Closed cold therapy cabin environment control method

Abstract

The invention discloses a closed cold therapy cabin environment control method. S1; eliminating gross errors of sensors by adopting a Dick's criterion, wherein the sensors comprise an in-cabin oxygen sensor, a CO2 sensor, a pressure sensor, a humidity sensor, a temperature sensor and an illuminance sensor; the sensor acquires data of treatment parameters and survival parameters in the cabin and participates in environment control of the cold therapy cabin; s2; the sensor data is processed through Kalman filtering to improve the measurement precision; according to the control method, firstly, a Dick's criterion is adopted to eliminate gross errors of the sensor, then, sensor data are processed through Kalman filtering to improve the measurement precision, and a PID neural network controller optimized based on an ant lion algorithm is established to achieve real-time control over treatment parameters and survival parameters. Therefore, a patient can receive safe and effective physical treatment in the cold therapy cabin, and the situation that the environment in the cabin is not matched with a treatment mode due to sensor data errors is eliminated.

Description

Technical Field

[0001] This invention relates to the field of cryotherapy systems, specifically to a method for controlling the environment of a closed cryotherapy chamber. Background Technology

[0002] Closed cryotherapy chambers are subject to numerous environmental factors during treatment. Typically, sensors are installed in various environmental channels, and a main control system receives and processes these signals, issuing commands to control valves and regulate the chamber's environment. However, the sensors within the channels have inherent errors; components from different batches and with varying usage times result in data with significant inaccuracies. Therefore, the acquired data needs to be filtered and processed before use. Algorithms are used to achieve real-time control of treatment parameters, ensuring patients receive safe and effective physical therapy within the cryotherapy chamber. Summary of the Invention

[0003] The purpose of this invention is to provide a method for controlling the environment of a closed cryotherapy chamber, so as to solve the problem of achieving real-time control of the environment of a closed cryotherapy chamber as mentioned in the background art.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] A method for controlling the environment of a closed cryotherapy chamber includes the following steps;

[0006] S1; The Dixon criterion is used to eliminate gross errors in the sensors, which include an in-cabin oxygen sensor, CO2 sensor, pressure sensor, humidity sensor, temperature sensor, and illuminance sensor; The sensors collect data on in-cabin treatment parameters and survival parameters and participate in the environmental control of the cryotherapy chamber;

[0007] S2; Sensor data is processed using Kalman filtering to improve measurement accuracy;

[0008] S3; Establish a PID neural network controller based on the antlion algorithm to control the environment inside the chamber, so as to achieve real-time control of treatment parameters and survival parameters, and ensure that patients receive safe and effective physical therapy in the cryotherapy chamber.

[0009] Cryotherapy chamber environmental control refers to the control of treatment parameters and survival parameters to achieve safe and effective treatment.

[0010] The method of eliminating gross errors in sensors using the Dixon criterion includes the following steps:

[0011] S1-1; Collect the data; x1, x2, ..., x n Sort x from smallest to largest (1) ≤x (2) ≤x (3) ≤…≤x (n) ;

[0012] S1-2; Constructing and checking high-end outliers x (n) and low-end outlier x (1) The test statistic, due to varying sample sizes n, falls into the following clusters: 3 ≤ n ≤ 7, minimum residual value x. (1) , Test the maximum value of the residual x (n) , Minimum residual value x for testing 8≤n≤10 (1) , Test the maximum value of the residual x (n) , The minimum residual value x for testing 11≤n≤13 (1) , Test the maximum value of the residual x (n) , Minimum residual value x for the test of 14≤n (1) , Test the maximum value of the residual x (n) , The above r 10 、r' 10 r 11 、r' 11 r 21 、r' 21 r 22 、r' 22 Let it be r ij 、r' ij , where r ij It is to test for incomplete low-end outliers x (1) ,r' ij It is to test for high-end outliers in the residuals x (n) .

[0013] S1-3; Calculate r ij and r' ij The value is used to find the critical value r given a significance level α. ij (n,a), r ij (n,a) can be obtained by consulting probability statistics tables; two values ​​are used for judgment, r ij >r ij If (n,a) is a low-end outlier, it is removed; r' ij >r ij (n,a) is used to detect outliers, which are then removed. The removed outliers are the gross error values.

[0014] In step S2, Kalman filtering is used to process the sensor data after removing gross error values ​​in step S1. The estimated value at time k is obtained by using the estimated value of the state at time k-1 and the observed value of the state at time k. This can filter out environmental noise and improve the sensor measurement accuracy. The steps include the following steps.

[0015] S2-1; Establish the Karman filter equation using sensor-collected data;

[0016] X(k|k-1)=AX(k-1|k-1)+BU(k) Formula 1

[0017] P(k|k-1)=AP(k-1|k-1)A+Q Formula 2

[0018] In the formula, X(k|k-1) is the predicted value using the data collected by the sensor in the previous time, X(k-1|k-1) is the optimal result of the previous state, and since the sensor data at time k-1 is the same as at time k, then A = 1; U(k) is the control quantity of the current state; P(k|k-1) is the covariance of X(k|k-1), and P(k-1|k-1) is the covariance of X(k-1|k-1);

[0019] S2-2; The optimal estimate of the current state, corrected by the current state measurement, can be obtained using the following formula:

[0020] X(k|k)=X(k|k-1)+Kg(k)[Z(k)-HX(k|k-1)] Formula 3

[0021] In the formula, Kg(k) is the Kalman filter gain.

[0022] Xg(k)=P(k|k-1)H' / HP(k|k-1)H'+R] Formula 4

[0023] Formulas 1-4 are the Kalman gains for correcting the predictions to obtain the optimal estimate, where Q is the process noise variance and R is the observation noise variance, obtained from experimental data;

[0024] S2-3; The optimal value of the system state X(k|k) is obtained through an iterative process using the Kalman filter algorithm, therefore X(k|k) and P(k|k) need to be updated.

[0025] P(k|k)=(I-Kg(k)H)P(k|k-1) Formula 5

[0026] S2-4; Iterate the results calculated by Formulas 3 and 5 to Formulas 1 and 2, and obtain the filtered value through the iteration process of Formula 5.

[0027] The neural network constructed by S3 uses the K of the PID. p Ki K d The update method is introduced into the hidden layer structure of the neural network, so that the system has both the good dynamic characteristics of the neural network and the characteristics of the PID controller.

[0028] The parameters for treatment and survival in the cryotherapy chamber that need to be regulated include oxygen concentration, carbon dioxide concentration, and chamber pressure. The opening degree of the control terminal solenoid valves SOV_2 / SOV_3 / SOV_4 is a multi-input multi-output system. The constructed neural network is 6×9×3, with 6 input neurons forming the input layer, 9 hidden neurons forming the hidden layer, and 3 output neurons forming the output layer.

[0029] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0030] The control method of this invention first uses the Dixon criterion to eliminate gross errors in the sensor, then processes the sensor data through Kalman filtering to improve measurement accuracy, and establishes a PID neural network controller based on the antlion algorithm to achieve real-time control of treatment parameters and survival parameters, so as to ensure that patients receive safe and effective physical therapy in the cryotherapy chamber and eliminate the mismatch between the chamber environment and treatment mode caused by sensor data errors. Attached Figure Description

[0031] Figure 1 This is a schematic diagram of the circuit structure of the cryotherapy chamber environment control system of the present invention.

[0032] Figure 2 A schematic diagram of the oxygen generation channel of the present invention.

[0033] Figure 3 This is a schematic diagram of the drying and CO2 filtration channel traversal of the present invention.

[0034] Figure 4 This is a schematic diagram of the environmental replacement channel traversal channel of the present invention.

[0035] Figure 5 This is a schematic diagram of the pressure relief channel traversal of the present invention;

[0036] Figure 6 This is a schematic diagram of the PID neural network environmental control structure for the cryotherapy chamber of the present invention;

[0037] Figure 7 This is a schematic diagram of the Antlion algorithm flow of the present invention;

[0038] Figure 8 This is a schematic diagram of the optimized PID neural network structure for the Antlion algorithm of this invention. Detailed Implementation

[0039] The technical solution 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, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the protection scope of the present invention.

[0040] A method for controlling the environment of a closed cryotherapy chamber includes the following steps;

[0041] S1; The Dixon criterion is used to eliminate gross errors in the sensors, which include an in-cabin oxygen sensor, CO2 sensor, pressure sensor, humidity sensor, temperature sensor, and illuminance sensor; The sensors collect data on in-cabin treatment parameters and survival parameters and participate in the environmental control of the cryotherapy chamber;

[0042] S2; Sensor data is processed using Kalman filtering to improve measurement accuracy;

[0043] S3; Establish a PID neural network controller based on the antlion algorithm to control the environment inside the chamber, so as to achieve real-time control of treatment parameters and survival parameters, and ensure that patients receive safe and effective physical therapy in the cryotherapy chamber.

[0044] Cryotherapy chamber environmental control refers to the control of treatment parameters and survival parameters to achieve safe and effective treatment.

[0045] The method of eliminating gross errors in sensors using the Dixon criterion includes the following steps:

[0046] S1-1; Collect the data; x1, x2, ..., x n Sort x from smallest to largest (1) ≤x (2) ≤x (3) ≤…≤x (n) ;

[0047] S1-2; Constructing and checking high-end outliers x (n) and low-end outlier x (1) The test statistic, due to varying sample sizes n, falls into the following clusters: 3 ≤ n ≤ 7, minimum residual value x. (1) , Test the maximum value of the residual x (n) , Minimum residual value x for testing 8≤n≤10 (1) , Test the maximum value of the residual x (n) , The minimum residual value x for testing 11≤n≤13 (1) , Test the maximum value of the residual x (n) , Minimum residual value x for the test of 14≤n (1) , Test the maximum value of the residual x (n) , The above r 10 、r' 10 r 11 、r' 11 r 21 、r' 21 r 22 、r' 22 Let it be r ij 、r' ij , where r ij It is to test for incomplete low-end outliers x (1) ,r' ij It is to test for high-end outliers in the residuals x (n) .

[0048] S1-3; Calculate r ij and r' ij The value is used to find the critical value r given a significance level α. ij (n,a), r ij (n,a) can be obtained by consulting probability statistics tables; two values ​​are used for judgment, r ij >r ij If (n,a) is a low-end outlier, it is removed; r' ij >r ij (n,a) is used to detect outliers, which are then removed. The removed outliers are the gross error values.

[0049] In step S2, Kalman filtering is used to process the sensor data after removing gross error values ​​in step S1. The estimated value at time k is obtained by using the estimated value of the state at time k-1 and the observed value of the state at time k. This can filter out environmental noise and improve the sensor measurement accuracy. The steps include the following steps.

[0050] S2-1; Establish the Karman filter equation using sensor-collected data;

[0051] X(k|k-1)=AX(k-1|k-1)+BU(k) Formula 1

[0052] P(k|k-1)=AP(k-1|k-1)A+Q Formula 2

[0053] In the formula, X(k|k-1) is the predicted value using the data collected by the sensor in the previous time, X(k-1|k-1) is the optimal result of the previous state, and since the sensor data at time k-1 is the same as at time k, then A = 1; U(k) is the control quantity of the current state; P(k|k-1) is the covariance of X(k|k-1), and P(k-1|k-1) is the covariance of X(k-1|k-1);

[0054] S2-2; The optimal estimate of the current state, corrected by the current state measurement, can be obtained using the following formula:

[0055] X(k|k)=X(k|k-1)+Kg(k)[Z(k)-HX(k|k-1)] Formula 3

[0056] In the formula, Kg(k) is the Kalman filter gain.

[0057] Kg(k)=P(k|k-1)H' / HP(k|k-1)H'+R] Formula 4

[0058] Formulas 1-4 are the Kalman gains for correcting the predictions to obtain the optimal estimate, where Q is the process noise variance and R is the observation noise variance, obtained from experimental data;

[0059] S2-3; The optimal value of the system state X(k|k) is obtained through an iterative process using the Kalman filter algorithm, therefore X(k|k) and P(k|k) need to be updated.

[0060] P(k|k)=(I-Kg(k)H)P(k|k-1) Formula 5

[0061] S2-4; Iterate the results calculated by Formulas 3 and 5 to Formulas 1 and 2, and obtain the filtered value through the iteration process of Formula 5.

[0062] The neural network constructed by S3 uses the K of the PID. p K i K d The update method is introduced into the hidden layer structure of the neural network, enabling the system to simultaneously possess the excellent dynamic characteristics of the neural network and the characteristics of the PID controller.

[0063] The parameters for treatment and survival in the cryotherapy chamber that need to be regulated include oxygen concentration, carbon dioxide concentration, and chamber pressure. The opening degree of the control terminal solenoid valves SOV_2 / SOV_3 / SOV_4 is a multi-input multi-output system. The constructed neural network is 6×9×3, with 6 input neurons forming the input layer, 9 hidden neurons forming the hidden layer, and 3 output neurons forming the output layer.

[0064] like Figure 6The diagram shows the PID neural network environmental control structure of the cryotherapy chamber.

[0065] (1) Input layer

[0066] For a multi-input neural network system, at sensor sampling time k, its input and output are:

[0067] Un1(k)=y n (k) Formula 6

[0068] U n2 (k)=r n (k Formula 7

[0069] V ni (k)=U ni (k) Formula 8

[0070] Where n = 1, 2, 3 represents the nth control target, i = 1, 2; V ni (k) represents the output value of the i-th input layer of the n-th network, U ni (k) represents the input value of the i-th input layer of the n-th network.

[0071] (2) Hidden layer

[0072] The hidden layer represents K in the PID controller. p K i K d The input to its hidden layer is:

[0073]

[0074] The hidden layer output is:

[0075] B n1 (k)=P ni (k) Formula 10

[0076] B n2 (k)=B n2 (k-1)+P n2 (k) Formula 11

[0077] B n3 (k)=P n3 (k)-P n3 (k-1) Formula 12

[0078] (3) Output layer

[0079] The output layer contains the inputs to the control system. The inputs to this layer are:

[0080]

[0081] Where n represents the nth output layer of the neural network, ω nih These are the connection weights of the nth neural network.

[0082] The output value expression of the output layer is:

[0083]

[0084] Among them, O max O min This sets the maximum and minimum opening limits for solenoid valves SOV_2 / SOV_3 / SOV_4.

[0085] like Figure 1 The diagram shows the circuit diagram of the cryotherapy chamber's environmental control system. The aforementioned control method operates on this system, which manages the environment within the completely enclosed chamber during treatment. The patient is situated within this enclosed space, and environmental control involves regulating treatment and survival parameters to ensure safe and effective treatment. Treatment parameters include, but are not limited to, temperature, light intensity, treatment time, and oxygen levels. Survival parameters include, but are not limited to, carbon dioxide concentration, humidity, and chamber pressure. The control of treatment and survival parameters is achieved through system operating modes. These modes include, but are not limited to, oxygen generation, drying and CO2 filtration, environmental replacement, and depressurization. The oxygen generation mode consists of an oxygen generation channel, the drying and CO2 filtration mode consists of a drying and CO2 filtration channel, the environmental replacement mode consists of an environmental replacement channel, and the depressurization mode consists of a depressurization channel. The oxygen generation, drying and CO2 filtration, and environmental replacement channels are all channels for external gas to enter the chamber, sharing an evaporative heat exchanger and a drying condenser to cool and dry the gas within the pipes. The depressurization channel is the channel for the gas to exit the chamber. The oxygen generation mode controls the oxygen content inside the chamber; the drying and CO2 filtration modes control the humidity and CO2 concentration inside the chamber, respectively; the environmental exchange mode controls the exchange between the gas inside the chamber and the outside gas, i.e., gas exits from the chamber and gas enters from outside; and the pressure relief mode maintains the pressure inside the chamber at standard atmospheric pressure. The refrigerant from the evaporative heat exchanger is led to the refrigeration system to cool the gas entering the chamber. The refrigerant from the dryer condenser is led to the refrigeration system to dry the gas entering the chamber.

[0086] 1. Oxygen generation channel

[0087] like Figure 2The bold black lines in the diagram represent the oxygen generation channel. The oxygen generator (OG) draws air from the environment, passes it through an internal molecular sieve, and outputs pressurized oxygen at a concentration controlled at 93% ± 3%. This oxygen flows through the chamber's electrically adjustable solenoid valve SOV-2, and the main controller controls the oxygen flow rate by adjusting the opening of SOV-2. The pressurized oxygen, with its flow rate controlled, flows to the IN_2 port of the dryer / condenser (DCS) for filtration and drying to reduce humidity. The low-humidity pressurized oxygen output from the OUT_2 port of the dryer / condenser (DCS) flows to the IN_2 port of the evaporative heat exchanger (EHX-2), where its temperature is lowered to the chamber's ambient temperature. It is then output back into the chamber via the OUT_2 port of the evaporative heat exchanger (with the inlet solenoid valve SOV-10 open and the enhanced cryotherapy solenoid valve SOV-5 closed).

[0088] 2. Drying and CO2 filtration channels

[0089] like Figure 3 The bold black lines in the diagram represent the drying and CO2 filtration channels.

[0090] 2.1 CO2 Filtration

[0091] Air with excessive carbon dioxide concentration in the cabin is heated to the suction medium temperature range of the low-temperature oil-free air compressor LAC by the heater HTR through the cabin ventilation duct CVP-1. The air then flows through the oil-free air compressor LAC and the cabin drying and CO2 filtering solenoid valve SOV-1 (SOV-1 is open, SOV-3 is closed) to the carbon dioxide filter CO2F, where carbon dioxide in the air is adsorbed.

[0092] The filtered air flows to the IN_2 port of the dryer condenser DCS to filter and dry the air, reducing its humidity. The low-humidity air output from the OUT_2 port of the dryer condenser DCS flows to the IN_2 port of the evaporative heat exchanger EHX-2, reducing the temperature of the low-humidity air to the ambient temperature inside the chamber, and then outputs to the chamber through the OUT_2 port of the evaporative heat exchanger EHX-2 (with the inlet solenoid valve SOV-10 open and the enhanced cryotherapy solenoid valve SOV-5 closed).

[0093] 2.2 Drying

[0094] Air with excessive humidity inside the chamber is heated by the heater HTR to the suction medium temperature range of the low-temperature oil-free air compressor LAC through the chamber ventilation duct CVP-1. It then flows through the oil-free air compressor LAC, through the chamber drying and CO2 filtering solenoid valve SOV-1 (SOV-1 open, SOV-3 closed), to the CO2 filter CO2F, where carbon dioxide is adsorbed. The filtered air then flows to the IN_2 port of the dryer-condenser DCS for further filtration and humidity reduction. The low-humidity air output from the OUT_2 port of the dryer-condenser DCS flows to the IN_2 port of the evaporative heat exchanger EHX-2, where its temperature is lowered to the chamber's ambient temperature. Finally, it is output back into the chamber through the OUT_2 port of the evaporative heat exchanger EHX-2 (inlet solenoid valve SOV-10 open, enhanced cryotherapy solenoid valve SOV-5 closed).

[0095] 3. Environmental replacement channel

[0096] like Figure 4 The bold black lines in the diagram represent the environmental displacement channel traversal path. The oil-free air compressor AC draws pressurized air from the environment, which flows through the in-chamber displacement solenoid valve SOV-4. The main controller opens SOV-4, allowing the pressurized air to flow to the IN_2 port of the dryer condenser DCS for filtration and drying to reduce humidity. The low-humidity pressurized air output from the OUT_2 port of the dryer condenser DCS flows to the IN_2 port of the evaporative heat exchanger EHX-2, reducing its temperature to the ambient temperature inside the chamber. It then exits through the OUT_2 port of the evaporative heat exchanger EHX-2 back into the chamber (with the inlet solenoid valve SOV-10 open and the enhanced cooling solenoid valve SOV-5 closed).

[0097] To maintain the cabin pressure at the operating pressure, cabin pressure sensor P_1 detects a deviation from the operating pressure and initiates cabin depressurization. During depressurization, cabin air flows through cabin ventilation duct CVP-1, is heated by heater HTR to the temperature range of the suction medium of the low-temperature oil-free air compressor LAC, and is then discharged to the ambient atmosphere via cabin depressurization solenoid valve SOV-3. The operating pressure is the same as the cabin operating pressure described in section 2.1 (Oxygen Generation Channel): 0.08 MPa.

[0098] 4. Pressure relief channel

[0099] like Figure 4 The bold black lines in the diagram represent the pressure relief channel. The air inside the chamber passes through the chamber ventilation duct CVP-1, is heated by the heater HTR to the temperature range of the suction medium of the low-temperature oil-free air compressor LAC, and is then discharged to the ambient atmosphere through the chamber pressure relief solenoid valve SOV-3, thus maintaining the chamber pressure at the working pressure (≤0.08Mpa).

[0100] 5. Environmental control system control indicators

[0101] 5.1 Oxygen Concentration

[0102] The environmental control system controls the oxygen concentration inside the cabin within the oxygen-rich range through the oxygen generation channel, that is, the oxygen concentration inside the cabin is between 21% and 24%.

[0103] 5.2 CO2 (Carbon Dioxide Concentration)

[0104] The environmental control system keeps the carbon dioxide (CO2) concentration inside the cabin below 1% through drying and CO2 filtration channels.

[0105] 5.3 Humidity

[0106] The environmental control system keeps the relative humidity inside the cabin below 30% through drying and CO2 filtration channels.

[0107] 5.4. In-cabin pressure

[0108] The working pressure of the cabin is 0.08 MPa.

[0109] 6. Mathematical Model of Environmental Control Indicators

[0110] 6.1 Mathematical Model of In-Cabin Oxygen Concentration

[0111] Initial parameters inside the chamber: pressure P0, gas mass m0, relative molecular mass M0, gas constant R0, oxygen concentration n0.

[0112] After a certain time t, the parameters inside the cabin are: pressure P t gas mass m t relative molecular weight M t Gas constant R t Oxygen concentration n t oxygen mass m at time t 02t ;

[0113] The airflow q1 is output to the cabin via OUT_2 of the evaporative heat exchanger EHX-2;

[0114] The exhaust flow rate q2 output to the ambient atmosphere by the cabin depressurization solenoid valve SOV-3;

[0115] The oxygen production flow rate q3 output by the oxygen concentrator OG;

[0116] The volume V inside the cryotherapy chamber; ρ is the density of dry air, ρ o2 Dry oxygen density

[0117]

[0118] Based on the control of air intake and exhaust within the chamber, the instantaneous changes in the total amount of gas and the mass of oxygen within the chamber at a given time t can be obtained as follows:

[0119]

[0120] Wherein; dm is the mass of the gas mixture at any time; dm1 is the mass of the air entering the chamber after being cooled by the evaporator; and dm2 is the mass of the air discharged from the chamber to the atmosphere through the pressure relief valve.

[0121] The mass of the gas and the mass of oxygen inside the cabin at a certain time t are:

[0122]

[0123] Assuming that the mass of all gaseous substances except oxygen remains constant during the oxygen supply process, then the relative molecular mass and gas constant of the gas in the chamber at time t are:

[0124]

[0125] Substituting formulas 15-18 above into the gas state equation, we can obtain the cabin pressure and oxygen concentration at time t as follows:

[0126]

[0127] 6.2 Mathematical Model of CO2 Concentration Inside the Chamber

[0128] An empirical formula relating a patient's metabolic rate to their oxygen intake:

[0129]

[0130] In the formula, M represents the metabolic rate per unit patient surface area (W / m²). 2 QR respiratory entropy is the ratio of the rate of CO2 release to the rate of oxygen consumption during respiration. Human oxygen consumption rate (mL / s); A D Human body surface area (m) 2

[0131] The patient's CO2 release rate can be obtained from the above formula:

[0132]

[0133] In the formula, FR is the human body's CO2 release rate in L / s; H is the human body's height in m; W is the human body's weight in kg; and M is the human body's metabolic rate (1me = 58.1 W / m²). 2 .

[0134] For CO2 concentration control within the cryotherapy chamber, a calculation model is constructed based on the patient's CO2 release, the amount of fresh air introduced through the displacement channel, and the amount discharged through the depressurization channel. It is assumed that the air leakage within the cryotherapy chamber is zero and the chamber pressure remains constant, with the fresh air entering the chamber equal to the exhaust air volume. The equilibrium equation for the change in CO2 concentration within the chamber is obtained through mass conservation:

[0135]

[0136] In the formula, V is the internal volume of the cryotherapy chamber (m³). 3 t represents the treatment time (s) of the patient in the cryotherapy chamber; Q CVP The amount of fresh air entering the cryotherapy chamber per unit time (m³) 3 / l; Total CO2 release m from patients in cryotherapy chamber 3 / s;C i C0 represents the CO2 concentration inside the cryotherapy chamber; C0 represents the CO2 concentration of fresh air introduced from outside the cryotherapy chamber.

[0137] 7. Flow control of oxygen generation channel, displacement channel and decompression channel

[0138] The flow rate of each channel is controlled by the opening degree of SOV_2 / SOV_3 / SOV_4. The flow rate calculation formula when the valve is fully open is as follows:

[0139]

[0140]

[0141]

[0142] In the formula, F P K is the pipe geometry coefficient; for circular pipes with no change in diameter, it is taken as 1. V K is the valve flow coefficient. V =C V / 1.156; T is temperature, 293K; Z is gas compressibility coefficient, which can be obtained from a table based on temperature and pressure; X is the pressure difference ratio; P1 is the absolute pressure at the valve inlet, P2 is the absolute pressure at the valve outlet; X T Valve characteristics, critical pressure ratio; F K The adiabatic correction factor is 1;

[0143] The controlled flow rate Q can be obtained by adjusting the valve opening. SOV The formula for valve opening is:

[0144]

[0145] 8. Antlion Algorithm Optimization Principle

[0146] like Figure 7The diagram shows the antlion algorithm flowchart. The antlion algorithm is a biomimetic optimization algorithm that evolved from the behavior of trapping ants. Its core principle is to leverage the random walk strategy of ants to achieve good performance in global search, while the application of roulette wheel strategy and elite antlion strategy gives the algorithm good local optimization performance.

[0147] There are 3 roles in the algorithm;

[0148] Ants represent feasible solutions, and the global solution space is explored through the random walks of ants.

[0149] Antlion is the optimal solution, and it is updated by comparing the fitness of ants and antlions;

[0150] Elite antlions are the best overall; after all antlions are updated, the one with the highest adaptability is the elite antlion.

[0151] The elite antlion after the iteration is the optimal result.

[0152] 8.1 Initialize the ants and antlions

[0153] A certain number of ants and antlions are randomly generated, and the ant population location and fitness function are initialized as follows:

[0154]

[0155]

[0156] M Ant Ant position matrix, M OAnt Ant fitness matrix, where n is the number of ants, d is the feature dimension, and f(A) = ... nd ) represents the fitness of the ant at that location.

[0157] Similarly, the location and fitness function of the antlion population can be obtained:

[0158]

[0159]

[0160] M AntLion Ant position matrix, M OAntLion Ant fitness matrix, where n is the number of antlions, d is the feature dimension, and f(AL) nd ) represents the antlion's fitness at that location.

[0161] 8.2 Calculate the antlion fitness value

[0162] The antlion with the highest fitness is selected as the elite antlion, and an antlion is assigned to each ant according to its fitness value based on a roulette wheel.

[0163] 8.3 Constructing Traps

[0164] The ant moves around the hypersphere of the antlion. The ant moves within the solution space based on random walks, as follows:

[0165] X(t)=[0,cumsum(2r(t1)-1),cumsum(2r(t2)-1),…,cumsum(2r(t) n )-1)]

[0166] Formula 31

[0167] Where cumsum is the function for calculating the cumulative sum of the array, t is the current iteration time step, and r(t) is a user-defined random function, defined as follows:

[0168]

[0169] Rand is a random number uniformly distributed in the range [0,1]. To ensure that the ant's random wandering range is within the trap area, the ant position matrix in Formula 31 above is normalized:

[0170]

[0171] Let a represent the ant position of the i-th variable in the t-th iteration. i b i Let i be the minimum and maximum step size. Let be the maximum and minimum values ​​in the t-th iteration. To simulate antlion predation, the overall movement direction of the ants is the antlion's position:

[0172]

[0173]

[0174] Let c be the position of the j-th antlion after the t-th iteration. t d t These represent the minimum and maximum values ​​that occur during the ant's movement.

[0175] 8.4 Local Optimal Solution Exploration

[0176] The ant that falls into the trap gradually shrinks its range of movement, then:

[0177]

[0178] Where I represents the adaptive factor for the iteration number t:

[0179]

[0180] Where t represents the number of iterations, T is the maximum number of iterations, and w is a constant determined by t, which is:

[0181]

[0182] 8.5 Update Antlion Location

[0183] Calculate the fitness of the antlion and the ant in the trap. If the fitness of the antlion is greater than the fitness of the corresponding ant, the predation is considered successful, and the ant's position is updated to the antlion's position.

[0184]

[0185] 8.6 Update ant locations

[0186] Elite antlion position updates are achieved by comparing the fitness values ​​of all antlions during iteration and selecting the position with the highest fitness value. When ants randomly walk within the feasible solution space according to the random walk rules and constraints, their position iteration relationship is as follows:

[0187]

[0188] in, This represents the position of the elite antlion's trajectory during its random walk in the t-th iteration. This represents the position of the ant lion's trajectory during its random walk in the t-th iteration. The ant's position is updated by averaging the two values.

[0189] The following diagram illustrates the process of the Antlion algorithm described above, in order to find the optimal solution.

[0190] 9. Antlion algorithm optimizes PID neural network

[0191] like Figure 8 The diagram shows the PID neural network structure optimized by the antlion algorithm. Controlling the cryotherapy chamber environmental system using a PID neural network relies to some extent on the initial weight settings. Since the initial weights of the PID neural network are random values, the control effect is sometimes inferior to simple PID control. To enable rapid convergence to the target value at the start of control, the antlion algorithm is used to optimize the weights.

[0192] The control algorithm of this invention uses the Dixon criterion to eliminate gross errors in the sensor, and then processes the sensor data through Kalman filtering to improve measurement accuracy; a PID neural network controller based on the antlion algorithm is established to realize real-time control of treatment parameters and survival parameters, so as to ensure that patients receive safe and effective physical therapy in the cryotherapy chamber.

[0193] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the present invention without departing from its novel spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for controlling the environment of a closed cryotherapy chamber, characterized in that, Includes the following steps; S1; The Dixon criterion is used to eliminate gross errors in the sensors, which include an in-cabin oxygen sensor, CO2 sensor, pressure sensor, humidity sensor, temperature sensor, and illuminance sensor; The sensors collect data on in-cabin treatment parameters and survival parameters and participate in the environmental control of the cryotherapy chamber; S2; Sensor data is processed using Kalman filtering to improve measurement accuracy; S3; Establish a PID neural network controller based on the antlion algorithm optimization to control the cabin environment, so as to realize real-time control of treatment parameters and survival parameters.

2. The environmental control method for a closed cryotherapy chamber according to claim 1, characterized in that, Cryotherapy chamber environmental control refers to the control of treatment parameters and survival parameters to achieve safe and effective treatment.

3. The environmental control method for a closed cryotherapy chamber according to claim 1, characterized in that, The Dixon criterion is used to eliminate gross errors in the sensor. Includes the following steps; S1-1; Collect the data; x1, x2, ..., x n Sort x from smallest to largest (1) ≤x (2) ≤x (3) ≤…≤x (n) ; S1-2; Constructing and checking high-end outliers x (n) and low-end outlier x (1) The test statistic, due to varying sample sizes n, falls into the following clusters: 3 ≤ n ≤ 7, minimum residual value x. (1) , Test the maximum value of the residual x (n) , Minimum residual value x for testing 8≤n≤10 (1) , Test the maximum value of the residual x (n) , The minimum residual value x for testing 11≤n≤13 (1) , Test the maximum value of the residual x (n) , Minimum residual value x for the test of 14≤n (1) , Test the maximum value of the residual x (n) , The above r 10 、r' 10 r 11 、r' 11 r 21 、r' 21 r 22 、r' 22 Let it be r ij 、r' ij , where r ij It is to test for incomplete low-end outliers x (1) ,r' ij It is to test for high-end outliers in the residuals x (n) . S1-3; Calculate r ij and r' ij The value is used to find the critical value r given a significance level α. ij (n,a), r ij (n,a) can be obtained by consulting probability statistics tables; two values ​​are used for judgment, r ij >r ij If (n,a) is a low-end outlier, it is removed; r' ij >r ij (n,a) is used to detect outliers, which are then removed. The removed outliers are the gross error values.

4. The environmental control method for a closed cryotherapy chamber according to claim 1, characterized in that, In step S2, Kalman filtering is used to process the sensor data after removing gross error values ​​in step S1. The estimated value at time k is obtained by using the estimated value of the state at time k-1 and the observed value of the state at time k. This can filter out environmental noise and improve the sensor measurement accuracy. The steps include the following steps. S2-1; Establish the Karman filter equation using sensor-collected data; X(k|k-1)=AX(k-1|k-1)+BU(k) Formula 1 P(k|k-1)=AP(k-1|k-1)A+Q Formula 2 In the formula, X(k|k-1) is the predicted value using the data collected by the sensor in the previous time, X(k-1|k-1) is the optimal result of the previous state, and since the sensor data at time k-1 is the same as at time k, then A = 1; U(k) is the control quantity of the current state; P(k|k-1) is the covariance of X(k|k-1), and P(k-1|k-1) is the covariance of X(k-1|k-1); S2-2; The optimal estimate of the current state, corrected by the current state measurement, can be obtained using the following formula: X(k|k)=X(k|k-1)+Kg(k)[Z(k)-HX(k|k-1)] Formula 3 In the formula, Kg(k) is the Kalman filter gain. Kg(k)=P(k|k-1)H' / [HP(k|k-1)H'+R] Formula 4 Formulas 1-4 are the Kalman gains for correcting the predictions to obtain the optimal estimate, where Q is the process noise variance and R is the observation noise variance, obtained from experimental data; S2-3; The optimal value of the system state X(k|k) is obtained through an iterative process using the Kalman filter algorithm, therefore X(k|k) and P(k|k) need to be updated. P(k|k)=(I-Kg(k)H)P(k|k-1) Formula 5 S2-4； The results calculated by formulas 3 and 5 are iterated into formulas 1 and 2. The filtered value is obtained through the iterative process of formula 5.

5. The environmental control method for a closed cryotherapy chamber according to claim 1, characterized in that, The neural network constructed by S3 uses the K of the PID. p K i K d The update method is introduced into the hidden layer structure of the neural network, so that the system has both the good dynamic characteristics of the neural network and the characteristics of the PID controller.

6. The environmental control method for a closed cryotherapy chamber according to claim 1, characterized in that, The parameters for treatment and survival in the cryotherapy chamber that need to be regulated include oxygen concentration, carbon dioxide concentration, and chamber pressure, as well as the control terminal electrical parameters. The opening degrees of solenoid valves SOV_2 / SOV_3 / SOV_4 are determined by a 6×9×3 neural network, which is a multiple-input multiple-output system. The input layer consists of 6 input neurons, the hidden layer consists of 9 hidden neurons, and the output layer consists of 3 output neurons.

Images