Method for accurately controlling and adjusting temperature of heating non-combustion appliance

Through the combination of fuzzy control algorithm and least squares method, the problem of inaccurate temperature control and large energy consumption of heating non-burning instruments is solved, and rapid response and precise control are achieved, which reduces energy consumption and improves the safety and use of equipment. Experience.

CN119033168APending Publication Date: 2024-11-29HONGTA LIAONING TOBACCO CO LTD
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411230174.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-04
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

The existing heating-free burning appliances have problems such as excessive temperature, large energy consumption and unreasonable model in terms of temperature control, resulting in fragrance loss and low energy efficiency.

Method used

Using a combination of fuzzy control algorithm and least squares method, a rule base based on fuzzy rules is established by fuzzing the temperature, voltage, current and resistance parameters, and a relationship model between temperature and electrical parameters is established by using the least squares method to realize Precise temperature control and energy optimization.

Benefits of technology

It realizes rapid response and precise temperature control of heating equipment, reduces energy consumption, avoids the generation of harmful gases caused by excessive temperature, and improves the user experience and safety of the equipment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure SMS_1
    Figure SMS_1
  • Figure SMS_2
    Figure SMS_2
  • Figure SMS_3
    Figure SMS_3
Patent Text Reader

Abstract

The invention discloses a precise temperature control and adjustment method for a heating non-combustion appliance, and relates to the field of tobacco equipment control. According to the invention, optimization algorithms such as a fuzzy control algorithm and a least square method are adopted to improve a temperature control system of the heat-not-burn appliance. A fuzzy control algorithm is applied, the method is suitable for solving the problem that a model is not clear or has uncertainty in the control process, the least square method is applied to the method, data of a temperature sensor are precisely fitted, and a reliable temperature model is established. By means of the method, the temperature of the heater can be accurately predicted and adjusted so that the ideal heating effect can be achieved, and meanwhile the situation that harmful substances are generated or excessive energy is consumed due to the too high temperature is avoided. And by utilizing the least square method, the model parameters can be effectively adjusted, and the equipment is ensured to work in a reasonable temperature range, so that the overall use experience and safety are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of tobacco equipment control, and in particular to a method for precise temperature control and adjustment of a heat-not-burn device. Background Art

[0002] In the current heating process of non-combustion heating appliances, CN201910453766.8 is a heating method for electronic cigarettes. The heating temperature is too high, which produces a large amount of harmful gases, causing the smoke to lose its fragrance and consuming a lot of energy. The modeling is unreasonable. CN201510894747.0 Electronic cigarette and its heating atomization control method, CN201610358924.8 Electronic cigarette and its smoking method, its combustion CN202011383356.X is a heating control method for electronic cigarettes and electronic cigarettes. It cannot solve the temperature and time problems, the temperature is high, the modeling is unreasonable, the energy consumption is large, and the parameter introduction is unreasonable. CN 114009843 A is a method and system for adaptively adjusting the temperature control parameters of heating non-combustion appliances. It only realizes adaptive temperature adjustment but cannot achieve rapid adjustment and precise control of temperature. Summary of the invention

[0003] The invention proposes a new technical solution, that is, using fuzzy control algorithm and least square method and other optimization algorithms to improve the temperature control system of the heating without burning appliance.

[0004] In order to solve the above technical problems, the present invention creates a technical solution: a method for accurately controlling and adjusting the temperature of a heating-not-burning appliance, the steps of which are:

[0005] Step 1: Fuzzify the temperature, voltage, current, power and resistance parameters:

[0006] 1.1) Set the membership functions of temperature, voltage and resistance, and convert the specific parameter values ​​into the membership of the fuzzy set "low", "medium" and "high";

[0007] Temperature fuzzification: MembershipFunctionTemperature(value)→FuzzySetTemperature

[0008] Voltage fuzzification: MembershipFunctionVoltage(value)→FuzzySetVoltage

[0009] Fuzzy resistance: MembershipFunctionResistance(value)→FuzzySetResistance

[0010] Current fuzzification: MembershipFunctionCurrent(value)→FuzzySetCurrent

[0011] Power fuzzification: MembershipFunctionPower(value)→FuzzySetPower

[0012] In the fuzzification process, for each input parameter, the membership functions μT(x), μV(x), μR(x), μI(x), and μP(x) for temperature T, voltage V, resistance R, current I, and power P are defined as triangular functions or trapezoidal functions to represent the fuzzy set of each parameter;

[0013] a) The triangular membership function μtri(a,b,c;x) is defined as follows:

[0014]

[0015] Among them: a is the left endpoint, that is, the lowest value, b is the peak point, that is, the point with the highest membership degree 1, and c is the right endpoint, that is, the highest value;

[0016] b) The trapezoidal membership function μtrap(a1, a2, a3, a4; x) is defined as follows:

[0017]

[0018] Among them: a1 is the left endpoint, that is, the lowest value; a2 is the point where the flat top begins, a3 is the point where the flat top ends, and a4 is the right endpoint, that is, the highest value;

[0019] According to different parameter characteristics, we select triangular membership function or trapezoidal membership function by the following method:

[0020] The temperature T changes relatively linearly, and the temperature T adopts a triangular membership function: μT(x)=μtri(aT,bT,cT;x);

[0021] The voltage V remains stable within a certain range, so a trapezoidal membership function is used: μV(x) = μtrap(aV, bV, cV, dV; x) The resistance R changes little with the factors, so a triangular membership function is used: μR(x) = μtri(aR, bR, cR; x)

[0022] The current I changes relatively linearly, and the current I adopts a triangular membership function: μI(x)=μtri(aI,bI,cI;x)

[0023] Power P changes within a certain range. Power P uses a trapezoidal membership function: μP(x) = μ

[0024] trap(aP,bP,cP,dP;x).

[0025] 1.2) Establish a rule base based on fuzzy rules, determine the adjustment of heating power according to the input fuzzy set, and convert each input value into a fuzzy value through its corresponding membership function;

[0026] Consider that the fuzzy system is represented by a parameter vector θ, where θ contains the key parameters of all membership functions and the weights of fuzzy rules. Define a performance function J(θ). In the fuzzy system, the performance function J(θ) is a measure of the difference between the system output and the expected output. By optimizing this performance function, the parameter vector θ of the fuzzy system is adjusted to improve the system performance. θ contains the key parameters of all membership functions and the weights of fuzzy rules.

[0027] Definition of performance function J(θ):

[0028] Mean Squared Error (MSE): One of the most commonly used performance functions is the mean squared error, defined as the average of the sum of squares of the differences between the system output youtputy and the desired output ydesiredy:

[0029]

[0030] N is the number of samples, youtput,i(θ) is the output value calculated by the fuzzy system at the i-th sample according to the current parameter vector θ, and ydesired is the corresponding expected output value; the parameter θ determines the behavior of the fuzzy system: the membership function parameters and fuzzy rule weights contained in θ directly affect the reasoning process of the system, thereby determining the system output youtput;

[0031] J(θ) evaluates the performance of the current parameter θ by calculating the error between the output and the expected output, and optimizes J(θ) to guide the adjustment of θ to improve the control accuracy of the system, thereby being used to evaluate the accuracy of the system output;

[0032] The mean square error is used to evaluate the difference between the output of the fuzzy system and the expected output.

[0033] J(θ)=N1∑i=1N(yi-y^i(θ))2

[0034] Among them, yi is the actual output value of the i-th data point, y^i(θ) is the predicted value given by the fuzzy system, N is the total number of data points, and θ is adjusted by minimizing the performance function J(θ), that is, finding the entry value that minimizes J(θ) and optimizing the control performance of the fuzzy system.

[0035]

[0036] Defuzzification is to calculate a specific power adjustment value. The centroid of a fuzzy output set is calculated for each case. Let output μ output (P) be the membership function of the fuzzy output set of power adjustment obtained based on the above rule reasoning. The centroid P* of defuzzification is expressed as:

[0037] P*=∫PminPmaxuoutput(P)dPJPminPmaxPuoutput(P)dP

[0038] Pmin and Pmax are the possible ranges of power adjustment. Through this formula, a specific power adjustment command, namely P*, is obtained.

[0039] The rule base in 1.2) specifically includes:

[0040] In order to implement the control strategy, the system needs to identify two main stages: rapid temperature increase and temperature maintenance and step adjustment. In order for the system to achieve accurate temperature control in various scenarios, rules are formulated based on different input condition combinations in the rapid temperature increase stage and the temperature maintenance and step adjustment stage.

[0041] 1.2.1) Phase 1: Rapid temperature increase, the goal is to quickly increase the system temperature to the target range within 20 seconds;

[0042] Rule 1:

[0043] The current stage is "rapid temperature increase", the temperature T is far below the target value, the time t < 10 seconds, the maximum voltage and current are output, and the temperature is increased at the highest speed;

[0044] Rule 2:

[0045] The current stage is "rapid temperature increase", the temperature T is slightly lower than the target value, the time t < 20t seconds, the output voltage and current are medium, the temperature is increased rapidly, but overshoot is avoided;

[0046] Rule 3:

[0047] The current stage is "rapidly raising the temperature", the temperature T reaches the upper limit of the target value, the time t < 20t seconds, and the current and voltage are reduced to prevent the temperature from exceeding the target range;

[0048] Rule 4:

[0049] The current stage is "rapid temperature increase", the temperature T is close to the target value, the time t is between 15-20 seconds, and the current and voltage are gradually reduced to prepare for the temperature maintenance stage;

[0050] 1.2.1) Stage 2: Temperature maintenance and step adjustment

[0051] After the rapid pull-up is completed, stabilize the temperature within the target range and make fine adjustments as needed;

[0052] Rule 5:

[0053] The current stage is "temperature maintenance and step adjustment", the temperature T is stable within the target range, the current voltage and current are maintained, and the temperature is continuously stabilized;

[0054] Rule 6:

[0055] The current stage is "temperature maintenance and step adjustment". The temperature T starts to drop slightly, and the current is increased to increase the temperature slightly and maintain it at the target temperature.

[0056] Rule 7:

[0057] The current stage is "temperature maintenance and step adjustment". The temperature T is slightly higher than the target range, so the voltage is reduced and the temperature is slightly lowered.

[0058] Rule 8:

[0059] Condition: The current stage is "temperature maintenance and step adjustment". The temperature T is at the lower limit of the target range, but the current and voltage output are close to the minimum limit. Maintain the minimum current and voltage and observe the temperature changes to prevent over-adjustment.

[0060] Rule 9:

[0061] The current stage is "temperature maintenance and step adjustment". The external environment changes cause the temperature T to drop sharply, and the current and voltage are increased quickly to immediately respond to the temperature change and restore the target temperature;

[0062] Rule 10:

[0063] The current stage is "temperature maintenance and step adjustment". The system detects that there is no obvious temperature change for a long time, keeps the current setting unchanged, and makes periodic small adjustments to verify system stability;

[0064] 1.2.3) Comprehensive control rules

[0065] Rule 11:

[0066] The current stage is "temperature maintenance and step adjustment". The temperature T continues to be within the target range, but the system detects that the power P is gradually increasing, and slightly reduces the current to optimize energy consumption and maintain temperature stability;

[0067] Rule 12:

[0068] The current stage is "temperature maintenance and step adjustment". The voltage V fluctuates significantly, causing the temperature T to be unstable. The system response speed is adjusted and the current is increased to buffer the impact of voltage fluctuations on temperature.

[0069] Rule 13:

[0070] The current stage is "rapid temperature increase". The time t exceeds 20 seconds, but the temperature T still does not reach the target range. It enters the temperature maintenance and step adjustment stage and outputs appropriate adjustment commands to continue heating.

[0071] Rule 14:

[0072] The current stage is "temperature maintenance and step adjustment". It is detected that the system temperature regulation is invalid and the target temperature cannot be maintained. It switches back to the fast pull-up stage and outputs the highest voltage and current to restore the target temperature.

[0073] Step 2: Introduce the least square method to estimate the temperature model of the heating system;

[0074] 2.1) In the heating non-combustion equipment, the key parameters of the heating process are collected in real time, and the collected data are cleaned, standardized and preprocessed for feature selection; the key parameters include heating power P, voltage U, current I, resistance R, set temperature T2, current temperature T1, time constant τ, heat loss coefficient α, mass m and specific heat capacity c.

[0075] 2.2) The least square method is used to establish a relationship model between temperature and electrical parameters. The model attempts to find a set of parameters so that the predicted heating temperature is consistent with the actual heating temperature Q = m·c·(t2-t1) model;

[0076] P=ΔtQ=UI=I2R

[0077] Q represents the required heat, m and c are the mass and specific heat capacity of the smoking matrix, t2 and t1 are the set temperature and current temperature, P is the heating power, U, I, and R are the voltage, current, and resistance, respectively;

[0078] Considering that the heating process is affected by multiple factors, the model should include basic heat formulas and power calculations. The model is expanded to use the following model:

[0079] Q=m·c(t)·(t2-t1)+AQ

[0080] c(t) represents the function of the specific heat capacity of the smoke matrix changing with temperature, and ΔQ takes into account additional factors, including heat loss, heating efficiency,

[0081] The least square method is used to establish a relationship model between temperature and electrical parameters. The goal of the model is to find a set of parameters that minimizes the sum of squares of the difference between the predicted heating temperature and the actual heating temperature.

[0082] Where Q represents the required heat, m and c are the mass and specific heat capacity of the smoking substrate, t2 and t1 are the set temperature and current temperature, P is the heating power, U, I and R are the voltage, current and resistance, respectively;

[0083] Considering the temperature dependence of the internal resistance of the heater and achieving more precise temperature control, a temperature feedback mechanism is introduced in this model, that is, the heating power is adjusted according to the difference between the actual measured temperature and the target temperature:

[0084] P adjust (t)=P(t)+K p ·((t2-T actual (t))+K i ·(t2-T actual (t))dt+K d ·dtd(t2-T actual (t))

[0085] Tactual(t) is the actual measured temperature, Kp, Ki, and Kd are the proportional, integral, and differential coefficients of the PID controller, respectively.

[0086] Application of the method for precise temperature control and adjustment of heating-without-combustion appliances: Application in the startup phase of heating-without-combustion appliances: Optimize the temperature rise process by combining a fuzzy control model and a least squares model. The fuzzy control model processes the uncertainty and nonlinearity of the system and provides an adjustment strategy based on empirical rules, so that the heating appliance responds quickly and approaches the set temperature in a short time; the least squares rule is used to accurately model the relationship between the heat demand and the electrical parameters during the heating process, and the error between the predicted temperature and the actual temperature is minimized by continuously adjusting the control parameters, thereby accelerating the heating process;

[0087] After the heating device reaches a temperature close to the set temperature, stable temperature control is achieved through the dynamic adjustment capability of the fuzzy control model and the precise parameter optimization of the least squares method. When faced with temperature fluctuations, the fuzzy control model balances the heating power and heat demand by adjusting the control strategy to maintain temperature stability; the least squares method continuously corrects the parameters in the model to reflect changes under actual operating conditions and ensure that the temperature remains near the set value.

[0088] The beneficial effects of the present invention are:

[0089] 1. Apply fuzzy control algorithm. Fuzzy control algorithm is based on fuzzy logic and is suitable for dealing with unclear or uncertain models in the control process. In the field of temperature control of heating non-combustion equipment such as electronic cigarettes, fuzzy control can effectively deal with the complexity and variables of the heating process. This algorithm achieves intelligent regulation of the heating process by establishing a set of fuzzy rule bases, which can quickly respond to temperature changes and ensure the stability and accuracy of temperature control.

[0090] 2. Apply the least squares method Here, the data of the temperature sensor can be accurately fitted to establish a reliable temperature model. The model optimized by this method can accurately predict and adjust the temperature of the heater to achieve the ideal heating effect, while avoiding the generation of harmful substances or excessive energy consumption due to excessive temperature. Using the least squares method, the model parameters can be effectively adjusted to ensure that the device works within a reasonable temperature range, thereby improving the overall user experience and safety.

[0091] The present invention can handle the fuzziness and uncertainty in the temperature control process through the fuzzy control algorithm, build a fuzzy rule base to achieve intelligent adjustment of the system, quickly respond to temperature changes, and provide a stable temperature control effect. The least squares rule is used to fit the temperature sensor data and establish an accurate temperature model, and control and adjust according to the model parameters so that the temperature of the equipment reaches the expected set value and remains within a reasonable temperature range. DETAILED DESCRIPTION

[0092] A method for accurately controlling and adjusting the temperature of a heating-not-burning appliance, comprising the steps of:

[0093] Step 1: Fuzzify the temperature, voltage, current, power and resistance parameters:

[0094] 1.1) Set the membership functions of temperature, voltage and resistance, and convert the specific parameter values ​​into the membership of the fuzzy set "low", "medium" and "high";

[0095] Temperature fuzzification: MembershipFunctionTemperature(value)→FuzzySetTemperature

[0096] Voltage fuzzification: MembershipFunctionVoltage(value)→FuzzySetVoltage

[0097] Fuzzy resistance: MembershipFunctionResistance(value)→FuzzySetResistance

[0098] Current fuzzification: MembershipFunctionCurrent(value)→FuzzySetCurrent

[0099] Power fuzzification: MembershipFunctionPower(value)→FuzzySetPower

[0100] In the fuzzification process, for each input parameter, the membership functions μT(x), μV(x), μR(x), μI(x), and μP(x) for temperature T, voltage V, resistance R, current I, and power P are defined as triangular functions or trapezoidal functions to represent the fuzzy set of each parameter;

[0101] a) The triangular membership function μtri(a,b,c;x) is defined as follows:

[0102]

[0103] Among them: a is the left endpoint, that is, the lowest value, b is the peak point, that is, the point with the highest membership degree 1, and c is the right endpoint, that is, the highest value;

[0104] b) The trapezoidal membership function μtrap(a1, a2, a3, a4; x) is defined as follows:

[0105]

[0106] Among them: a1 is the left endpoint, that is, the lowest value; a2 is the point where the flat top begins, a3 is the point where the flat top ends, and a4 is the right endpoint, that is, the highest value;

[0107] According to different parameter characteristics, we select triangular membership function or trapezoidal membership function by the following method:

[0108] The temperature T changes relatively linearly, and the temperature T adopts a triangular membership function: μT(x)=μtri(aT,bT,cT;x);

[0109] The voltage V remains stable within a certain range, so a trapezoidal membership function is used: μV(x) = μtrap(aV, bV, cV, dV; x) The resistance R changes little with the factors, so a triangular membership function is used: μR(x) = μtri(aR, bR, cR; x)

[0110] The current I changes relatively linearly, and the current I adopts a triangular membership function: μI(x)=μtri(aI,bI,cI;x)

[0111] Power P changes within a certain range. Power P uses a trapezoidal membership function: μP(x) = μ

[0112] trap(aP,bP,cP,dP;x).

[0113] 1.2) Establish a rule base based on fuzzy rules, determine the adjustment of heating power according to the input fuzzy set, and convert each input value into a fuzzy value through its corresponding membership function;

[0114] Consider that the fuzzy system is represented by a parameter vector θ, where θ contains the key parameters of all membership functions and the weights of fuzzy rules. Define a performance function J(θ). In the fuzzy system, the performance function J(θ) is a measure of the difference between the system output and the expected output. By optimizing this performance function, the parameter vector θ of the fuzzy system is adjusted to improve the system performance. θ contains the key parameters of all membership functions and the weights of fuzzy rules.

[0115] Definition of performance function J(θ):

[0116] Mean Squared Error (MSE): One of the most commonly used performance functions is the mean squared error, defined as the average of the sum of squares of the differences between the system output youtputy and the desired output ydesiredy:

[0117]

[0118] N is the number of samples, youtput,i(θ) is the output value calculated by the fuzzy system at the i-th sample according to the current parameter vector θ, and ydesired is the corresponding expected output value; the parameter θ determines the behavior of the fuzzy system: the membership function parameters and fuzzy rule weights contained in θ directly affect the reasoning process of the system, thereby determining the system output youtput;

[0119] J(θ) evaluates the performance of the current parameter θ by calculating the error between the output and the expected output, and optimizes J(θ) to guide the adjustment of θ to improve the control accuracy of the system, thereby being used to evaluate the accuracy of the system output;

[0120] The mean square error is used to evaluate the difference between the output of the fuzzy system and the expected output.

[0121] J(θ)=N1∑i=1N(yi-y^i(θ))2

[0122] Among them, yi is the actual output value of the i-th data point, y^i(θ) is the predicted value given by the fuzzy system, N is the total number of data points, and θ is adjusted by minimizing the performance function J(θ), that is, finding the entry value that minimizes J(θ) and optimizing the control performance of the fuzzy system.

[0123]

[0124] Defuzzification is to calculate a specific power adjustment value. The centroid of a fuzzy output set is calculated for each case. Let output μ output (P) be the membership function of the fuzzy output set of power adjustment obtained based on the above rule reasoning. The centroid P* of defuzzification is expressed as:

[0125] P*=∫PminPmaxuoutput(P)dPJPminPmaxPuoutput(P)dP

[0126] Pmin and Pmax are the possible ranges of power adjustment. Through this formula, a specific power adjustment command, namely P*, is obtained.

[0127] The rule base in 1.2) specifically includes:

[0128] In order to implement the control strategy, the system needs to identify two main stages: rapid temperature increase and temperature maintenance and step adjustment. In order for the system to achieve accurate temperature control in various scenarios, rules are formulated based on different input condition combinations in the rapid temperature increase stage and the temperature maintenance and step adjustment stage.

[0129] 1.2.1) Phase 1: Rapid temperature increase, the goal is to quickly increase the system temperature to the target range within 20 seconds;

[0130] Rule 1:

[0131] The current stage is "rapid temperature increase", the temperature T is far below the target value, the time t < 10 seconds, the maximum voltage and current are output, and the temperature is increased at the highest speed;

[0132] Rule 2:

[0133] The current stage is "rapid temperature increase", the temperature T is slightly lower than the target value, the time t < 20t seconds, the output voltage and current are medium, the temperature is increased rapidly, but overshoot is avoided;

[0134] Rule 3:

[0135] The current stage is "rapidly raising the temperature", the temperature T reaches the upper limit of the target value, the time t < 20t seconds, and the current and voltage are reduced to prevent the temperature from exceeding the target range;

[0136] Rule 4:

[0137] The current stage is "rapid temperature increase", the temperature T is close to the target value, the time t is between 15-20 seconds, and the current and voltage are gradually reduced to prepare for the temperature maintenance stage;

[0138] 1.2.1) Stage 2: Temperature maintenance and step adjustment

[0139] After the rapid pull-up is completed, stabilize the temperature within the target range and make fine adjustments as needed;

[0140] Rule 5:

[0141] The current stage is "temperature maintenance and step adjustment", the temperature T is stable within the target range, the current voltage and current are maintained, and the temperature is continuously stabilized;

[0142] Rule 6:

[0143] The current stage is "temperature maintenance and step adjustment". The temperature T starts to drop slightly, and the current is increased to increase the temperature slightly and maintain it at the target temperature.

[0144] Rule 7:

[0145] The current stage is "temperature maintenance and step adjustment". The temperature T is slightly higher than the target range, so the voltage is reduced and the temperature is slightly lowered.

[0146] Rule 8:

[0147] Condition: The current stage is "temperature maintenance and step adjustment". The temperature T is at the lower limit of the target range, but the current and voltage output are close to the minimum limit. Maintain the minimum current and voltage and observe the temperature changes to prevent over-adjustment.

[0148] Rule 9:

[0149] The current stage is "temperature maintenance and step adjustment". The external environment changes cause the temperature T to drop sharply, and the current and voltage are increased quickly to immediately respond to the temperature change and restore the target temperature;

[0150] Rule 10:

[0151] The current stage is "temperature maintenance and step adjustment". The system detects that there is no obvious temperature change for a long time, keeps the current setting unchanged, and makes periodic small adjustments to verify system stability;

[0152] 1.2.3) Comprehensive control rules

[0153] Rule 11:

[0154] The current stage is "temperature maintenance and step adjustment". The temperature T continues to be within the target range, but the system detects that the power P is gradually increasing, and slightly reduces the current to optimize energy consumption and maintain temperature stability;

[0155] Rule 12:

[0156] The current stage is "temperature maintenance and step adjustment". The voltage V fluctuates significantly, causing the temperature T to be unstable. The system response speed is adjusted and the current is increased to buffer the impact of voltage fluctuations on temperature.

[0157] Rule 13:

[0158] The current stage is "rapid temperature increase". The time t exceeds 20 seconds, but the temperature T still does not reach the target range. It enters the temperature maintenance and step adjustment stage and outputs appropriate adjustment commands to continue heating.

[0159] Rule 14:

[0160] The current stage is "temperature maintenance and step adjustment". It is detected that the system temperature regulation is invalid and the target temperature cannot be maintained. It switches back to the fast pull-up stage and outputs the highest voltage and current to restore the target temperature.

[0161] Step 2: Introduce the least square method to estimate the temperature model of the heating system;

[0162] 2.1) In the heating non-combustion equipment, the key parameters of the heating process are collected in real time, and the collected data are cleaned, standardized and preprocessed for feature selection; the key parameters include heating power P, voltage U, current I, resistance R, set temperature T2, current temperature T1, time constant τ, heat loss coefficient α, mass m and specific heat capacity c.

[0163] 2.2) The least square method is used to establish a relationship model between temperature and electrical parameters. The model attempts to find a set of parameters so that the predicted heating temperature is consistent with the actual heating temperature Q = m·c·(t2-t1) model;

[0164] P=ΔtQ=UI=I2R

[0165] Q represents the required heat, m and c are the mass and specific heat capacity of the smoking matrix, t2 and t1 are the set temperature and current temperature, P is the heating power, U, I, and R are the voltage, current, and resistance, respectively;

[0166] Considering that the heating process is affected by multiple factors, the model should include basic heat formulas and power calculations. The model is expanded to use the following model:

[0167] Q=m·cc(t)·(t2-t1)+ΔQ

[0168] c(t) represents the function of the specific heat capacity of the smoke matrix changing with temperature, and ΔQ takes into account additional factors, including heat loss, heating efficiency,

[0169] The least square method is used to establish a relationship model between temperature and electrical parameters. The goal of the model is to find a set of parameters that minimizes the sum of squares of the difference between the predicted heating temperature and the actual heating temperature.

[0170] Where Q represents the required heat, m and c are the mass and specific heat capacity of the smoking substrate, t2 and t1 are the set temperature and current temperature, P is the heating power, U, I and R are the voltage, current and resistance, respectively;

[0171] Considering the temperature dependence of the internal resistance of the heater and achieving more precise temperature control, a temperature feedback mechanism is introduced in this model, that is, the heating power is adjusted according to the difference between the actual measured temperature and the target temperature:

[0172] P adjust (t)=P(t)+K p ·(t2-T actual (t))+K i ∫(t2-T actual (t))dt+K d ·dtd(t2-T actual (t))

[0173] Tactual(t) is the actual measured temperature, Kp, Ki, and Kd are the proportional, integral, and differential coefficients of the PID controller, respectively.

[0174] Application of the method for precise temperature control and adjustment of heating-without-combustion appliances: Application in the startup phase of heating-without-combustion appliances: Optimize the temperature rise process by combining a fuzzy control model and a least squares model. The fuzzy control model processes the uncertainty and nonlinearity of the system and provides an adjustment strategy based on empirical rules, so that the heating appliance responds quickly and approaches the set temperature in a short time; the least squares rule is used to accurately model the relationship between the heat demand and the electrical parameters during the heating process, and the error between the predicted temperature and the actual temperature is minimized by continuously adjusting the control parameters, thereby accelerating the heating process;

[0175] After the heating device reaches a temperature close to the set temperature, stable temperature control is achieved through the dynamic adjustment capability of the fuzzy control model and the precise parameter optimization of the least squares method. When faced with temperature fluctuations, the fuzzy control model balances the heating power and heat demand by adjusting the control strategy to maintain temperature stability; the least squares method continuously corrects the parameters in the model to reflect changes under actual operating conditions and ensure that the temperature remains near the set value.

[0176] Embodiment 1:

[0177] Optimization and comparative analysis of temperature control system based on fuzzy control and least squares method

[0178] 1. In this embodiment, we take a certain type of heating without burning appliance as the research object to explore the performance difference between the temperature control method of the present invention and the traditional PID control method. The experiment verifies the advantages of the method of the present invention in terms of temperature control accuracy, response speed, energy efficiency and stability through specific parameter setting and comparative analysis.

[0179] Experimental conditions and parameter settings:

[0180] Target temperature: 220℃

[0181] Initial temperature: 25℃

[0182] Allowable temperature fluctuation range: ±2℃

[0183] Sampling time interval: 1 second

[0184] Total sampling times: 1000 times

[0185] Ambient temperature: 20℃

[0186] Heating power range: 0-1500W

[0187] Control algorithm:

[0188] 1) Traditional PID control

[0189] 2) The method of combining fuzzy control with least squares method of the present invention

[0190] 2. Experimental process and methods:

[0191] 1) Traditional PID control method:

[0192] The system uses 250℃ as the target temperature and uses the traditional PID control algorithm to control the heating device.

[0193] The PID controller parameters are set as Kp=2.0K, Ki=0.5, Kd=0.

[0194] During the experiment, record the time required to reach the target temperature, the fluctuation range after the temperature stabilizes, and the actual power consumption. Observe the dynamic response of the system at different temperature stages, especially the overshoot phenomenon and stability.

[0195] 2) Control method of the present invention:

[0196] Under the same experimental conditions, the method combining fuzzy control and least square method of the present invention is adopted.

[0197] Fuzzy control part: Use membership function to perform fuzzy processing on temperature error and error change rate, and generate preliminary control strategy.

[0198] Least square method part: optimize the parameters in the control strategy through real-time data fitting to reduce the error between the system output and the expected output.

[0199] Record the same performance indicators and analyze the system's temperature regulation process and results.

[0200] 3. Experimental results and data analysis:

[0201]

[0202]

[0203] 1. Temperature response time analysis:

[0204] Traditional PID control: The system reaches the target temperature after 40 seconds. Since the PID control method responds to temperature changes more linearly, it exhibits a large hysteresis in the initial stage.

[0205] The control method of the present invention: the system quickly reaches the target temperature within 20 seconds, which is 20 seconds faster than the traditional PID control. Since the fuzzy control combined with the set rule base can adjust the heating power more quickly, the response time is reduced.

[0206] 2. Temperature fluctuation range analysis:

[0207] Traditional PID control: After the temperature reaches the target value, the system temperature fluctuation range is ±3.5°C. This large fluctuation range reflects the shortcomings of traditional PID control when facing temperature fluctuations, especially when the load changes greatly.

[0208] The control method of the present invention: after the temperature is stabilized, the fluctuation range is only ±1°C, indicating that the method of the present invention has higher accuracy and stability in temperature control.

[0209] 3. Power consumption analysis:

[0210] Traditional PID control: In the experiment, the traditional PID control system consumed 135W of power, which was mainly attributed to the system frequently adjusting power to maintain a stable state after reaching the target temperature.

[0211] The control method of the present invention: the power consumption is reduced to 100 W, indicating that the control strategy of the present invention can utilize energy more efficiently and reduce unnecessary power waste.

[0212] 4. Overshoot phenomenon analysis:

[0213] Traditional PID control: When the system reaches the target temperature, there is a slight overshoot phenomenon, and the temperature temporarily exceeds the target value, which will cause short-term temperature instability and affect the accuracy of the system.

[0214] The control method of the present invention: Since the fuzzy control can predict the system and adjust it in advance, the least square method further optimizes the parameter configuration, and the system does not have overshoot during the entire heating process, ensuring stable temperature control.

[0215] 5. System stability analysis:

[0216] Traditional PID control: Due to overshoot and large temperature fluctuations, the system stability is relatively low. Especially when the ambient temperature changes or the load fluctuates, the PID control is difficult to adjust quickly.

[0217] The control method of the present invention: the system exhibits extremely high stability, the temperature control process is smooth, and can quickly adapt to environmental changes, ensuring the accuracy and consistency of temperature control.

Claims

1. A method for accurately controlling and adjusting the temperature of a heating-not-burning appliance, characterized in that: The steps are: Step 1: Fuzzify the temperature, voltage, current, power and resistance parameters: 1.1) Set the membership functions of temperature, voltage and resistance, and convert the specific parameter values ​​into the membership of fuzzy sets "low", "medium" and "high"; Temperature fuzzification: MembershipFunctionTemperature(value)→FuzzySetTemperature Voltage fuzzification: MembershipFunctionVoltage(value)→FuzzySetVoltage Fuzzy resistance: MembershipFunctionResistance(value)→FuzzySetResistance Current fuzzification: MembershipFunctionCurrent(value)→FuzzySetCurrent Power fuzzification: MembershipFunctionPower(value)→FuzzySetPower In the fuzzification process, for each input parameter, the membership functions μT(x), μV(x), μR(x), μI(x), and μP(x) for temperature T, voltage V, resistance R, current I, and power P are defined as triangular functions or trapezoidal functions to represent the fuzzy set of each parameter; a) The triangular membership function μtri(a,b,c;x) is defined as follows: Among them: a is the left endpoint, that is, the lowest value, b is the peak point, that is, the point with the highest membership degree 1, and c is the right endpoint, that is, the highest value; b) The trapezoidal membership function μtrap(a1, a2, a3, a4; x) is defined as follows: Among them: a1 is the left endpoint, that is, the lowest value; a2 is the point where the flat top begins, a3 is the point where the flat top ends, and a4 is the right endpoint, that is, the highest value; According to different parameter characteristics, we select triangular membership function or trapezoidal membership function by the following method: The temperature T changes relatively linearly, and the temperature T adopts a triangular membership function: μT(x)=μtri(aT,bT,cT;x); The voltage V remains stable within a certain range, so a trapezoidal membership function is used: μV(x) = μtrap(aV, bV, cV, dV; x) The resistance R changes little with the factors, so a triangular membership function is used: μR(x) = μtri(aR, bR, cR; x) The current I changes relatively linearly, and the current I adopts a triangular membership function: μI(x)=μtri(aI,bI,cI;x) Power P changes within a certain range, and power P uses a trapezoidal membership function: μP(x) = μtrap(aP, bP, cP, dP; x); 1.2) Establish a rule base based on fuzzy rules, determine the adjustment of heating power according to the input fuzzy set, and convert each input value into a fuzzy value through its corresponding membership function; Consider that the fuzzy system is represented by a parameter vector θ, where θ contains the key parameters of all membership functions and the weights of fuzzy rules. Define a performance function J(θ). In the fuzzy system, the performance function J(θ) is a measure of the difference between the system output and the expected output. By optimizing this performance function, the parameter vector θ of the fuzzy system is adjusted to improve the system performance. θ contains the key parameters of all membership functions and the weights of fuzzy rules. Definition of performance function J(θ): Mean Squared Error (MSE): One of the most commonly used performance functions is the mean squared error, defined as the average of the sum of squares of the differences between the system output youtputy and the desired output ydesiredy: N is the number of samples, youtput,i(θ) is the output value calculated by the fuzzy system at the i-th sample according to the current parameter vector θ, and ydesired is the corresponding expected output value; the parameter θ determines the behavior of the fuzzy system: the membership function parameters and fuzzy rule weights contained in θ directly affect the reasoning process of the system, thereby determining the system output youtput; J(θ) evaluates the performance of the current parameter θ by calculating the error between the output and the expected output, and optimizes J(θ) to guide the adjustment of θ to improve the control accuracy of the system, thereby being used to evaluate the accuracy of the system output; The mean square error is used to evaluate the difference between the output of the fuzzy system and the expected output. J(θ)=N1∑i=1N(yi-y^i(θ))2 Among them, yi is the actual output value of the i-th data point, y^i(θ) is the predicted value given by the fuzzy system, N is the total number of data points, and θ is adjusted by minimizing the performance function J(θ), that is, finding the entry value that minimizes J(θ) and optimizing the control performance of the fuzzy system. Defuzzification is to calculate a specific power adjustment value. The centroid of a fuzzy output set is calculated for each case. Let outputμoutput(P) be the membership function of the fuzzy output set of power adjustment obtained based on the above rule reasoning. The centroid P* of defuzzification is expressed as: P*=∫PminPmaxμoutput(P)dP∫PminPmaxPμoutput(P)dP Pmin and Pmax are the possible ranges of power adjustment. Through this formula, a specific power adjustment command, namely P*, is obtained; Step 2: Introduce the least square method to estimate the temperature model of the heating system; 2.1) Collect key parameters of the heating process in real time in the heating non-combustion equipment, and perform data cleaning, standardization and feature selection preprocessing on the collected data; 2.2) The least square method is used to establish a relationship model between temperature and electrical parameters. The model attempts to find a set of parameters that makes the predicted heating temperature and the actual heating temperature Q = m·c·(t2-t1) square and the minimum basic model; P=ΔtQ=UI=I2R Q represents the required heat, m and c are the mass and specific heat capacity of the smoking matrix, t2 and t1 are the set temperature and current temperature, P is the heating power, U, I, and R are the voltage, current, and resistance, respectively; Considering that the heating process is affected by multiple factors, the model should include basic heat formulas and power calculations. The model is expanded to use the following model: Q=m·c(t)·(t2-t1)+AQ c(t) represents the function of the specific heat capacity of the smoke matrix changing with temperature, and ΔQ takes into account additional factors, including heat loss, heating efficiency, The least square method is used to establish a relationship model between temperature and electrical parameters. The goal of the model is to find a set of parameters that minimizes the sum of squares of the difference between the predicted heating temperature and the actual heating temperature. Where Q represents the required heat, m and c are the mass and specific heat capacity of the smoking substrate, t2 and t1 are the set temperature and current temperature, P is the heating power, U, I and R are the voltage, current and resistance, respectively; Considering the temperature dependence of the internal resistance of the heater and achieving more precise temperature control, a temperature feedback mechanism is introduced in this model, that is, the heating power is adjusted according to the difference between the actual measured temperature and the target temperature: P adjust (t)=P(t)+K p ·(t2-T actual (t))+K i ·∫(t2-T actual (t))dt+K d · dtd (t2-T actual (t)) Tactual(t) is the actual measured temperature, Kp, Ki, and Kd are the proportional, integral, and differential coefficients of the PID controller, respectively.

2. The method for precise temperature control and adjustment of a heating-not-burning appliance according to claim 1, characterized in that: The rule base in 1.2) specifically includes: In order to implement the control strategy, the system needs to identify two main stages: rapid temperature increase and temperature maintenance and step adjustment. In order for the system to achieve accurate temperature control in various scenarios, rules are formulated based on different input condition combinations in the rapid temperature increase stage and the temperature maintenance and step adjustment stage. 1.2.1) Phase 1: Rapid temperature increase, the goal is to quickly increase the system temperature to the target range within 20 seconds; Rule 1: The current stage is "rapid temperature increase", the temperature T is far below the target value, the time t < 10 seconds, the maximum voltage and current are output, and the temperature is increased at the highest speed; Rule 2: The current stage is "rapid temperature increase", the temperature T is slightly lower than the target value, the time t<20t seconds, the output voltage and current are medium, the temperature is increased rapidly, but overshoot is avoided; Rule 3: The current stage is "rapid temperature increase", the temperature T reaches the upper limit of the target value, the time t < 20t seconds, reduce the current and voltage to prevent the temperature from exceeding the target range; Rule 4: The current stage is "rapid temperature increase", the temperature T is close to the target value, the time t is between 15-20 seconds, and the current and voltage are gradually reduced to prepare for the temperature maintenance stage; 1.2.1) Stage 2: Temperature maintenance and step adjustment After the rapid pull-up is completed, stabilize the temperature within the target range and make fine adjustments as needed; Rule 5: The current stage is "temperature maintenance and step adjustment", the temperature T is stable within the target range, the current voltage and current are maintained, and the temperature is continuously stabilized; Rule 6: The current stage is "temperature maintenance and step adjustment". The temperature T starts to drop slightly, and the current is increased to increase the temperature slightly and maintain the target temperature. Rule 7: The current stage is "temperature maintenance and step adjustment". The temperature T is slightly higher than the target range. The voltage is reduced to reduce the temperature slightly. Rule 8: Condition: The current stage is "temperature maintenance and step adjustment". The temperature T is at the lower limit of the target range, but the current and voltage output are close to the minimum limit. Maintain the minimum current and voltage and observe the temperature changes to prevent over-adjustment. Rule 9: The current stage is "temperature maintenance and step adjustment". The external environment changes cause the temperature T to drop sharply, and the current and voltage are increased quickly to respond to the temperature change immediately and restore the target temperature. Rule 10: The current stage is "temperature maintenance and step adjustment". The system detects that there is no obvious temperature change for a long time, keeps the current settings unchanged, and makes periodic small adjustments to verify system stability; 1.2.3) Comprehensive control rules Rule 11: The current stage is "temperature maintenance and step adjustment". The temperature T remains within the target range, but the system detects that the power P is gradually increasing, and slightly reduces the current to optimize energy consumption and maintain temperature stability. Rule 12: The current stage is "temperature maintenance and step adjustment". The voltage V fluctuates significantly, causing the temperature T to be unstable. The system response speed is adjusted and the current is increased to buffer the impact of voltage fluctuations on temperature. Rule 13: The current stage is "rapid temperature increase". The time t exceeds 20 seconds, but the temperature T still does not reach the target range. It enters the temperature maintenance and step adjustment stage and outputs appropriate adjustment commands to continue heating. Rule 14: The current stage is "temperature maintenance and step adjustment". It is detected that the system temperature regulation is invalid and the target temperature cannot be maintained. It switches back to the fast pull-up stage and outputs the highest voltage and current to restore the target temperature.

3. An application of the method for precise temperature control and adjustment of a heating-without-combustion appliance according to claim 1, characterized in that: Application in the startup phase of heating-not-burning appliances: The temperature rise process is optimized by combining the fuzzy control model and the least squares model. The fuzzy control model handles the uncertainty and nonlinearity of the system and provides a regulation strategy based on empirical rules, so that the heating appliance responds quickly and approaches the set temperature in a short time. The least squares rule is used to accurately model the relationship between heat demand and electrical parameters during the heating process, and the error between the predicted temperature and the actual temperature is minimized by continuously adjusting the control parameters to accelerate the heating process. After the heating device reaches a temperature close to the set temperature, stable temperature control is achieved through the dynamic adjustment capability of the fuzzy control model and the precise parameter optimization of the least squares method. When faced with temperature fluctuations, the fuzzy control model balances the heating power and heat demand by adjusting the control strategy to maintain temperature stability; the least squares method continuously corrects the parameters in the model to reflect changes under actual operating conditions and ensure that the temperature remains near the set value.

4. The method for precise temperature control and adjustment of a heating-not-burning appliance according to claim 1, characterized in that: In the step 2.1), the key parameters include heating power P, voltage U, current I, resistance R, set temperature T2, current temperature T1, time constant τ, heat loss coefficient α, mass m and specific heat capacity c.

Citation Information

Cited By

  • Heater temperature control performance detection method and system based on vacuum environment

    CN119690056A

  • Self-adaptive adjustment control system and method for NOx generation mechanism combustion experiment

    CN120295138A

  • Biochemical analysis control method and system

    CN120595884A