Ultralow-temperature fluorine road all-in-one machine

Abstract

The invention discloses an ultralow-temperature fluorine path all-in-one machine, and relates to the technical field of refrigeration. A compressor, a condenser, a liquid storage tank, a fluorine pump and an evaporator are connected to form a circulation loop; the liquid storage tank is provided with a first passage and a second passage, each passage is provided with an electronic expansion valve, the condenser is firstly connected with the first temperature sensor, the heat exchange piece is communicated with the liquid storage tank, the four-way valve is communicated with the compressor, the heat exchange piece, the evaporator and the condenser, and an electromagnetic valve and a booster pump are connected in parallel between the condenser and the four-way valve; an inlet of the first channel electronic expansion valve communicates with a heat exchange piece, meanwhile, temperature and pressure sensors are arranged at important positions of the system, and a regulation and control module receives data of the temperature and pressure sensors and adjusts the opening degree of the electronic expansion valve. According to the electronic expansion valve, the temperature and pressure in equipment are monitored, the opening degree of the electronic expansion valve is regulated and controlled through the regulation and control module, damage to the equipment due to too high or too high temperature and pressure is avoided, and the service life of the equipment is prolonged.

Description

Technical Field

[0001] The present invention relates to the technical field of refrigeration, and more specifically, to an ultra-low temperature integrated fluorine circuit machine. Background Art

[0002] At present, with the booming development of the communication industry, high-power density communication devices such as blade servers and rack servers in data centers are increasing continuously. These devices generate a large amount of heat during operation, and the heat generation of a single cabinet often exceeds 20 kW.

[0003] Traditional room-level computer room precision air conditioners are difficult to meet their uniform cooling requirements, which prompts data centers to adopt more efficient and powerful refrigeration systems. However, the high cooling demand makes the refrigeration equipment work at a high intensity for a long time, accelerating the wear of the equipment and reducing the service life of the equipment. For example, as a key heat exchange component in the refrigeration system, the heat exchanger is in a high-load state for a long time, and the heat exchange efficiency of the heat exchanger will gradually decrease. For example, the surface of its heat exchange tubes may be oxidized, corroded, etc. due to long-term contact with high-temperature refrigerant, resulting in a reduction in heat exchange performance, thereby affecting the performance of the entire refrigeration system and shortening its service life. Under different seasons and working conditions, the working conditions of the heat exchanger vary greatly. In the fluorine pump heat pipe refrigeration mode, the heat exchanger needs to efficiently transfer the cold quantity of outdoor cold air to the refrigerant; in the heat recovery mode, it is necessary to achieve effective heat exchange between the waste heat and the refrigerant; in the conventional refrigeration mode, it is necessary to ensure normal condensation or evaporation of the refrigerant in it. Frequent mode switching and different working requirements make the heat exchanger frequently withstand drastic changes in parameters such as temperature and pressure, easily generating thermal stress and mechanical stress, thus accelerating the damage of the heat exchanger and reducing its service life.

[0004] Therefore, how to extend the service life of refrigeration equipment is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention provides an ultra-low temperature integrated fluorine circuit machine, which monitors the internal temperature and pressure of the equipment, and uses a control module to adjust the opening degree of the electronic expansion valve, so as to avoid damage to the equipment caused by too high or too strong temperature and pressure, and extends the service life of the equipment.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A cryogenic fluorine circuit integrated machine, comprising: a compressor, a condenser, a liquid storage tank, a fluorine pump, an evaporator, a first temperature sensor, a second temperature sensor, a first pressure sensor, a second pressure sensor, a four-way valve, a control module, a heat exchanger, a first electronic expansion valve, a second electronic expansion valve, a solenoid valve and a booster pump. The compressor, the condenser, the liquid storage tank, the fluorine pump and the evaporator are connected to form a circulation loop. The liquid storage tank is provided with a first passage and a second passage. One end of the condenser is first connected to the first temperature sensor, and the other end of the first temperature sensor is connected to the first passage and the second passage. The first passage and the second passage extend into the stratified liquid in the liquid storage tank, and the outlets of the first passage and the second passage are arranged at the bottom of the liquid storage tank. A second temperature sensor is arranged between the evaporator and the compressor. The two ends of the heat exchanger are respectively communicated with the C port of the four-way valve and the liquid storage tank. The exhaust port of the compressor is communicated with the D port of the four-way valve. The evaporator is communicated with the S port of the four-way valve. A parallel-connected solenoid valve and booster pump are connected between one end of the condenser and the E port of the four-way valve. The booster pump makes the fluid flow from the condenser to the E port of the four-way valve. A first pressure sensor is arranged between the exhaust port of the compressor and the D port. A second pressure sensor is arranged in the top region inside the liquid storage tank. The first passage is provided with a first electronic expansion valve. The heat exchanger is communicated between the first temperature sensor and the first electronic expansion valve. The second passage is provided with a second electronic expansion valve. The control module receives the data of each temperature sensor and pressure sensor and adjusts the opening degrees of each electronic expansion valve.

[0008] Preferably, the outlet of the first passage is connected to the inlet of the evaporator through a first check valve. The first check valve allows the fluid to flow from the liquid storage tank to the evaporator. The outlet of the second passage is communicated with the fluorine pump, and the fluorine pump is communicated to the inlet of the evaporator.

[0009] Preferably, a second check valve is further included between the first temperature sensor and the first electronic expansion valve. The second check valve only allows the fluid to flow from the condenser to the first electronic expansion valve.

[0010] Preferably, the heat exchanger is communicated between the second check valve and the first electronic expansion valve. The heat exchanger is connected to the inlet of the first electronic expansion valve through a third check valve. The third check valve only allows the fluid to flow from the heat exchanger to the first electronic expansion valve.

[0011] Preferably, the oil-containing refrigerant in the liquid storage tank is transported to the evaporator through the outlet of the first passage. When the density of the lubricating oil in the liquid storage tank is less than the density of the refrigerant, the outlet of the first passage is provided with a hose, one end of which extends into the liquid storage tank and is connected to a floating ball. The floating ball floats with the change of the liquid level height inside the liquid storage tank. After the oil and liquid are stratified in the liquid storage tank, the floating ball sucks the oil-containing refrigerant in the liquid storage tank, and the other end of the hose communicates with the outlet of the first passage to transport the oil-containing refrigerant to the evaporator.

[0012] Preferably, an expansion valve opening prediction model is preset inside the regulation module, which specifically includes:

[0013] S1: Using the condenser outlet temperature, evaporator outlet temperature, compressor discharge port pressure, and the pressure in the top space of the liquid storage tank collected by the first temperature sensor, second temperature sensor, first pressure sensor, and second pressure sensor, combined with the outdoor temperature as training data, then normalizing the training data, arranging the processed variable data in a time series as input data, and constructing a 5-variable time series input layer;

[0014] S2: Perform convolution pooling and fully connected operations on the input variable data. First, perform convolution operations on the input data, and then perform pooling on the input data after convolution operations. After multiple convolution pooling operations, perform fully connected operations on the output data to complete the forward training of the expansion valve opening prediction model;

[0015] S3: Use the backpropagation technique to update the weight parameters to improve the prediction accuracy. Design independent adaptive learning rates for different parameters by calculating the first-order moment estimate and second-order moment estimate of the gradient, update the weights and biases of the convolutional layer, complete the parameter fine-tuning, and minimize the output error;

[0016] S4: Use the trained expansion valve opening prediction model to predict the openings of the first electronic expansion valve and the second electronic expansion valve after a fixed time, and adjust the openings of the first electronic expansion valve and the second electronic expansion valve according to the predicted openings.

[0017] Preferably, in S2, perform convolution operations on the variable data of the input layer in different directions. The vertical convolution extracts the features between variables, and the horizontal convolution extracts the data features of the same variable at different times. After each convolution calculation, perform a pooling operation to reduce the operation parameters. Performing convolution operations on the variable data of the input layer in different directions specifically includes:

[0018] S21: Use t convolutional kernels to perform vertical convolution on the variable data of the input layer. The vertical convolution is calculated using the following formula:

[0019] a i,j =f(w m x i,j +zm ), m = 1, 2, ..., t;

[0020] Among them, x i,j is the element in the i-th row and j-th column of the input layer data; w m represents the convolutional kernel weight, z m is the bias term of the convolutional kernel; a i,j is the element in the i-th row and j-th column of the data after convolution; f represents the activation function;

[0021] S22: After the input layer data is convolved by t convolutional kernels, it is activated by the relu function, and t neurons are output after activation. Each neuron contains a data matrix;

[0022] S23: Average pooling is performed on the t neurons. The specific formula is as follows:

[0023]

[0024] Among them, q represents the size of the pooling area, D and F are the length and width of the data matrix of a single neuron. Since the pooling kernel is vertically pooled, only the length of the neuron matrix will decrease. p i / q,j represents the element in the i / q-th row and j-th column of the output neuron matrix of the pooling layer;

[0025] S24: The neurons output by the pooling layer are then horizontally convolved by g convolutional kernels and average pooling is performed again, outputting g neurons;

[0026] S25: The neurons output after two convolutions and poolings are used as the input of the connection layer. The connection layer integrates the feature information represented by the elements of all neuron matrices into the neurons of the connection layer. Let the number of neurons in the connection layer be T. Each neuron output after multiple convolutions and poolings contains k rows and l columns of elements. Each neuron in the connection layer convolves the elements in each neuron matrix by g [k, l] convolutional kernels. The specific formula is as follows:

[0027]

[0028] Among them, x k,l is the element in the k-th row and l-th column of the input layer data; w n is the convolutional kernel weight, z n is the bias term of the convolutional kernel; the size of the convolutional kernel is the same as the size of the neuron matrix. After convolution, a single value is output, y u represents the value in each neuron of the fully connected layer, that is, the sum of the g values after convolution.

[0029] Preferably, in S25, the output layer is a linear weighted summation calculation of the output vector of the connection layer. The number of input neurons is T, and the output is The specific formula is as follows:

[0030]

[0031] Among them, is the predicted output result, is the predicted opening degree of the first electronic expansion valve, is the predicted opening degree of the second electronic expansion valve.

[0032] As can be seen from the above technical solutions, compared with the prior art, the present invention discloses a cryogenic fluorine circuit integrated machine. By setting a first temperature sensor, a second temperature sensor, a first pressure sensor, and a second pressure sensor at key positions, important parameters such as the condenser outlet temperature, the evaporator outlet temperature, the compressor discharge port pressure, and the liquid storage tank top space pressure can be accurately obtained in real time. The control module, based on these precise sensor data, utilizes an advanced convolutional neural network (CNN) algorithm and electronic expansion valve technology to achieve precise adjustment of the opening degrees of the first electronic expansion valve and the second electronic expansion valve; precise control and stable operating conditions reduce wear and damage of each component of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained according to the provided drawings without creative efforts.

[0034] Figure 1 is the structural schematic diagram provided by the present invention.

[0035] In the figure: 1 compressor; 2 condenser; 3 liquid storage tank; 4 fluorine pump; 5 evaporator; 6 first temperature sensor; 7 second temperature sensor; 8 first pressure sensor; 9 second pressure sensor; 10 four-way valve; 11 heat exchanger; 12 control module; 13 first check valve; 14 gas-liquid separator; 15 first electronic expansion valve; 16 second electronic expansion valve; 17 second check valve; 18 third check valve; 19 solenoid valve; 20 booster pump; 21 float ball. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0036] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0037] An embodiment of the present invention discloses an ultra-low temperature fluorine circuit integrated machine, as Figure 1 shown, including: a compressor 1, a condenser 2, a liquid storage tank 3, a fluorine pump 4, an evaporator 5, a first temperature sensor 6, a second temperature sensor 7, a first pressure sensor 8, a second pressure sensor 9, a four-way valve 10, a control module 12, a heat exchanger 11, a first electronic expansion valve 15, a second electronic expansion valve 16, a solenoid valve 19, and a booster pump 20. The compressor 1, the condenser 2, the liquid storage tank 3, the fluorine pump 4, and the evaporator 5 are connected to form a circulation loop; the liquid storage tank 3 is provided with a first passage and a second passage. One end of the condenser 2 is first connected to the first temperature sensor 6, and the other end of the first temperature sensor 6 is connected to the first passage and the second passage. The first passage and the second passage extend into the stratified liquid in the liquid storage tank 3, and the outlets of the first passage and the second passage are arranged at the bottom of the liquid storage tank 3; a second temperature sensor 7 is arranged between the evaporator 5 and the compressor 1; both ends of the heat exchanger 11 are respectively communicated with the C port of the four-way valve 10 and the liquid storage tank 3, the exhaust port of the compressor 1 is communicated with the D port of the four-way valve 10, the evaporator 5 is communicated with the S port of the four-way valve 10. Between one end of the condenser 2 and the E port of the four-way valve 10, a parallel-connected solenoid valve 19 and a booster pump 20 are connected. The booster pump 20 enables the fluid to flow from the condenser 2 to the E port of the four-way valve 10. A first pressure sensor 8 is arranged between the exhaust port and the D port of the compressor 1; a second pressure sensor 9 is arranged in the top region inside the liquid storage tank 3; a first electronic expansion valve 15 is arranged on the first passage, the heat exchanger 11 is communicated between the first temperature sensor 6 and the first electronic expansion valve 15, a second electronic expansion valve 16 is arranged on the second passage, and the control module 12 receives the data of each temperature sensor and pressure sensor and adjusts the opening degrees of each electronic expansion valve.

[0038] In a specific embodiment, the outlet of the first passage is connected to the inlet of the evaporator 5 through a first one-way valve 13. The first one-way valve 13 allows the fluid to flow from the liquid storage tank 3 to the evaporator 5. The outlet of the second passage is communicated with the fluorine pump 4, and the fluorine pump 4 is communicated to the inlet of the evaporator 5. The second pressure sensor 9 is arranged in the region 0.05 - 0.2 meters away from the top inside the liquid storage tank 3;

[0039] In a specific embodiment, it further includes a gas-liquid separator 14. The gas outlet of the evaporator 5 is communicated with a gas-liquid separator 14. The gas-liquid separator 14 includes a first inlet, a second inlet, and a discharge port. The first inlet is communicated with the evaporator 5, the second inlet is communicated to the S port of the four-way valve 10, and the discharge port is communicated with the suction port of the compressor 1.

[0040] In a specific embodiment, a second one-way valve 17 is further included between the first temperature sensor 6 and the first electronic expansion valve 15. The second one-way valve 17 only allows the fluid to flow from the condenser 2 to the first electronic expansion valve 15.

[0041] In a specific embodiment, a heat exchanger 11 is connected between the second one-way valve 17 and the first electronic expansion valve 15. The heat exchanger 11 is connected to the inlet of the first electronic expansion valve 15 through a third one-way valve 18. The third one-way valve 18 only allows fluid to flow from the heat exchanger 11 to the first electronic expansion valve 15.

[0042] In a specific embodiment, the oil-containing refrigerant in the liquid storage tank 3 is transported to the evaporator 5 through the outlet of the first passage. When the density of the lubricating oil in the liquid storage tank 3 is less than the density of the refrigerant, the outlet of the first passage is set as a hose, one end of which extends into the liquid storage tank 3 and is connected to the floating ball 21. The floating ball 21 floats with the change of the liquid level height inside the liquid storage tank 3. After the oil and liquid are stratified in the liquid storage tank 3, the floating ball 21 sucks the oil-containing refrigerant in the liquid storage tank 3, and the other end of the hose is connected to the outlet of the first passage to transport the oil-containing refrigerant to the evaporator 5.

[0043] In a specific embodiment, an expansion valve opening prediction model is preset inside the control module 12, which specifically includes:

[0044] S1: Using the condenser outlet temperature, evaporator outlet temperature, compressor discharge port pressure, and liquid storage tank top space pressure collected by the first temperature sensor 6, second temperature sensor 7, first pressure sensor 8, and second pressure sensor 9, combined with the outdoor temperature as training data, then normalizing the training data, arranging the processed variable data in a time series as input data, and constructing a 5-variable time series input layer;

[0045] S2: Performing convolution pooling and fully connected operations on the input variable data. First, performing convolution operations on the input data, and performing pooling on the input data after convolution operations. After multiple convolution poolings, performing fully connected operations on the output data to complete the forward training of the expansion valve opening prediction model;

[0046] S3: Using the backpropagation technique to update the weight parameters to improve the prediction accuracy, designing independent adaptive learning rates for different parameters by calculating the first moment estimate and second moment estimate of the gradient, updating the weights and biases of the convolutional layer, completing the parameter fine-tuning, and minimizing the output error;

[0047] S4: Using the trained expansion valve opening prediction model to predict the openings of the first electronic expansion valve and the second electronic expansion valve after a fixed time, and adjusting the openings of the first electronic expansion valve and the second electronic expansion valve according to the predicted openings.

[0048] In a specific embodiment, in S2, convolution operations are performed on the variable data of the input layer in different directions. Vertical convolution is used to extract features between variables, and horizontal convolution is used to extract data features of the same variable at different times. After each convolution calculation, a pooling operation is performed to reduce the number of operation parameters. The convolution operations on the variable data of the input layer in different directions specifically include:

[0049] S21: Use t convolution kernels to perform vertical convolution on the variable data of the input layer. The vertical convolution is calculated using the following formula:

[0050] a i,j = f(w m x i,j + z m ), m = 1, 2,..., t;

[0051] where x i,j is the element at the i-th row and j-th column of the input layer data; w m represents the convolution kernel weight, z m is the bias term of the convolution kernel; a i,j is the element at the i-th row and j-th column of the data after convolution; f represents the activation function;

[0052] S22: The input layer data is activated by the relu function after convolution calculation with t convolution kernels. After activation, t neurons are output, and each neuron contains a data matrix;

[0053] S23: Perform average pooling on the t neurons. The specific formula is as follows:

[0054]

[0055] where q represents the size of the pooling area, D and F are the length and width of the data matrix of a single neuron. Since the pooling kernel is vertical pooling, only the length of the neuron matrix will decrease. p i / q,j represents the element at the i / q-th row and j-th column of the output neuron matrix of the pooling layer;

[0056] S24: The neurons output by the pooling layer are then subjected to horizontal convolution by g convolution kernels and average pooling again, and g neurons are output;

[0057] S25: The neurons output after two convolution and pooling operations are used as the input of the connection layer. The connection layer integrates the feature information represented by the elements of all neuron matrices into the neurons of the connection layer. Let the number of neurons in the connection layer be T. Each neuron output after multiple convolution and pooling operations contains k rows and l columns of elements. Each neuron in the connection layer is convolved with g [k, l] convolution kernels on the elements in each neuron matrix. The specific formula is as follows:

[0058]

[0059] Among them, x k,l is the element at the k-th row and l-th column of the input layer data; w n is the convolution kernel weight, and z n is the bias term of the convolution kernel; the size of the convolution kernel is the same as the size of the neuron matrix, and after convolution, a single value y u represents the value in each neuron of the fully connected layer, that is, the value obtained by adding the g values after convolution.

[0060] In a specific embodiment, in S25, the output layer is a linear weighted summation calculation of the output vector of the connection layer. The number of input neurons is T, and the output is The specific formula is as follows:

[0061]

[0062] Among them, is the predicted result of the output, is the predicted result of the opening degree of the first electronic expansion valve, is the predicted result of the opening degree of the second electronic expansion valve.

[0063] The system of the present invention includes a compression refrigeration mode, a single evaporator refrigeration mode, a dual evaporator dual evaporation temperature refrigeration mode, a dual evaporator single evaporation temperature refrigeration mode, and a fluorine pump heat pipe refrigeration mode. First, it is judged which mode the system is currently in.

[0064] Compression refrigeration mode control: Close the booster pump 20; Open or close the fluorine pump 4 according to the actual situation; Open the solenoid valve 19, the first electronic expansion valve 15, and the second electronic expansion valve 16; Make the D port and the E port of the four-way valve 10 conduct, and the C port and the S port conduct; At this time, the refrigerant flow direction is: Compressor 1 → DE port of the four-way valve 10 → Solenoid valve 19 → Condenser 2 → Second electronic expansion valve 16 (and / or second check valve 17 → First electronic expansion valve 15) → Liquid storage tank 3 → First check valve 13 (or fluorine pump 4) → Evaporator 5 → Gas-liquid separator 14 → Compressor 1.

[0065] Single evaporator refrigeration mode control: Close the second electronic expansion valve 16, the solenoid valve 19, and the booster pump 20; Make the D port and the C port of the four-way valve 10 conduct, and the E port and the S port conduct; Open or close the fluorine pump 4 according to the actual situation, and open the first electronic expansion valve 15; At this time, the refrigerant flow direction is: Compressor 1 → DC port of the four-way valve 10 → Heat exchange element → Third check valve 18 → First electronic expansion valve 15 → Liquid storage tank 3 → First check valve 13 (or fluorine pump 4) → Evaporator 5 → Gas-liquid separator 14 → Compressor 1.

[0066] Dual-evaporator dual-evaporation temperature refrigeration mode control: Close solenoid valve 19; Open or close the fluorine pump 4 according to the actual situation to make the D port and C port of the four-way valve 10 conduct, and the E port and S port conduct; Open the first electronic expansion valve 15, the second electronic expansion valve 16 and the booster pump 20; At this time, the refrigerant flow direction is: Compressor 1 → DC port of four-way valve 10 → Heat exchanger 11 → Third check valve 18 → First electronic expansion valve 15 → Liquid storage tank 3 → [(First check valve 13 (or fluorine pump 4) → Evaporator 5) + (Second electronic expansion valve 16 → Condenser 2 → Booster pump 20 → ES port of four-way valve 10)] → Gas-liquid separator 14 → Compressor 1.

[0067] Dual-evaporator single-evaporation temperature refrigeration mode control: Close the booster pump 20; Open or close the fluorine pump 4 according to the actual situation to make the D port and C port of the four-way valve 10 conduct, and the E port and S port conduct; Open the first electronic expansion valve 15, the second electronic expansion valve 16 and the solenoid valve 19. At this time, the refrigerant flow direction is: Compressor 1 → DC port of four-way valve 10 → Heat exchanger 11 → Third check valve 18 → First electronic expansion valve 15 → Liquid storage tank 3 → [(First check valve 13 (or fluorine pump 4) → Evaporator 5) + (Second electronic expansion valve 16 → Condenser 2 → Solenoid valve 19 → ES port of four-way valve 10)] → Gas-liquid separator 14 → Compressor 1.

[0068] Fluorine pump 4 heat pipe refrigeration mode control: Close the booster pump 20 and open the fluorine pump 4; Make the D port and C port of the four-way valve 10 conduct, and the E port and S port conduct; Open the first electronic expansion valve 15, the second electronic expansion valve 16 and the solenoid valve 19. At this time, the refrigerant flow direction is: Fluorine pump 4 → Evaporator 5 → Gas-liquid separator 14 → SE port of four-way valve 10 → Solenoid valve 19 → Condenser 2 → [(Second check valve 17 → First electronic expansion valve 15) + (Second electronic expansion valve 16)] → Liquid storage tank 3 → Fluorine pump 4.

[0069] Temperature difference detection and control of dual-evaporator related modes

[0070] Temperature difference detection: In the dual-evaporator dual-evaporation temperature refrigeration mode or the dual-evaporator single-evaporation temperature refrigeration mode, detect the real-time temperature difference ΔT (ΔT = T 冷凝器 -T 储液罐 ) between the ambient temperature where the condenser 2 is located and the refrigerant liquid temperature in the liquid storage tank 3.

[0071] Judgment and control

[0072] When ΔT < ΔT1, control to close the solenoid valve 19, open the booster pump 20, and reduce the opening degree of the second electronic expansion valve 16. Since the temperature difference between the low-temperature and low-pressure refrigerant and the ambient temperature where the condenser 2 is located is not large, the evaporation heat transfer driving force is insufficient. At this time, it is necessary to reduce the evaporation temperature at the condenser 2. When the booster pump 20 is turned on and the opening degree of the second electronic expansion valve 16 is reduced, the refrigerant liquid flowing through the second electronic expansion valve 16 in the liquid storage tank 3 is expanded again to reduce the temperature and pressure. The booster pump 20 can increase the pressure and flow rate.

[0073] When ΔT ≥ ΔT1, control to close the booster pump 20, open the solenoid valve 19, and adjust the opening degree of the second electronic expansion valve 16 according to the refrigerant flow demand. At this time, dual evaporation temperature operation is not required, and the booster pump 20 can be turned off. When the booster pump 20 is turned off and the solenoid valve 19 is opened, the refrigeration mode with dual evaporation temperatures of the heat recovery dual evaporator is converted to "parallel dual evaporators to achieve a single evaporation temperature" operation. At this time, the condenser 2 is in parallel with the evaporator, and the second electronic expansion valve 16 is mainly used for refrigerant flow regulation.

[0074] Throttle valve adjustment in the fluorine pump heat pipe refrigeration mode

[0075] When the system is in the fluorine pump heat pipe refrigeration mode, control to adjust the opening degrees of the first electronic expansion valve 15 and the second electronic expansion valve 16 to the maximum to avoid consuming too much fluorine pump head at the throttle valve. At this time, the gas-liquid separator 14 is equivalent to an oil separator, which is beneficial for the refrigerant flowing in the system to bring the lubricating oil back into the gas-liquid separator 14 for separation and storage, reducing the lubricating oil content in the heat pipe system and improving the heat pipe heat transfer efficiency.

[0076] The various embodiments in this specification are described in a progressive manner. The key points of each embodiment are the differences from other embodiments. The same or similar parts among the various embodiments can be referred to each other. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple. For the relevant parts, refer to the description in the method section.

[0077] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A cryogenic fluorine circuit integrated machine, characterized in that, Including: A compressor, a condenser, a liquid storage tank, a fluorine pump, an evaporator, a first temperature sensor, a second temperature sensor, a first pressure sensor, a second pressure sensor, a four-way valve, a control module, a heat exchanger, a first electronic expansion valve, a second electronic expansion valve, a solenoid valve, and a booster pump. The compressor, the condenser, the liquid storage tank, the fluorine pump, and the evaporator are connected to form a circulation loop. The liquid storage tank is provided with a first passage and a second passage. One end of the condenser is first connected to the first temperature sensor, and the other end of the first temperature sensor is connected to the first passage and the second passage. The first passage and the second passage extend into the stratified liquid in the liquid storage tank, and the outlets of the first passage and the second passage are arranged at the bottom of the liquid storage tank. A second temperature sensor is arranged between the evaporator and the compressor. Both ends of the heat exchanger are respectively communicated with the C port of the four-way valve and the liquid storage tank. The exhaust port of the compressor is communicated with the D port of the four-way valve. The evaporator is communicated with the S port of the four-way valve. A parallel-connected solenoid valve and booster pump are connected between one end of the condenser and the E port of the four-way valve. The booster pump makes the fluid flow from the condenser to the E port of the four-way valve. A first pressure sensor is arranged between the exhaust port of the compressor and the D port. A second pressure sensor is arranged in the top region inside the liquid storage tank. The first passage is provided with a first electronic expansion valve. The heat exchanger is communicated between the first temperature sensor and the first electronic expansion valve. The second passage is provided with a second electronic expansion valve. The control module receives the data of each temperature sensor and pressure sensor and adjusts the opening degrees of each electronic expansion valve.

2. The cryogenic fluorine circuit integrated machine according to claim 1, characterized in that, The outlet of the first passage is connected to the inlet of the evaporator through a first one-way valve. The first one-way valve allows the fluid to flow from the liquid storage tank to the evaporator. The outlet of the second passage is communicated with the fluorine pump, and the fluorine pump is communicated to the inlet of the evaporator.

3. The cryogenic fluorine circuit integrated machine according to claim 1, wherein A second one-way valve is further included between the first temperature sensor and the first electronic expansion valve. The second one-way valve only allows the fluid to flow from the condenser to the first electronic expansion valve.

4. The cryogenic fluorine circuit integrated machine according to claim 3, characterized in that, The heat exchanger is communicated between the second one-way valve and the first electronic expansion valve. The heat exchanger and the inlet of the first electronic expansion valve are connected through a third one-way valve. The third one-way valve only allows the fluid to flow from the heat exchanger to the first electronic expansion valve.

5. A cryogenic fluorine circuit integrated machine according to claim 1, characterized in that, The oil-containing refrigerant in the liquid storage tank is transported to the evaporator through the outlet of the first passage. When the density of the lubricating oil in the liquid storage tank is less than the density of the refrigerant, the outlet of the first passage is set as a hose, one end of which extends into the liquid storage tank and is connected to a floating ball. The floating ball floats with the change of the liquid level height inside the liquid storage tank. After the oil and liquid are stratified in the liquid storage tank, the floating ball sucks the oil-containing refrigerant in the liquid storage tank, and the other end of the hose is communicated with the outlet of the first passage to transport the oil-containing refrigerant to the evaporator.

6. The cryogenic fluorine circuit integrated machine according to claim 1, wherein The control module internally presets an expansion valve opening prediction model, which specifically includes: S1: Using the condenser outlet temperature, evaporator outlet temperature, compressor discharge port pressure, and top space pressure of the liquid storage tank collected by the first temperature sensor, second temperature sensor, first pressure sensor, and second pressure sensor, combined with the outdoor temperature as training data, then normalizing the training data, arranging the processed variable data in time series as input data, and constructing a 5-variable time series input layer; S2: Perform convolution pooling and fully connected operations on the input variable data. First, perform convolution operations on the input data, and then perform pooling on the input data after convolution operations. After multiple convolution poolings, perform fully connected operations on the output data to complete the forward training of the expansion valve opening prediction model; S3: Use the backpropagation technique to update the weight parameters to improve the prediction accuracy. Design independent adaptive learning rates for different parameters by calculating the first-order moment estimate and second-order moment estimate of the gradient, update the weights and biases of the convolutional layer, complete the parameter fine-tuning, and minimize the output error; S4: Use the trained expansion valve opening prediction model to predict the openings of the first electronic expansion valve and the second electronic expansion valve after a fixed time, and adjust the openings of the first electronic expansion valve and the second electronic expansion valve according to the predicted openings.

7. The cryogenic fluorine circuit integrated machine according to claim 6, characterized in that, In S2, perform convolution operations on the variable data of the input layer in different directions. Longitudinal convolution extracts the features between variables, and horizontal convolution extracts the data features of the same variable at different times. After each convolution calculation, perform pooling operations to reduce the operation parameters. Performing convolution operations on the variable data of the input layer in different directions specifically includes: S21: Use t convolutional kernels to perform longitudinal convolution on the variable data of the input layer. The longitudinal convolution is calculated using the following formula: a i,j = f(w m x i,j + z m ), m = 1, 2, ..., t; Among them, x i,j is the element in the i-th row and j-th column of the input layer data; w m represents the convolution kernel weight, z m is the bias term of the convolution kernel; a i,j is the element in the i-th row and j-th column of the data after convolution; f represents the activation function; S22: After the input layer data is convolved by t convolutional kernels, it is activated by the relu function. After activation, t neurons are output, and each neuron contains a data matrix; S23: Perform average pooling on the t neurons. The specific formula is as follows: Among them, q represents the size of the pooling region, D and F are the length and width of the single-neuron data matrix. Since the pooling kernel performs vertical pooling, only the length of the neuron matrix will decrease. p i / q,j represents the element in the i / q-th row and j-th column of the output neuron matrix of the pooling layer; S24: The neurons output by the pooling layer are then horizontally convolved by g convolutional kernels and average pooling is performed again, outputting g neurons; S25: The neurons output after two convolution poolings are used as the input of the connection layer. The connection layer integrates the feature information represented by the elements of all neuron matrices into the neurons of the connection layer. Suppose the number of neurons in the connection layer is T. After multiple convolution poolings, each neuron output contains k rows and l columns of elements. Each neuron in the connection layer is convolved by g [k, l] convolutional kernels for the elements in each neuron matrix. The specific formula is as follows: Among them, x k,l is the element in the k-th row and l-th column of the input layer data; w n is the convolution kernel weight, and z n is the bias term of the convolution kernel; the size of the convolution kernel is the same as that of the neuron matrix, and a single value is output after convolution, and y u represents the value in each neuron of the fully connected layer, that is, the value obtained by adding the g values after convolution.

8. The cryogenic fluorine circuit integrated machine according to claim 7, characterized in that, In the above S25, the output layer performs a linear weighted summation calculation on the output vector of the connection layer. The number of input neurons is T, and the output is The specific formula is as follows: Among them, is the predicted output result, is the predicted opening degree result of the first electronic expansion valve, is the predicted opening degree result of the second electronic expansion valve.

Images