adjust the frequency change step size of the variable frequency compressor according to the frequency of the variable frequency compressor. The loop module is used to enter the next control cycle if the shutdown condition is not met, and return to the steps of using the model predictive control algorithm to predict the system output and to solve for the optimal control action by rolling optimization with the goal of minimizing the cost function.

[0019] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the control method for a refrigerant circulation system described in the first aspect or any corresponding embodiment thereof.

[0020] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the control method for a refrigerant circulation system described in the first aspect or any corresponding embodiment thereof. Attached Figure Description

[0021] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of the intelligent control architecture of a refrigerant circulation system according to an embodiment of the present invention; Figure 2 This is a schematic flowchart of a control method for a refrigerant circulation system according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the control flow for a refrigerant circulation system according to an embodiment of the present invention; Figure 4 This is a timing diagram illustrating the effects of MPC model predictive control and EEV collaborative look-ahead control according to an embodiment of the present invention. Figure 5 This is a diagram illustrating the effect of LSTM fault prediction and stepped protection advance timing according to an embodiment of the present invention. Figure 6 This is a structural block diagram of a control device for a refrigerant circulation system according to an embodiment of the present invention; Figure 7 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention.

[0023] Explanation of reference numerals in the attached figures: 1. Variable frequency compressor; 2. Condenser; 3. Electronic expansion valve; 4. Evaporator; 5. Sensor group; 6. Main controller; 7. Edge gateway; 8. Cloud server. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0026] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0027] In related technologies, the control strategy for refrigerant circulation systems typically treats the compressor, electronic expansion valve (EEV), fan, and other actuators as separate units for control, which mainly suffers from the following technical drawbacks: (1) Throttling control lags and poor system dynamic response: Electronic expansion valves generally adopt PID control based on suction superheat. This control is a reactive control. Under dynamic conditions such as compressor start-up and shutdown, frequency change or evaporator facing the risk of frosting, there is an inherent delay, which causes the electronic expansion valve opening adjustment to lag behind the system state change, which can easily cause evaporator frosting or liquid refrigerant backflow impacting the compressor. (2) The compressor protection logic is simple and the operation is not continuous enough: For abnormal working conditions such as high pressure, high temperature and overcurrent, the existing technology mostly adopts the hard protection strategy of "stopping when the threshold is reached". Although this kind of either-or protection method can prevent equipment damage, frequent unplanned shutdowns disrupt the continuity of the dehumidification process, and each restart is accompanied by a large current surge (8 to 12 times the rated current) and mechanical wear. (3) The control parameters are statically fixed, resulting in weak full-frequency adaptive capability: The frequency adjustment step size of the compressor is usually a fixed value, which makes it impossible to achieve optimal control that balances stability and response speed in the low-frequency, medium-frequency and high-frequency ranges. At the same time, the system lacks the ability to actively avoid the inherent mechanical resonance frequency of the compressor, which leads to a significant increase in the noise and vibration of the whole machine when operating in certain specific frequency ranges; (4) Lack of predictability and inability to intervene before failure: Existing control strategies are all reactive and can only be adjusted after deviation or failure occurs. They cannot predict future load changes (such as sudden changes in ambient temperature and humidity, system frost trend, and excessive exhaust temperature trend), which causes control to always lag behind the system state.

[0028] This invention provides a refrigerant circulation system, such as... Figure 1 As shown, the system includes a variable frequency compressor 1, a condenser 2, an electronic expansion valve 3, an evaporator 4, a sensor group 5, a main controller 6, an edge gateway 7, and a cloud server 8.

[0029] Specifically, the variable frequency compressor 1 adjusts its speed via a variable frequency drive to control the refrigerant circulation flow. The condenser 2 condenses the high-temperature, high-pressure gaseous refrigerant into a high-pressure liquid, releasing heat to the environment. The electronic expansion valve 3 precisely adjusts its opening via a stepper motor to control the refrigerant flow into the evaporator. The evaporator 4 evaporates the low-temperature, low-pressure liquid refrigerant into a low-pressure gas, absorbing heat and moisture from the environment. The sensor group 5 includes an exhaust temperature sensor, an intake temperature sensor, a fin temperature sensor, a high / low pressure sensor, and an ambient temperature and humidity sensor. The main controller 6 (MCU) runs MPC (Model Predictive Control) and FNN (Fuzzy Neural Network) lightweight models to perform real-time control. The edge gateway 7 runs LSTM (Long Short-Term Memory) and RL (Reinforcement Learning) models to perform fault prediction and parameter self-learning. The cloud server 8 is responsible for the offline training and periodic updates of complex models.

[0030] Specifically, the discharge port of the variable frequency compressor 1 is connected to the inlet of the condenser 2, and the suction port is connected to the outlet of the evaporator 4. The electronic expansion valve 3 is connected between the outlet of the condenser 2 and the inlet of the evaporator 4. The sensor group 5 is distributed in key positions of the system, and the signal output terminals are all connected to the analog / digital input ports of the main controller 6. The main controller 6 is connected to the compressor variable frequency driver through an RS485 bus and to the stepper motor of the electronic expansion valve 3 through a pulse signal line. The main controller 6 is connected to the edge gateway 7 through Ethernet / WiFi, and the edge gateway 7 is connected to the cloud server 8 through the Internet.

[0031] Lightweight models (MPC, FNN) are deployed on the edge MCU for real-time control, while computationally intensive models (LSTM, RL) are deployed on the edge gateway 7. Offline training is deployed in the cloud, achieving an optimal balance between computing power and real-time performance. Through model predictive control algorithms, the frequency of the variable frequency compressor 1 and the opening of the electronic expansion valve 3 are coordinated and pre-adjusted, reducing response latency by approximately 70%. A fuzzy neural network is constructed to dynamically generate the optimal superheat setpoint based on the current operating conditions, and it possesses online learning capabilities, enabling superheat control to evolve from "static optimal" to "dynamic adaptive optimal." Long short-term memory networks are used to learn the temporal characteristic patterns before faults occur, issuing early warnings and implementing tiered protection, reducing unplanned downtime by approximately 75%. The compressor frequency change step size is dynamically adjusted according to the current frequency, and reinforcement learning is used to autonomously discover and avoid mechanical resonance frequency points, reducing overall machine noise by approximately 7 dB(A).

[0032] According to an embodiment of the present invention, a control method for a refrigerant circulation system is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0033] This embodiment provides a control method for a refrigerant circulation system. Figure 2 This is a flowchart of a control method for a refrigerant circulation system according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps: Step S201: When the main controller receives the power-on command, it reads the high-pressure sensor value and the low-pressure sensor value. When the pressure difference between the high-pressure sensor value and the low-pressure sensor value is greater than the starting pressure difference threshold, it controls the electronic expansion valve to open to the pressure difference opening degree until the pressure difference falls back to the safe range.

[0034] In this embodiment of the invention, when the main controller receives the power-on command, it reads the high-pressure sensor value P_high and the low-pressure sensor value P_low, and calculates the pressure difference |P_high-P_low| between the high-pressure sensor value and the low-pressure sensor value.

[0035] When the pressure difference between the high-pressure sensor value and the low-pressure sensor value is greater than the starting pressure difference threshold ΔP_start (default 3.0 Bar), i.e. |P_high-P_low|≥ΔP_start, the electronic expansion valve is controlled to open to the pressure difference opening degree, with a default of 420 steps, and active pressure relief is performed until the pressure difference falls back to the safe range.

[0036] This soft-start protection logic prevents the inverter compressor from starting with high pressure, thus preventing damage to the equipment caused by excessive starting current and excessive mechanical load.

[0037] Step S202: Before the variable frequency compressor starts, control the electronic expansion valve to open to the initial opening degree and maintain the preset opening time.

[0038] In this embodiment of the invention, before the variable frequency compressor starts, the electronic expansion valve is controlled to open to the initial opening degree in advance and maintain the opening for a preset duration. The initial opening degree is 200 steps by default and the opening duration is 30 seconds by default.

[0039] It should be noted that only after the differential pressure has fallen back to a safe range and the electronic expansion valve has been pre-opened through the above steps S201 to S202 can the process proceed to step S203. Otherwise, the process will remain in the variable frequency compressor start-up preparation stage and will not execute subsequent operation control.

[0040] Step S203: After the variable frequency compressor starts, a working mode decision is made and the corresponding working mode is switched.

[0041] In this embodiment of the invention, after the variable frequency compressor starts, it makes a working mode decision based on the external input signal, or the indoor temperature, or the indoor humidity, and switches to the corresponding working mode. The working modes include forced dehumidification mode, cooling mode, and dehumidification mode.

[0042] Step S204: Within the control cycle, the overheating setpoint output by the fuzzy neural network is used as the overheating tracking target value. The system output is predicted using the model predictive control algorithm, and the optimal control action is solved by rolling optimization with the goal of minimizing the cost function.

[0043] In this embodiment of the invention, within each control cycle, a fuzzy neural network is used to output an overheating setpoint. The overheating setpoint output by the fuzzy neural network is used as the target value for overheating tracking. The model predictive control algorithm is used to predict the system output. With the goal of minimizing the cost function, the optimal control action is solved through rolling optimization.

[0044] Step S205: Use a long short-term memory network to predict the probability that the exhaust temperature exceeds the alarm threshold, and decide the system state based on the probability that the exhaust temperature exceeds the alarm threshold.

[0045] In this embodiment of the invention, a long short-term memory network is run on the edge gateway. The long short-term memory network is used to predict the probability that the exhaust temperature exceeds the alarm threshold. Based on the decision mechanism, the system state is decided according to the probability that the exhaust temperature exceeds the alarm threshold. The system state includes frequency limiting preparation state and frequency reduction preparation state.

[0046] Step S206: When any one of the exhaust temperature, high pressure, or variable frequency compressor current reaches the corresponding threshold, the corresponding actions are executed according to the pre-set variable frequency compressor step protection logic.

[0047] In this embodiment of the invention, a stepped protection logic for the variable frequency compressor is set. When any one of the following is detected to reach the corresponding threshold: exhaust temperature, high pressure, or variable frequency compressor current, the corresponding actions are executed according to the stepped protection logic of the variable frequency compressor, including maintaining the current frequency, reducing the frequency, or shutting down the variable frequency compressor.

[0048] Step S207: Force intervention on the electronic expansion valve for high and low evaporation temperatures.

[0049] In this embodiment of the invention, a forced intervention logic is added on top of the basic superheat PID regulation, that is, the electronic expansion valve is subjected to forced intervention at high and low evaporation temperatures.

[0050] Step S208: Dynamically adjust the frequency change step size of the variable frequency compressor according to the frequency of the variable frequency compressor.

[0051] In this embodiment of the invention, the frequency change step size of the variable frequency compressor is dynamically adjusted according to the frequency of the variable frequency compressor, so that the variable frequency compressor can obtain a smoother frequency change in the low speed range and a faster response speed in the high speed range.

[0052] If the shutdown conditions are not met in step S209, proceed to the next control cycle and return to step S204.

[0053] In embodiments of the present invention, such as Figure 3 As shown, if the shutdown adjustment is met, the process ends; if the shutdown condition is not met, the process enters the next control cycle after completing the entire control cycle, returns to step S204, and repeats the above steps S204~S208.

[0054] The control method for refrigerant circulation systems provided in this embodiment combines look-ahead collaborative control of model predictive control algorithms, dynamic superheat adaptive setting of fuzzy neural networks, fault prediction of long short-term memory networks, and step-by-step protection logic of variable frequency compressors to achieve intelligent control from passive response to active prediction, thereby improving the adaptability, operational continuity, and reliability of the refrigerant circulation system.

[0055] This embodiment provides a control method for a refrigerant circulation system, the process of which includes the following steps: Step S301: When the main controller receives the power-on command, it reads the high-pressure sensor value and the low-pressure sensor value. When the pressure difference between the high-pressure sensor value and the low-pressure sensor value is greater than the starting pressure difference threshold, it controls the electronic expansion valve to open to the pressure difference opening degree until the pressure difference falls back to the safe range.

[0056] Please see details Figure 2 Step S201 of the illustrated embodiment will not be described again here.

[0057] Step S302: Before starting the variable frequency compressor, control the electronic expansion valve to open to the initial opening degree and maintain it for a preset opening time.

[0058] Please see details Figure 2 Step S202 of the illustrated embodiment will not be described again here.

[0059] Step S303: After the variable frequency compressor starts, a working mode decision is made and the compressor switches to the corresponding working mode.

[0060] Specifically, step S303 includes: Step S3031: If the forced dehumidification input signal is valid, switch to forced dehumidification mode; Step S3032: If the forced dehumidification input signal is invalid and the indoor temperature is greater than or equal to the temperature threshold, then switch to cooling mode. Step S3033: After exiting the cooling mode, if the indoor humidity is greater than or equal to the humidity threshold, switch to the dehumidification mode.

[0061] In this embodiment of the invention, if the external forced dehumidification input signal is valid, for example, if the digital input port DI5 is at a high level, then the forced dehumidification mode is switched and the inverter compressor runs at full frequency.

[0062] If the external forced dehumidification input signal is invalid, the indoor temperature T_in is detected. If the indoor temperature T_in is greater than or equal to the temperature threshold T_set + ΔT_db, the system switches to cooling mode. The temperature threshold is the sum of the temperature setting threshold T_set and the temperature hysteresis ΔT_db, which defaults to 2.0℃.

[0063] After exiting cooling mode, the indoor humidity H_in is detected. If the indoor humidity H_in is greater than or equal to the humidity threshold H_set + ΔH_db, the system switches to dehumidification mode. The humidity threshold is the sum of the humidity setting threshold H_set and the humidity hysteresis ΔH_db, which is 5% by default.

[0064] It should be noted that the forced dehumidification mode is an intelligent mode that runs automatically and stops when the set humidity is reached, while the dehumidification mode is a manual mode that ignores sensor readings and runs continuously at full power.

[0065] Through an automatic switching mechanism for operating modes, the system responds to dehumidification needs and adaptively adjusts temperature and humidity. When the temperature exceeds the standard, it prioritizes cooling to avoid excessive dehumidification, and when the humidity exceeds the standard, it prioritizes dehumidification to maintain a dry environment and avoid energy efficiency losses caused by frequent start-stop of the inverter compressor.

[0066] Step S304: Within the control cycle, the overheating setpoint output by the fuzzy neural network is used as the overheating tracking target value. The system output is predicted using the model predictive control algorithm, and the optimal control action is solved by rolling optimization with the goal of minimizing the cost function.

[0067] Specifically, in step S304 above, the system output is predicted using a model predictive control algorithm, and the optimal control action is solved through rolling optimization with the goal of minimizing the cost function. This includes: Step S3041: Read system status parameters; Step S3042: Using the system state parameters as initial conditions, predict the system output using a lightweight system response model; Step S3043: Substitute the predicted sequence into the cost function, take the control sequence corresponding to the minimum cost function as the optimal control sequence, and extract the first control action as the optimal control action.

[0068] In this embodiment of the invention, rolling time-domain optimization is performed within each control cycle (control cycle Δt = 1 second).

[0069] First, the status is read, including the variable frequency compressor frequency f(k), electronic expansion valve opening v(k), exhaust temperature T_dis(k), suction superheat SH(k), and evaporation temperature T_evap(k).

[0070] Then, model prediction is performed using a lightweight system response model (e.g., first-order inertia plus pure time delay) to predict the system output within the next Np = 60 seconds:

[0071] in, The predicted system state (frequency change, electronic expansion valve opening change), k is the current control time, and i is the prediction step number (1~60), i=1, 2, ...Np. To predict the system output at time k+i, f_model is the system response model function. Let u(k) = [Δf(k), Δv(k)] be the system output at time k+i-1, and u(k) = [Δf(k), Δv(k)] be the control input (frequency, electronic expansion valve opening). u(k+i-1) is the control input at time k+i-1.

[0072] Solve for the cost function minimization, the cost function is as follows:

[0073] Where J(k) is the cost function, w1, w2, and w3 are weighting coefficients that control the overheat tracking accuracy, the smoothness of the control action, and the priority of the anti-frost constraint, respectively, and SH_target is the overheat setting value. The model predicts the superheat at second k+i, where T_evap_min is the lower limit of the evaporation temperature, T_evap(k+i|k) is the evaporation temperature at time k+i, and Δu(k+i-1)=u(k+i-1)-u(k+i-2) is the control increment penalty term to prevent drastic actions.

[0074] For example, w1=1.0, w2=0.5, w3=10.0, T_evap_min=1℃.

[0075] like Figure 4 As shown, MPC prediction has the ability to anticipate changes in advance. The MPC prediction curve is always above the actual response curve, enabling forward-looking pre-adjustment.

[0076] By using model predictive control algorithms, the system output is predicted based on system state parameters, and the optimal control action is solved through rolling optimization, thereby achieving look-ahead control and shortening the system response delay.

[0077] It should be noted that, due to disturbances in the actual operation of the system, the state of the next cycle may deviate from the predicted value. Therefore, it is not necessary to execute all planned actions, but only the first control action u(k) obtained by optimization is executed.

[0078] This approach avoids the accumulation of errors in long-cycle feedforward control. Each cycle is re-optimized based on the latest real-time status to balance foresight and anti-interference capabilities, adapting to the dynamic changes in the refrigerant circulation system.

[0079] Specifically, the fuzzy neural network includes an input layer, a fuzzification layer, a fuzzy inference layer, and a defuzzification layer. The fuzzy neural network outputs an overheating setpoint in the following manner: Step S304a: Receive node variables using the input layer; Step S304b: Use the fuzzing layer to fuzzify the node variables received by the input layer; Step S304c: Use the fuzzy inference layer to perform fuzzy inference based on the fuzzification processing result and the stored fuzzy rules, and output the rule trigger strength; Step S304d: The output of the fuzzy inference layer is converted into an overheat setting value using the defuzzification layer.

[0080] In this embodiment of the invention, the input layer receives node variables, which are evaporation temperature T_evap, ambient temperature T_amb, ambient humidity H_amb, and variable frequency compressor frequency f.

[0081] The fuzzification layer uses Gaussian membership functions to convert the node variables received by the input layer into membership degrees between 0 and 1. Each input node variable corresponds to 3 Gaussian membership functions, representing the degree to which the node variable belongs to the three fuzzy sets of "low", "medium" and "high".

[0082] The Gaussian membership function is shown in the following equation: μ_Ai_j(xi)=exp(-(xi-c_ij)² / (2σ_ij²)) Where μ_Ai_j(xi) is the Gaussian membership degree, xi is the input node variable, c_ij is the center of the Gaussian function, and σ_ij is the width of the Gaussian function. The closer xi is to c_ij, the closer μ_Ai_j(xi) is to 1, which means it belongs to the fuzzy set. The farther xi is from c_ij, the closer μ_Ai_j(xi) is to 0, which means it gradually does not belong to the fuzzy set.

[0083] For example, after fuzzing, the evaporation temperature of 8°C is given as "low temperature 0.2, medium temperature 0.9, high temperature 0.03", which means that the evaporation temperature "is basically medium temperature, slightly cold, and not high temperature". This is only an example and is not intended to be a limitation.

[0084] The fuzzy inference layer combines the membership degrees of the input variables output by the fuzzification layer to calculate the trigger strength of each fuzzy rule, thereby determining the matching strength between the current working condition and each fuzzy rule. Each node variable corresponds to three membership functions, and each node corresponds to one fuzzy rule. This layer has a total of 3... 4 =27 nodes.

[0085] For the k-th rule, the formula for calculating its trigger strength is as follows: ω_k=Π_{i=1}^{4}μ_Ai_ji(xi) Where ω_k is the trigger strength. The closer ω_k is to 1, the higher the degree of matching between the current working condition and the fuzzy rule. The closer ω_k is to 0, the lower the degree of matching between the current working condition and the fuzzy rule.

[0086] The defuzzification layer converts the output of the fuzzy inference layer into a precise overheat setting value using a weighted average method, as shown in the following expression: SH_target=(Σ_{k=1}^{27}ω_k×θ_k) / (Σ_{k=1}^{27}ω_k) Where SH_target is the overheat setting value, θ_k is the output parameter corresponding to the k-th rule, Σ_{k=1}^{27} is the weighted sum of the 27 rules, and Σ_{k=1}^{27} is the sum of the trigger strength of the 27 rules.

[0087] By dynamically outputting the superheat setpoint through a fuzzy neural network, the traditional fixed superheat setpoint is replaced. The superheat setpoint is dynamically adjusted according to operating conditions such as evaporation temperature, ambient temperature and humidity, and variable frequency compressor frequency, transforming from static to dynamic adaptive, taking into account energy efficiency, heat exchange efficiency, and anti-liquid slugging requirements.

[0088] Furthermore, when there is an error between the overheat setting value output by the fuzzy neural network and the actual overheat, the output parameters are fine-tuned using the backpropagation algorithm through online learning. The parameter update expression is as follows:

[0089] Where θ_k(t+1) is the updated output parameter, θ_k(t) is the output parameter before the update, E is the error between the superheat setpoint and the actual superheat, E=(SH_target-SH_actual)², and η is the learning rate, η=0.01.

[0090] Step S305: Use a long short-term memory network to predict the probability that the exhaust temperature exceeds the alarm threshold, and decide the system state based on the probability that the exhaust temperature exceeds the alarm threshold.

[0091] Specifically, step S305 above, based on the probability of the exhaust temperature exceeding the alarm threshold, determines the system status, including: Step S3051: If the probability of the exhaust temperature exceeding the alarm threshold is greater than or equal to the first probability threshold, the control system enters the frequency limiting preparation state in advance. In step S3052, if the probability that the exhaust temperature exceeds the alarm threshold is greater than or equal to the second probability threshold, the control system will perform frequency reduction action in advance.

[0092] In this embodiment of the invention, the Long Short-Term Memory (LSTM) network runs on an edge gateway and achieves selective memorization and forgetting of data through three gating structures: a forget gate, an input gate, and an output gate.

[0093] The forget gate determines how much information needs to be discarded, and its expression is: f t =σ(W f ·[h t-1 x t ]+b f ), where f t For the output of the forget gate, f t Forget when the value is close to 0, retain when it is close to 1, thus filtering out unimportant information. fLet h be the weight matrix of the forget gate. t-1 Let x be the hidden state at time t-1. t Let b be the input variable at time t. f This is the bias term for the forget gate.

[0094] The input gate filters which information is stored in the cell state; its expression is: i t =σ(W i ·[h t-1 x t ]+b i ), where i t W is the output of the input gate. i Let b be the weight matrix of the input gate. i This is the bias term for the input gate.

[0095] The candidate state extracts candidate content for updating long-term memory from the input data at the current moment, and its expression is: ,in, As a candidate state, W C Let b be the weight matrix of the candidate states. C The bias term for the candidate state.

[0096] Cell state updates combine the long-term memory from the previous time step with the information from the current time step to form updated long-term memory, the expression of which is: , where C t This represents the updated cell state. For element-wise multiplication, For long-term memory of the previous moment, This refers to the information at the current moment.

[0097] The output gate determines how much information from the current cell state needs to be output to the hidden state; its expression is: o t =σ(W o ·[h t-1 x t ]+b o ), where o t W is the output of the output gate. o Let b be the weight matrix of the output gate. o This is the bias term for the output gate.

[0098] The hidden state, as the final output of the Long Short-Term Memory (LSTM) network, is combined with the long-term memory of the output gate and cell state to obtain the output result, which is expressed as: , where h t It is in a hidden state.

[0099] The output layer of the Long Short-Term Memory (LSTM) network predicts the probability that the exhaust temperature will exceed the alarm threshold in the next 30 seconds, expressed as: ,in, Here, t+30 represents the predicted probability value, and t+30 is the prediction time window, i.e., predicting the next 30 seconds. p This is the bias term for the output layer.

[0100] The probability that the exhaust temperature exceeds the alarm threshold If the probability is greater than or equal to the first probability threshold of 0.6, the edge gateway sends a command to the main controller, and the control system enters the frequency limiting preparation state in advance.

[0101] The probability that the exhaust temperature exceeds the alarm threshold If the value is greater than or equal to the second probability threshold of 0.85, the edge gateway sends a command to the main controller, and the control system performs frequency reduction in advance.

[0102] like Figure 5 As shown, the LSTM prediction curve is always higher than the actual temperature curve, indicating its high fitting accuracy.

[0103] By predicting the probability of exhaust temperature exceeding the alarm threshold based on the long short-term memory network, protection decisions are made in stages to achieve intelligent control with early prediction and staged intervention. The control system enters the frequency limiting preparation state in advance to prevent the frequency from rising further, and the control system executes frequency reduction actions in advance to reduce the load.

[0104] Step S306: When any one of the following is detected to reach the corresponding threshold: exhaust temperature, high pressure, or variable frequency compressor current, the corresponding actions are executed according to the pre-set variable frequency compressor step protection logic.

[0105] Specifically, the first-level protection strategy of the variable frequency compressor's stepped protection logic is frequency limiting. That is, if the monitored exhaust temperature T_dis ≥ frequency limiting temperature threshold T_limit (e.g., 100℃), or the monitored high pressure P_high ≥ frequency limiting pressure threshold P_limit (e.g., 32 Bar), or the monitored variable frequency compressor current I_dis ≥ frequency limiting current threshold I_limit, the action is to prohibit frequency increase and maintain the current frequency.

[0106] The second-level protection strategy of the variable frequency compressor's stepped protection logic is frequency reduction. That is, if the monitored exhaust temperature T_dis ≥ frequency reduction temperature threshold T_reduce (e.g., 110℃) or the monitored high pressure P_high ≥ frequency reduction P_reduce (e.g., 35 Bar), the action is to reduce the frequency by 3Hz every 10 seconds, down to the minimum frequency.

[0107] The third-level protection strategy of the variable frequency compressor's stepped protection logic is to shut down the compressor. That is, if the monitored exhaust temperature T_dis is greater than or equal to the shutdown temperature threshold T_stop (e.g., 115℃) for 10 seconds, the action is to shut down the variable frequency compressor and report the fault.

[0108] like Figure 5 As shown, three safety zones are defined: frequency limiting, frequency reduction, and shutdown, forming a gradient protection logic.

[0109] By adding a high and low evaporation temperature forced intervention logic to the electronic expansion valve, the opening of the electronic expansion valve can be increased or decreased to achieve active intervention, maintain the evaporation temperature within a safe range, and ensure the safe and reliable operation of the system.

[0110] Step S307: Force intervention is performed on the electronic expansion valve to control the high and low evaporation temperatures.

[0111] Specifically, step S307 includes: Step S3071: If the evaporation temperature is less than or equal to the low evaporation temperature setting value and continues for a first preset time, then increase the opening of the electronic expansion valve. In step S3072, if the evaporation temperature is greater than or equal to the high evaporation temperature setting value and continues for a second preset duration, the opening of the electronic expansion valve is reduced.

[0112] In this embodiment of the invention, a forced intervention logic is added to the electronic expansion valve based on the basic superheat PID regulation. When the evaporation temperature T_evap ≤ the low evaporation temperature setpoint T_evap_low - 2℃, and this continues for a first preset duration (60 seconds), the opening of the electronic expansion valve is increased at a fixed rate, i.e., v(k+1) = v(k) + Δv_open, where v(k+1) is the expansion valve opening at time k+1, v(k) is the expansion valve opening at time k, and Δv_open is the valve opening rate, until the evaporation temperature T_evap > the low evaporation temperature setpoint T_evap_low. The default values ​​are T_evap_low = 3℃ and Δv_open = 10 steps / second.

[0113] When the evaporation temperature T_evap ≥ the high evaporation temperature setpoint T_evap_high + 1.5℃, and this continues for the second preset duration (60 seconds), the opening of the electronic expansion valve is reduced at a fixed rate, i.e., v(k+1) = v(k) - Δv_close, where v(k+1) is the expansion valve opening at time k+1, v(k) is the expansion valve opening at time k, and Δv_close is the valve closing rate, until the evaporation temperature T_evap < the high evaporation temperature setpoint T_evap_high. The default values ​​are T_evap_high = 15℃ and Δv_close = 10 steps / second.

[0114] By adding a high and low evaporation temperature forced intervention logic to the electronic expansion valve, the opening of the electronic expansion valve can be increased or decreased to achieve active intervention, maintain the evaporation temperature within a safe range, and ensure the safe and reliable operation of the system.

[0115] Step S308: Dynamically adjust the frequency change step size of the variable frequency compressor according to the frequency of the variable frequency compressor.

[0116] Specifically, the frequency change step size S_f of the variable frequency compressor is dynamically adjusted according to the frequency f of the variable frequency compressor, and its expression is: S_f=S_min+(S_max-S_min)×(f-f_min) / (f_max-f_min).

[0117] Where S_f is the frequency change step size of the variable frequency compressor, S_min is the lower limit of the step size ratio, S_max is the upper limit of the step size ratio, f is the frequency of the variable frequency compressor, f_min is the minimum frequency of the variable frequency compressor, and f_max is the maximum frequency of the variable frequency compressor.

[0118] For example, S_min=0.3Hz / s, S_max=1.0Hz / s, f_min=30Hz, f_max=50Hz.

[0119] This method allows the variable frequency compressor to achieve smoother frequency changes in the low-speed range, avoiding over-adjustment, and achieve faster response speed in the high-speed range, meeting the needs of heavy loads.

[0120] In addition, when the variable frequency compressor needs to cross the resonance interval [f_jump_start, f_jump_end] to increase the frequency from the current frequency to the target frequency f_target, the variable frequency compressor stays running at f_jump_start for Δt_hold=4 seconds, and then directly jumps to f_jump_end+1Hz.

[0121] By setting the step size of the variable frequency compressor frequency change to adaptively adjust with the frequency of the variable frequency compressor, the inherent defects of the traditional fixed step size are solved.

[0122] If the shutdown conditions are not met in step S309, proceed to the next control cycle and return to step 304, repeating steps S304 to S308.

[0123] The control method for refrigerant circulation systems provided in this embodiment has been verified to have the following beneficial effects: (1) Improved energy efficiency: The annual performance factor (APF) increased from 3.2 to 4.1, an increase of about 28%; annual electricity consumption was reduced by about 25%. (2) Enhanced stability: The system response delay was reduced from 15-25 seconds to 3-8 seconds, a reduction of about 70%; the overheat overshoot was reduced from 12℃ to 4.5℃, a reduction of 62.5%; (3) Improved reliability: The LSTM fault prediction accuracy reached 94.2% (30 seconds in advance), with an average early warning time of 25 seconds; the number of unplanned downtimes decreased from 12-18 times per year to 2-4 times, a reduction of about 75%; the MTBF (Mean Time Between Failures) of the variable frequency compressor increased from 8,000 hours to 14,000 hours, an increase of 75%; (4) Extended lifespan: Differential pressure pre-balanced start-up reduces starting current surge by about 60% and bearing wear by 65%; electronic expansion valve forced intervention eliminates the risk of frosting and liquid slugging; dynamic step size and frequency jump point control reduce overall noise by about 7 dB(A) (sound energy reduction of about 80%), extending the design life of the variable frequency compressor from 8 years to 12 years. (5) Reduced total life cycle cost: Annual maintenance costs are reduced by 50%, and the total life cycle cost over 15 years is reduced from approximately RMB 42,000 to approximately RMB 30,000, a reduction of approximately 28.6%.

[0124] This embodiment also provides a control device for a refrigerant circulation system, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0125] This embodiment provides a control device for a refrigerant circulation system, which includes a variable frequency compressor, an electronic expansion valve, and a main controller, such as... Figure 6 As shown, it includes: The first control module 601 is used to read the high pressure sensor value and the low pressure sensor value when the main controller receives the power-on command. When the pressure difference between the high pressure sensor value and the low pressure sensor value is greater than the start pressure difference threshold, it controls the electronic expansion valve to open to the pressure difference opening degree until the pressure difference falls back to the safe range. The second control module 602 is used to control the electronic expansion valve to open to the initial opening degree and maintain the preset opening time before the variable frequency compressor starts. The switching module 603 is used to make a working mode decision after the inverter compressor starts and switch to the corresponding working mode. The working modes include forced dehumidification mode, cooling mode and dehumidification mode. The solver module 604 is used to predict the system output using the overheat setpoint output by the fuzzy neural network as the overheat tracking target value within the control cycle, and to solve for the optimal control action by rolling optimization with the goal of minimizing the cost function. The decision module 605 is used to predict the probability of exhaust temperature exceeding the alarm threshold using a long short-term memory network, and to decide the system state based on the probability of exhaust temperature exceeding the alarm threshold. The system state includes frequency limiting preparation state and frequency reduction preparation state. The execution module 606 is used to execute corresponding actions according to the pre-set step protection logic of the variable frequency compressor when any one of the exhaust temperature, high pressure, or variable frequency compressor current reaches the corresponding threshold. Forced intervention module 607 is used to force intervention on the electronic expansion valve at high and low evaporation temperatures; The dynamic adjustment module 608 is used to dynamically adjust the frequency change step size of the variable frequency compressor according to the frequency of the variable frequency compressor. The loop module 609 is used to enter the next control cycle if the shutdown condition is not met, and return to the steps of predicting the system output using the model predictive control algorithm and solving the optimal control action by rolling optimization with the goal of minimizing the cost function.

[0126] In some alternative implementations, the switching module 603 includes: The first switching unit is used to switch to the forced dehumidification mode if the forced dehumidification input signal is valid. The second switching unit is used to switch to cooling mode if the forced dehumidification input signal is invalid and the indoor temperature is greater than or equal to the temperature threshold. The third switching unit is used to switch to dehumidification mode after exiting the cooling mode if the indoor humidity is greater than or equal to the humidity threshold.

[0127] In some alternative implementations, the solver module 604 includes: The reading unit is used to read system status parameters, including the frequency of the variable frequency compressor, the opening degree of the electronic expansion valve, the exhaust temperature, the suction superheat, and the evaporation temperature. The output unit is used to predict the system output using a lightweight system response model with the system state parameters as initial conditions. The extraction unit is used to substitute the predicted sequence into the cost function, select the control sequence corresponding to the minimum cost function as the optimal control sequence, and extract the first control action as the optimal control action.

[0128] In some alternative implementations, the solver module 604 further includes: The receiving unit is used to receive node variables using the input layer. The node variables include evaporation temperature, ambient temperature, ambient humidity, and variable frequency compressor frequency. The fuzzing processing unit is used to fuzzify the node variables received by the input layer using the fuzzing layer; The fuzzy inference unit is used to perform fuzzy inference based on the fuzzification processing results and stored fuzzy rules using the fuzzy inference layer, and outputs the rule trigger strength. The conversion unit is used to convert the output of the fuzzy inference layer into an overheat setting value using the defuzzification layer.

[0129] In some alternative implementations, the decision module 605 includes: The first control unit is used to enter the frequency limiting preparation state in advance if the probability that the exhaust temperature exceeds the alarm threshold is greater than or equal to the first probability threshold. The second control unit is used to execute a frequency reduction action in advance if the probability that the exhaust temperature exceeds the alarm threshold is greater than or equal to the second probability threshold.

[0130] In some alternative implementations, the mandatory intervention module 607 includes: An increasing unit is used to increase the opening of the electronic expansion valve if the evaporation temperature is less than or equal to the low evaporation temperature set value and continues for a first preset time. The reduction unit is used to reduce the opening of the electronic expansion valve if the evaporation temperature is greater than or equal to the high evaporation temperature set value and continues for a second preset time.

[0131] The control device for a refrigerant circulation system provided in this embodiment of the invention can execute the control method for a refrigerant circulation system provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the various modules and units described above are the same as in the corresponding embodiments described above, and will not be repeated here.

[0132] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0133] The following is a detailed reference. Figure 7 This diagram illustrates a suitable structural schematic for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 701, which can perform various appropriate actions and processes based on a program stored in read-only memory (ROM) 702 or a program loaded from memory 708 into random access memory (RAM) 703. The RAM 703 also stores various programs and data required for the operation of the electronic device. The processor 701, ROM 702, and RAM 703 are interconnected via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0134] Typically, the following devices can be connected to I / O interface 705: input devices 706 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 707 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 708 including, for example, magnetic tapes, hard disks, etc.; and communication devices 709. Communication device 709 allows electronic devices to exchange data via wireless or wired communication with other devices. Although Figure 7 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0135] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 709, or installed from a memory 708, or installed from a ROM 702. When the computer program is executed by the processor 701, it performs the functions defined in the control method for a refrigerant circulation system according to embodiments of the present invention.

[0136] Figure 7 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0137] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the control method for a refrigerant circulation system shown in the above embodiments is implemented.

[0138] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0139] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and all such modifications and variations fall within the scope defined by the appended invention.

Claims

1. A control method for a refrigerant circulation system, characterized in that, The refrigerant circulation system includes a variable frequency compressor, an electronic expansion valve, and a main controller; the method includes: When the main controller receives the power-on command, it reads the high-pressure sensor value and the low-pressure sensor value. When the pressure difference between the high-pressure sensor value and the low-pressure sensor value is greater than the start-up pressure difference threshold, it controls the electronic expansion valve to open to the pressure difference opening degree until the pressure difference falls back to the safe range. Before the variable frequency compressor starts, the electronic expansion valve is controlled to open to the initial opening degree and maintain the preset opening time. After the inverter compressor starts, it makes a working mode decision and switches to the corresponding working mode, which includes forced dehumidification mode, cooling mode and dehumidification mode; Within the control cycle, the overheating setpoint output by the fuzzy neural network is used as the overheating tracking target value. The system output is predicted using the model predictive control algorithm, and the optimal control action is solved by rolling optimization with the goal of minimizing the cost function. The probability of exhaust temperature exceeding the alarm threshold is predicted using a long short-term memory network, and the system state is determined based on the probability of exhaust temperature exceeding the alarm threshold. The system state includes frequency limiting preparation state and frequency reduction preparation state. When any of the exhaust temperature, high pressure, or variable frequency compressor current is detected to reach the corresponding threshold, the corresponding actions are executed according to the pre-set variable frequency compressor step protection logic. Forced intervention at high and low evaporation temperatures using electronic expansion valves; The frequency change step size of the variable frequency compressor is dynamically adjusted according to the frequency of the variable frequency compressor. If the shutdown conditions are not met, the system enters the next control cycle and returns to the steps of using the model predictive control algorithm to predict the system output and to continuously optimize and solve for the optimal control action with the goal of minimizing the cost function.

2. The method according to claim 1, characterized in that, The process of making a work mode decision and switching to the corresponding work mode includes: If the forced dehumidification input signal is valid, switch to forced dehumidification mode; If the forced dehumidification input signal is invalid and the indoor temperature is greater than or equal to the temperature threshold, then switch to cooling mode; After exiting the cooling mode, if the indoor humidity is greater than or equal to the humidity threshold, it will switch to the dehumidification mode.

3. The method according to claim 1, characterized in that, The method of using model predictive control algorithms to predict system output and then performing rolling optimization to find the optimal control action with the goal of minimizing the cost function includes: Read system status parameters, including variable frequency compressor frequency, electronic expansion valve opening, exhaust temperature, suction superheat, and evaporation temperature; Using the system state parameters as initial conditions, the system output is predicted using a lightweight system response model; Substitute the predicted sequence into the cost function, and take the control sequence that minimizes the cost function as the optimal control sequence. Extract the first control action as the optimal control action.

4. The method according to claim 1, characterized in that, The fuzzy neural network includes an input layer, a fuzzification layer, a fuzzy inference layer, and a defuzzification layer. It outputs an overheating setpoint in the following manner: The node variables are received using the input layer, including evaporation temperature, ambient temperature, ambient humidity, and variable frequency compressor frequency. The node variables received by the input layer are blurred using a blurring layer; The fuzzy inference layer performs fuzzy inference based on the fuzzification results and stored fuzzy rules, and outputs the rule trigger strength. The output of the fuzzy inference layer is converted into an overheat setting value using a defuzzification layer.

5. The method according to claim 1, characterized in that, The system state decision based on the probability of the exhaust temperature exceeding the alarm threshold includes: If the probability that the exhaust temperature exceeds the alarm threshold is greater than or equal to the first probability threshold, the control system will enter the frequency limiting preparation state in advance. If the probability that the exhaust temperature exceeds the alarm threshold is greater than or equal to the second probability threshold, the control system will perform frequency reduction action in advance.

6. The method according to claim 1, characterized in that, The forced intervention of the electronic expansion valve at high and low evaporation temperatures includes: If the evaporation temperature is less than or equal to the low evaporation temperature setting value and remains so for the first preset duration, the opening of the electronic expansion valve will be increased. If the evaporation temperature is greater than or equal to the high evaporation temperature setting value and continues for a second preset duration, the opening of the electronic expansion valve will be reduced.

7. The method according to claim 1, characterized in that, The frequency change step size of the variable frequency compressor is calculated using the following formula: S_f=S_min+(S_max-S_min)×(f-f_min) / (f_max-f_min) Where S_f is the frequency change step size of the variable frequency compressor, S_min is the lower limit of the step size ratio, S_max is the upper limit of the step size ratio, f is the frequency of the variable frequency compressor, f_min is the minimum frequency of the variable frequency compressor, and f_max is the maximum frequency of the variable frequency compressor.

8. A control device for a refrigerant circulation system, characterized in that, The refrigerant circulation system includes a variable frequency compressor, an electronic expansion valve, and a main controller; the device includes: The first control module is used to read the high-pressure sensor value and the low-pressure sensor value when the main controller receives the power-on command. When the pressure difference between the high-pressure sensor value and the low-pressure sensor value is greater than the start-up pressure difference threshold, it controls the electronic expansion valve to open to the pressure difference opening degree until the pressure difference falls back to the safe range. The second control module is used to control the electronic expansion valve to open to the initial opening degree before the variable frequency compressor starts, and to maintain the preset opening time. The switching module is used to make a working mode decision after the inverter compressor starts and switch to the corresponding working mode, including forced dehumidification mode, cooling mode and dehumidification mode. The solution module is used to predict the system output using the overheat setpoint output by the fuzzy neural network as the overheat tracking target value within the control cycle, and to solve for the optimal control action by rolling optimization with the goal of minimizing the cost function. The decision module is used to predict the probability of exhaust temperature exceeding the alarm threshold using a long short-term memory network, and to decide the system state based on the probability of exhaust temperature exceeding the alarm threshold. The system state includes frequency limiting preparation state and frequency reduction preparation state. The execution module is used to execute corresponding actions according to the pre-set step protection logic of the variable frequency compressor when any one of the exhaust temperature, high pressure, or variable frequency compressor current reaches the corresponding threshold. The forced intervention module is used to force the electronic expansion valve to operate at high and low evaporation temperatures. The dynamic adjustment module is used to dynamically adjust the frequency change step size of the variable frequency compressor according to the frequency of the variable frequency compressor. The loop module is used to enter the next control cycle if the shutdown condition is not met, and return to the steps of using the model predictive control algorithm to predict the system output and to solve for the optimal control action by rolling optimization with the goal of minimizing the cost function.

9. An electronic device, characterized in that, include: A memory and a processor are communicatively connected, the memory storing computer instructions, and the processor executing the computer instructions to perform the control method for a refrigerant circulation system as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the control method for a refrigerant circulation system as described in any one of claims 1 to 7.

Images