Fuzzy self-adaptive PID (Proportion Integration Differentiation) control method for high-temperature environment of small closed space

Abstract

The invention discloses a fuzzy self-adaptive PID (Proportion Integration Differentiation) control method for a high-temperature environment of a small closed space. According to the method, a fuzzy logic controller and PID controller cooperative control architecture based on the variable universe theory is constructed, and PID parameters are adjusted on line by utilizing a fuzzy inference rule by collecting the temperature deviation and the change rate of the closed space in real time. The fuzzy rule base design adopts a variable universe theory, a membership function and a universe range are dynamically adjusted according to a system state, and the adjustment direction and amplitude of PID parameters are accurately mapped. A real-time monitoring mechanism based on feedback of a temperature sensor is introduced, high-frequency sampling is carried out on the temperature of the space, and timeliness and accuracy of control signals are ensured. According to the method, by setting a temperature threshold value and an overshoot limiting mechanism, the phenomena of temperature overshoot and oscillation are effectively avoided, and the stability and reliability of the system are improved. According to the invention, the temperature can be quickly and accurately controlled in a high-temperature environment of a complex small closed space, and the temperature fluctuation range is obviously reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of temperature control, and particularly relates to a fuzzy adaptive PID control method for a high-temperature environment in a small enclosed space. Background Technique

[0002] Although the traditional Proportional-Integral-Derivative (PID) control algorithm has long dominated the industrial temperature control field with advantages such as a simple structure and ease of engineering use, its core defects are significantly amplified in the high-temperature scenario of a small enclosed space. The parameters of the traditional PID controller rely on offline tuning strategies such as the Ziegler-Nichols method and cannot adapt to dynamic thermal disturbances in real time, resulting in a large response lag and overshoot. The coupled effect of multiple physical fields in the enclosed space causes uneven heat flow distribution, and the steady-state error exceeds ±2.5%. Moreover, the thermal drift (±0.3°C / 100°C temperature rise) and measurement noise (SNR < 40 dB) of the sensor in the high-temperature environment further cause phase lag and bandwidth reduction (a reduction of 62%). In addition, the non-linear dead zone characteristic of the actuator (greater than 5% input range) exacerbates the control deviation, especially in the stage of temperature mutation, which is prone to output saturation and continuous oscillation, highlighting the dynamic mismatch and lack of disturbance rejection ability of the traditional PID in complex high-temperature scenarios.

[0003] Although the fuzzy PID control improves the adaptability by dynamically adjusting the parameters, there are still defects in the small high-temperature enclosed scenario. First, the membership function and the range of the universe of discourse of the traditional fuzzy control are fixed and cannot dynamically adjust the sensitivity according to the error, resulting in a slow adjustment speed in the large deviation stage and easy oscillation in the small deviation stage. Second, the fuzzy rules rely on expert experience and lack the global search ability of intelligent algorithms (such as genetic algorithms and particle swarm optimization) for the initial values of the parameters, and are prone to falling into local optima. Third, most of the existing solutions focus on algorithm simulation and do not fully consider the real-time constraints of the embedded system (such as processor computing power and sensor sampling frequency), and it is difficult to achieve a microsecond-level control period during actual deployment. Summary of the Invention

[0004] The purpose of the present invention is to provide a fuzzy adaptive PID control method for a high-temperature environment in a small enclosed space aiming at problems such as low temperature control accuracy, slow dynamic response, and poor adaptability to environmental changes in a small enclosed space. Through the variable universe of discourse fuzzy adaptive PID algorithm and the embedded real-time control system, this method can effectively solve the temperature fluctuation problem caused by thermal disturbances, sensor drift, and actuator non-linearity in a high-temperature environment, and is particularly suitable for industrial scenarios with strict requirements for temperature response speed (≤1 ms control period) and stability (steady-state error < 0.5%).

[0005] The technical solution for achieving the object of the present invention is: a fuzzy adaptive PID control method for a high-temperature environment in a small enclosed space, and the method includes:

[0006] Step 1, build an adaptive PID control system for adjusting the ambient temperature based on the fuzzy PID control principle, and this system supports a closed-loop control cycle in milliseconds;

[0007] Step 2, determine the reference values of the fuzzy PID parameters;

[0008] Step 3, construct a variable universe fuzzy control layer to adjust the fuzzy PID parameters online; the variable universe fuzzy control layer adjusts the universe of discourse range of the input variable / output variable to adjust the fuzzy PID parameters online;

[0009] Step 4, optimize the PID parameters through the particle swarm optimization algorithm;

[0010] Step 5, achieve the fuzzy adaptive PID control of the ambient temperature through the optimized adaptive PID control system.

[0011] Further, in step 2, the reference values of the fuzzy PID parameters are determined by the Ziegler-Nichols method, which specifically includes:

[0012] Step 1-1, initialize the integral gain and the derivative gain as: and ; start from a preset minimum value for the proportional gain , gradually increase it, and at the same time observe the output response of the adaptive PID control system. When the adaptive PID control system starts to exhibit equal-amplitude oscillations, record the proportional gain at this time as , and the oscillation period as ;

[0013] Step 1-2, calculate the reference values of the PID controller parameters according to the Ziegler-Nichols method:

[0014]

[0015]

[0016]

[0017] Step 1-3, analyze the thermodynamic model of the controlled object, and the transfer function is represented by as:

[0018]

[0019] In the formula, is the gain of the adaptive PID control system, which is the ratio of the steady-state output to the input of the system; is the time constant, is the lag time, the pure time-delay effect; s represents the complex variable;

[0020] Steps 1 - 4, when is higher than its preset threshold, the proportional gain in Steps 1 - 2 is decreased by a preset step size until it is lower than its preset threshold; when is higher than its preset threshold, the integral gain in Steps 1 - 2 is increased by a preset step size until it is lower than its preset threshold; when is higher than its preset threshold, the derivative gain is increased by a preset step size until it is lower than its preset threshold.

[0021] Further, Step 3 specifically includes:

[0022] Step 3 - 1, according to the characteristics of the adaptive PID control system, select the input variables as the error and the error change rate , and the output variable as the voltage ; the error is the difference between the current ambient temperature and the desired temperature;

[0023] Step 3 - 2, set the physical ranges for the input and output variables: set the initial universe of discourse of the error as , the initial universe of discourse of the error change rate as , and the initial universe of discourse of the output variable as ;

[0024] Step 3 - 3, normalize the initial universe of discourse of the error and the initial universe of discourse of the error change rate so that the ranges of the two initial universes of discourse are the same;

[0025] Step 3 - 4, use the Gaussian membership function to divide the universe of discourse of the input / output variables into 7 fuzzy subsets from negative large to positive large: , and map the output variable to the actual control quantity through anti-normalization; where represent negative large, negative medium, negative small, zero, positive small, positive medium, and positive large respectively;

[0026] Step 3 - 5, construct the scaling factor of the input universe of discourse:

[0027]

[0028]

[0029] Among them, are all adjustment functions used to achieve the error When it is large, the universe of discourse is expanded, and when the error is small, the universe of discourse is contracted, and when the rate of change of the error is large, the universe of discourse is expanded; and are respectively the stretching factors corresponding to the universes of discourse of the error and the rate of change of the error ;

[0030] Based on the stretching factors of the input universe of discourse, the universe of discourse of the error is adjusted to , and the universe of discourse of the rate of change of the error is adjusted to ;

[0031] Step 3-6, construct the stretching factor of the output universe of discourse:

[0032]

[0033] Among them, is the attenuation coefficient, is the stretching factor corresponding to the universe of discourse of the output variable u;

[0034] Based on the stretching factor of the output universe of discourse, the universe of discourse of the output variable is adjusted to ;

[0035] Step 3-7, construct the fuzzy rule base, including the corresponding relationship between the fuzzy subsets of the input variables and the fuzzy subsets of the output variables; the fuzzy rule base is in the form of a list. In this list: the 7 fuzzy subsets of the two input variables are used as the 7 values of the rows and columns respectively, and the corresponding fuzzy subsets of the output variables are output at the intersection of each row and column;

[0036] Step 3-8, defuzzify by the centroid method to solve the integral gain , the derivative gain and the proportional gain respectively. The calculation formula is:

[0037]

[0038] Among them, is the exact value after defuzzification of the output variable u of the controller in the adaptive PID control system; is the i-th membership degree value of the output variable. This membership degree value is represented by a fuzzy subset. n + 1 represents that there are n + 1 membership degree values for the output variable, and one membership degree value corresponds to one fuzzy subset; The membership degree for the output variable belonging to .

[0039] Furthermore, step 3 also includes:

[0040] Step 3-9, in combination with the output universe scaling factor, to correct in real time the output range of the output variable u of the controller:

[0041]

[0042] where, represents the corrected u.

[0043] Furthermore, in step 3-7, construct a fuzzy rule base according to the following principle: when the error is large, i.e., , then enhance the role of the proportional gain; when the error is small, including , then enhance the role of the integral gain; when the rate of change of the error is large, i.e., , then enhance the differential gain role.

[0044] Furthermore, in step 3-8, perform defuzzification by the centroid method based on the membership degree values of the output variables within the m×m neighborhood range of the fuzzy rule base, specifically including:

[0045] Assume that there are multiple membership degree values of the output variables within the m×m neighborhood range;

[0046] For each membership degree value , solve through the following formula:

[0047] =

[0048] where q represents the total number of membership degree values within the m×m neighborhood range, represents the membership degree of the j-th membership degree value within the m×m neighborhood range.

[0049] Furthermore, the membership degree of the j-th membership degree value is calculated by the formula:

[0050] =

[0051] where, , are respectively the membership degrees of the membership degree values of the two input variables corresponding to the j-th membership degree value ; the specific numerical values of the membership degrees of the membership degree values of each input variable are set by oneself.

[0052] Further, in step 3-8, defuzzification is performed by the centroid method based on the 2×2 neighborhood range of the fuzzy rule base. Then each input variable includes two membership values. Set the membership of one membership value to , and the membership of the other membership value to 1 - , where .

[0053] Further, step 4 specifically includes:

[0054] Step 4-1, encode the parameters into the position vector of the particle , corresponding to respectively; randomly generate a particle swarm, and initialize the position and velocity of each particle;

[0055] Step 4-2, comprehensively consider the dynamic response indexes of the adaptive PID control system and design a fitness function;

[0056] Step 4-3, iterative optimization of the particle swarm: substitute the parameters of each particle into the simulation model, calculate the step response and extract the performance indexes; update the individual and group optima, and record the position corresponding to the historical best fitness value of each particle as ; record the position corresponding to the best fitness value in the whole particle swarm, i.e., the global optimal position, as ;

[0057] Adjust the particle movement direction according to the formula:

[0058]

[0059]

[0060] Among them, are the particle velocities corresponding to the (t + 1)-th iteration and the t-th iteration respectively, is the inertia weight, is the learning factor; is a random number; are the particle positions corresponding to the (t + 1)-th iteration and the t-th iteration respectively;

[0061] Step 4-4, continuously iterate according to step 4-3. When the fitness value has no obvious improvement in the continuous N iteration results, that is, the change rate is lower than the preset threshold, stop the iteration. N is customarily set;

[0062] Step 4-5, use the parameter combination corresponding to the global optimal position as the optimization result of the PID parameters.

[0063] Furthermore, in step 4-2, the fitness function is as follows:

[0064]

[0065] In the formula, is the integral of squared error, which is used to measure the steady-state error; is the settling time, which is used to evaluate the response speed; is the overshoot, which is used to reflect the stability; is the weight, which is adaptively adjusted according to requirements

[0066] Compared with the prior art, the significant advantages of the present invention are as follows:

[0067] (1) A deep fusion architecture of variable universe fuzzy control and adaptive PID is proposed. The dynamic universe adjustment mechanism introduces the scaling factors (α(e), α(ec)) of the error (e) and the rate of change of error (ec), and dynamically adjusts the input universe range of the fuzzy controller through the formulas α(e)=1-λe² and α(ec)=1-μec². When the error is large, the universe is expanded for fast response, and when the error is small, the universe is contracted for fine adjustment, breaking through the problems of "slow adjustment for large deviation and easy oscillation for small deviation" caused by the fixed universe of traditional fuzzy control.

[0068] The output universe is dynamically corrected through β(u)=1 / (1+γ|e|), and is designed in combination with the Lyapunov stability theory to ensure that the control quantity is within the stable range of the system and suppress output saturation and oscillation.

[0069] The design of the variable universe fuzzy rule base is constructed based on 7×7 = 49 fuzzy rules, and the adjustment direction of the PID parameters is mapped in real time according to the error and the rate of change. For example, when the error is large, the proportional action is enhanced to quickly eliminate the error, and when the error is small, the integral action is enhanced to improve the steady-state accuracy, effectively realizing the adaptive switching of the control strategy.

[0070] (2) Global parameter optimization through the particle swarm optimization (PSO) algorithm

[0071] The intelligent optimization framework of the PID parameters encodes the PID parameters ( ) as the particle position vector and performs global search through the PSO algorithm. An innovative fitness function is designed, which comprehensively considers the steady-state error, settling time and overshoot indexes, avoiding the defects of the traditional Ziegler-Nichols method that relies on off-line tuning and is prone to falling into local optimum.

[0072] During the particle swarm iteration process, the individual best (pBest) and the global best (gBest) are dynamically updated. By linearly decreasing the inertia weight ω (from 0.9 to 0.4) and the learning factors (c1, c2 = 2), the global exploration and local exploitation capabilities are balanced, improving the parameter convergence speed and control accuracy (steady-state error < 0.5%).

[0073] (3)Millisecond-level closed-loop implementation of the embedded real-time control architecture.

[0074] For the hardware and system-level integrated innovation, an ARM Cortex-M7 embedded controller (main frequency > 300MHz), a high-precision K-type thermocouple (accuracy ±0.5%), and a solid-state relay (response time < 5ms) are used to build a hardware architecture supporting a millisecond-level closed-loop control cycle, adapting to the space and power consumption limitations of miniaturized devices.

[0075] Integrate a real-time operating system (RTOS) and a PWM drive module to ensure that the control cycle ≤ 1ms, solving the problem of poor real-time performance caused by insufficient processor computing power in traditional solutions.

[0076] For the anti-disturbance and stability enhancement design, a temperature threshold and overshoot limit mechanism are introduced. Combining the dynamic proportional factor uactual = β(u)・Ku・u0 to correct the output control quantity in real time, effectively suppressing the temperature overshoot and continuous oscillation caused by the thermal drift of the sensor (±0.3℃ / 100℃ temperature rise), measurement noise (SNR<40dB), and the non-linear dead zone of the actuator (>5% input range) in a high-temperature environment.

[0077] (4)Breakthrough innovation at the technical effect level.

[0078] The control accuracy is significantly improved: the temperature fluctuation range is reduced from ±2.5% of the traditional method to ±0.5%, the overshoot is reduced by more than 40%, and the adjustment time is shortened by 30%, meeting the stringent requirements for steady-state error and response speed in high-temperature enclosed spaces (such as the cockpit of an amphibious armored vehicle, a food drying room).

[0079] Robustness and adaptability optimization: By combining the variable universe and the PSO algorithm, the system shows stronger anti-disturbance ability to the dynamic changes of multi-physical-field coupled thermal disturbances, lag time τ, and system gain Ke, effectively solving the dynamic mismatch problem of traditional PID in non-linear lag scenarios.

[0080] The present invention will be further described in detail below with reference to the accompanying drawings. Description of the Drawings

[0081] Figure 1 It is a flowchart of the fuzzy adaptive PID control method for the high-temperature environment in a small enclosed space of the present invention.

[0082] Figure 2 It is the schematic diagram of fuzzy PID control.

[0083] Figure 3 It is the flowchart of calculating fuzzy PID control parameters in an embodiment. Specific implementation manners

[0084] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0085] It should be noted that if there are directional indications (such as up, down, left, right, front, back,...) involved in the embodiments of the present invention, the directional indications are only used to explain the relative positional relationship and movement conditions between components in a specific posture (as shown in the drawings). If the specific posture changes, the directional indications will also change accordingly.

[0086] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only for descriptive purposes, and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In addition, the technical solutions between various embodiments can be combined with each other, but it must be based on the fact that those skilled in the art can implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the protection scope required by the present invention.

[0087] In view of the problems existing in the prior art, the present invention provides a variable universe fuzzy adaptive PID control algorithm. First, by introducing the scaling factors of the error (e) and the error change rate (ec), the universe range of the fuzzy controller is adjusted in real time to achieve fast response in the large deviation stage and fine adjustment in the small deviation stage. Secondly, combining the particle swarm optimization (PSO) algorithm with fuzzy inference, the PID parameters are globally optimized to avoid the subjectivity limitation of the traditional rule base and improve the convergence speed and control accuracy (steady-state error < 0.5%). Then, an embedded architecture is adopted to integrate a high-precision thermocouple, a PWM drive module and a real-time operating system (RTOS) to ensure that the control period ≤ 1 ms and adapt to the space and power consumption limitations of miniaturized high-temperature devices.

[0088] In one embodiment, in combination with Figure 1 , a fuzzy adaptive PID control method for a high-temperature environment in a small enclosed space is provided, and the method includes:

[0089] Step 1: Build an adaptive PID control system for adjusting the environmental temperature based on the fuzzy PID control principle. This system supports a closed-loop control period of milliseconds.

[0090] The adaptive PID control system is as Figure 2 shown. is the input of the system; is the error of the system; is the proportional coefficient of the PID controller; is the integral coefficient of the PID controller; is the derivative coefficient of the PID controller; is the output of the controller; is the output of the system;

[0091] Based on this, build the hardware architecture of the system. Preferably, select a high-precision temperature sensor, K-type thermocouple, with an accuracy of ±0.5% as the measuring element; select an embedded controller, ARM Cortex-M7, with a main frequency greater than 300 MHz as the PID controller; select a solid-state relay as the actuator, with a response time less than 5 ms. Make the whole system support a closed-loop control period of milliseconds.

[0092] Step 2: Determine the reference values of the fuzzy PID parameters;

[0093] Step 3: Build a variable universe fuzzy control layer to adjust the fuzzy PID parameters online; the variable universe fuzzy control layer adjusts the universe of discourse range of the input variable / output variable to adjust the fuzzy PID parameters online;

[0094] Step 4: Optimize the PID parameters through the particle swarm optimization algorithm;

[0095] Step 5: Implement fuzzy adaptive PID control of the environmental temperature through the optimized adaptive PID control system.

[0096] Furthermore, in one embodiment, the reference values of the fuzzy PID parameters are determined by the Ziegler-Nichols method in Step 2, specifically including:

[0097] Step 1-1: Initialize the integral gain and the derivative gain as: and ; Start from a preset minimum value for the proportional gain and gradually increase it. At the same time, observe the output response of the adaptive PID control system. When the adaptive PID control system starts to exhibit equal-amplitude oscillations, record the proportional gain at this time as , the oscillation period as ;

[0098] Step 1-2: Calculate the parameter reference values of the PID controller according to the Ziegler-Nichols method:

[0099]

[0100]

[0101]

[0102] Step 1-3: Analyze the thermodynamic model of the controlled object, and the transfer function is represented by as follows:

[0103]

[0104] In the formula, is the gain of the adaptive PID control system, which is the ratio of the system steady-state output to the input; is the time constant, is the lag time, the pure delay effect; s represents the complex variable;

[0105] Step 1-4: When the value of is higher than its preset threshold, decrease the value of the proportional gain in Step 1-2 by a preset step size until it is lower than its preset threshold to prevent overshoot; when the value of is higher than its preset threshold, increase the value of the integral gain in Step 1-2 by a preset step size until it is lower than its preset threshold to accelerate the steady-state convergence; when the value of is higher than its preset threshold, increase the value of the derivative gain in Step 1-2 by a preset step size until it is lower than its preset threshold to compensate for the delay.

[0106] Furthermore, in one of the embodiments, Step 3 specifically includes:

[0107] Step 3-1: According to the characteristics of the adaptive PID control system, select the input variables as the error and the error change rate , and the output variable as the voltage ; the error is the difference between the current ambient temperature and the desired temperature;

[0108] Step 3-2: Set the physical ranges for the input variables and the output variable: Set the initial universe of discourse of the error as , the initial universe of discourse of the error change rate as , and the initial universe of discourse of the output variable as ;

[0109] Exemplarily here, the error and the error change rate values are mapped to the fuzzy universe of discourse through the quantization factor ; ;

[0110] Step 3-3, normalize the initial universe of discourse of the error and the initial universe of discourse of the error change rate so that the ranges of the two initial universes of discourse are the same;

[0111] Step 3-4, use a Gaussian membership function to divide the universe of discourse of the input / output variables into 7 fuzzy subsets from negative large to positive large: , and map the output variable to the actual control quantity through anti-normalization; where respectively represent negative large, negative medium, negative small, zero, positive small, positive medium, and positive large;

[0112] Step 3-5, construct the input universe of discourse scaling factor:

[0113]

[0114]

[0115] Among them, are all adjustment functions, used to expand the universe of discourse when the error is large, contract the universe of discourse when the error is small, and expand the universe of discourse when the error change rate is large; , are respectively the scaling factors corresponding to the universes of discourse of the error and the error change rate ;

[0116] Based on the input universe of discourse scaling factor, adjust the universe of discourse of the error to , and adjust the universe of discourse of the error change rate to ;

[0117] Step 3-6, construct the output universe of discourse scaling factor:

[0118]

[0119] Among them, is the attenuation coefficient, is the scaling factor corresponding to the universe of discourse of the output variable u; designed in combination with the Lyapunov stability theory to ensure that the output is within the stable range of the system;

[0120] Based on the output universe scaling factor, adjust the universe of the output variable to ;

[0121] Here, in the large error stage, increases, expands the universe, enhances the proportional action, and accelerates the response; in the small error stage, decreases, contracts the universe, enhances the integral action, and improves the steady-state accuracy. A dynamic proportional factor is introduced into the output universe, designed in combination with the Lyapunov stability theory, to suppress overshoot and oscillation and enhance stability.

[0122] Step 3-7: Construct a fuzzy rule base, including the correspondence between the fuzzy subsets of the input variables and the fuzzy subsets of the output variables; the fuzzy rule base is in the form of a list. In this list: the 7 fuzzy subsets of the two input variables are used as the 7 values of the rows and columns respectively, and the intersection of each row and column outputs the corresponding fuzzy subset of the output variable;

[0123] Step 3-8: Combine Figure 3 , and defuzzify by the centroid method to solve the integral gain , the derivative gain and the proportional gain respectively. The calculation formulas are as follows:

[0124]

[0125] where, is the exact value after defuzzifying the output variable u of the controller in the adaptive PID control system; is the i-th membership value of the output variable, which is represented by a fuzzy subset. n + 1 represents that there are n + 1 membership values for the output variable, and one membership value corresponds to one fuzzy subset; is the membership degree of the output variable belonging to .

[0126] Here, preferably, in some embodiments, in Step 3-7, the fuzzy rule base is constructed according to the following principle: when the error is large, i.e., , then enhance the proportional gain effect; when the error is small, including , then enhance the integral gain effect; when the error change rate is large, i.e., , then enhance the derivative gain effect.

[0127] Exemplarily preferably, the fuzzy rule base established according to the above principle is shown in Table 1 below:

[0128] Table 1 Fuzzy Rule Base

[0129]

[0130] Preferably, in some embodiments, in step 3-8, defuzzification is performed by the centroid method based on the membership degree values of the output variables within the m×m neighborhood range of the fuzzy rule base, specifically including:

[0131] Suppose there are membership degree values of multiple output variables within the m×m neighborhood range;

[0132] For each membership degree value , solve through the following formula :

[0133] =

[0134] where q represents the total number of membership degree values within the m×m neighborhood range , represents the membership degree of the j-th membership degree value within the m×m neighborhood range .

[0135] Here preferably, the membership degree of the j-th membership degree value is calculated by the formula:

[0136] =

[0137] In the formula, , are respectively the membership degrees of the membership degree values of the two input variables corresponding to the j-th membership degree value ; the specific numerical values of the membership degrees of the membership degree values of each input variable are set customarily.

[0138] Further preferably, in step 3-8, defuzzification is performed by the centroid method based on the 2×2 neighborhood range of the fuzzy rule base, then each input variable includes two membership degree values, and it is set that the membership degree of one membership degree value is , and the membership degree of the other membership degree value is 1 - , where . Here, the value of a is set customarily according to requirements.

[0139] Exemplarily preferably, based on Table 1, assume the two membership degree values of are , the membership degree belonging to is , then the membership degree belonging to is . Then assume the two membership degree values of are The membership degree of is then it belongs to and the membership degree is .

[0140] Then it belongs to and the membership degree is , it belongs to and the membership degree is then the output value belongs to and the membership degree is . Similarly, the output value belongs to and the other two membership degrees are , , and the membership degree of the output value belonging to NS is . It can be seen that the sum of the membership degrees of the output value is , and the membership degree of the output value is .

[0141] Therefore, , . Then:

[0142] .

[0143] Furthermore, in one of the embodiments, step 3 further includes:

[0144] Step 3-9, combining the output domain scaling factor to real-time correct the output range of the output variable u of the controller:

[0145]

[0146] In the formula, represents the corrected u.

[0147] Furthermore, in one of the embodiments, step 4 specifically includes:

[0148] Step 4-1, encoding the parameter into the position vector of the particle, corresponding to respectively; randomly generating a particle swarm, and initializing the position and velocity of each particle;

[0149] Step 4-2, designing a fitness function by integrating the dynamic response indexes of the adaptive PID control system;

[0150] Step 4-3, particle swarm iterative optimization: for each particle Substitute the parameters into the simulation model, calculate the step response and extract the performance indicators; update the individual and group optima, and record the position corresponding to the historical best fitness value of each particle as ; record the position corresponding to the best fitness value in the entire particle swarm, i.e., the global optimal position, as ;

[0151] According to formula, adjust the particle movement direction:

[0152]

[0153]

[0154] where are the particle velocities corresponding to the (t + 1)-th iteration and the t-th iteration respectively, is the inertia weight, usually linearly decreasing from 0.9 to 0.4, is the learning factor, set to 2; is a random number; are the particle positions corresponding to the (t + 1)-th iteration and the t-th iteration respectively;

[0155] Step 4-4, continuously iterate according to Step 4-3. When the fitness value has no obvious improvement in the results of N consecutive iterations, that is, the change rate is lower than the preset threshold, stop the iteration, and N is customarily set;

[0156] Step 4-5, take the parameter combination corresponding to the global optimal position as the optimization result of the PID parameters.

[0157] Preferably, in some embodiments, the fitness function in Step 4-2 is:

[0158]

[0159] In the formula, is the integral of the squared error, used to measure the steady-state error; is the settling time, used to evaluate the response speed; is the overshoot, used to reflect the stability; is the weight, adaptively adjusted according to requirements.

[0160] Here, breaking through the limitation of traditional fuzzy rules relying on expert experience, encode the PID parameters ( ) as the particle positions, and comprehensively evaluate the system performance (steady-state error , settling time , overshoot ) through the fitness function. Dynamically update the particle swarm positions to achieve the global optimality of the parameters and improve the convergence speed and control accuracy.

[0161] In summary, the present invention constructs a cooperative control architecture based on a variable universe fuzzy logic controller and a PID controller, and real-time collects the temperature deviation (e) and its change rate (ec) in a closed space; dynamically adjusts the universe range of input / output variables according to the variable universe theory, and real-time corrects the membership function and universe mapping relationship of the fuzzy controller through the scaling factor formula, so as to achieve fast response in the large error stage and fine adjustment in the small error stage; combines the particle swarm optimization (PSO) algorithm to globally optimize the PID parameters, designs a fitness function to comprehensively consider the steady-state error, adjustment time and overshoot index, and outputs the control quantity through a dynamic scale factor; realizes a millisecond-level closed-loop control period based on an embedded system, and suppresses temperature overshoot and oscillation through a temperature threshold and overshoot limit mechanism.

[0162] The industrial application of the present invention mainly focuses on small high-temperature closed spaces that require high-precision temperature control, such as the cockpit of an amphibious armored vehicle, a food / drug drying room, etc. This method effectively solves the temperature fluctuation problem caused by thermal disturbance, sensor drift and actuator nonlinearity in a high-temperature environment through a variable universe fuzzy adaptive PID algorithm and an embedded real-time control system, and is particularly suitable for industrial scenarios with strict requirements for temperature response speed (≤1ms control period) and stability (steady-state error <0.5%).

[0163] In one embodiment, a fuzzy adaptive PID control system for a high-temperature environment in a small closed space is provided. The system includes:

[0164] A first module for building an adaptive PID control system for adjusting the ambient temperature based on the fuzzy PID control principle, and the system supports a millisecond-level closed-loop control period;

[0165] A second module for determining the reference values of the fuzzy PID parameters;

[0166] A third module for building a variable universe fuzzy control layer to online adjust the fuzzy PID parameters; the variable universe fuzzy control layer online adjusts the fuzzy PID parameters by adjusting the universe range of the input variables / output variables;

[0167] A fourth module for optimizing the PID parameters through the particle swarm optimization algorithm;

[0168] A fifth module for realizing the fuzzy adaptive PID control of the ambient temperature through the optimized adaptive PID control system.

[0169] For the specific limitations of the fuzzy adaptive PID control system for the high-temperature environment in a small enclosed space, reference can be made to the limitations of the fuzzy adaptive PID control method for the high-temperature environment in a small enclosed space in the above text, which will not be elaborated here. Each module in the above-mentioned fuzzy adaptive PID control system for the high-temperature environment in a small enclosed space can be implemented in whole or in part through software, hardware, and their combinations. The above-mentioned modules can be embedded in the processor of the computer device in hardware form or be independent of it, or can be stored in the memory of the computer device in software form, so as to facilitate the processor to call and execute the operations corresponding to each of the above modules.

[0170] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it realizes:

[0171] Step 1, build an adaptive PID control system for adjusting the environmental temperature based on the fuzzy PID control principle, and this system supports a closed-loop control cycle at the millisecond level;

[0172] Step 2, determine the reference values of the fuzzy PID parameters;

[0173] Step 3, construct a variable universe fuzzy control layer to adjust the fuzzy PID parameters online; the variable universe fuzzy control layer adjusts the fuzzy PID parameters online by adjusting the universe of discourse range of the input variable / output variable;

[0174] Step 4, optimize the PID parameters through the particle swarm optimization algorithm;

[0175] Step 5, realize the fuzzy adaptive PID control of the environmental temperature through the optimized adaptive PID control system.

[0176] For the specific limitations of each step, reference can be made to the limitations of the fuzzy adaptive PID control method for the high-temperature environment in a small enclosed space in the above text, which will not be elaborated here.

[0177] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by the processor, it realizes:

[0178] Step 1, build an adaptive PID control system for adjusting the environmental temperature based on the fuzzy PID control principle, and this system supports a closed-loop control cycle at the millisecond level;

[0179] Step 2, determine the reference values of the fuzzy PID parameters;

[0180] Step 3, construct a variable universe fuzzy control layer to adjust the fuzzy PID parameters online; the variable universe fuzzy control layer adjusts the fuzzy PID parameters online by adjusting the universe of discourse range of the input variable / output variable;

[0181] Step 4, optimize the PID parameters through the particle swarm optimization algorithm;

[0182] Step 5, realize the fuzzy adaptive PID control of the environmental temperature through the optimized adaptive PID control system.

[0183] For the specific limitations of each step, reference can be made to the limitations of the fuzzy adaptive PID control method for the high-temperature environment in a small enclosed space in the above text, which will not be elaborated here.

[0184] In summary, the present invention innovatively integrates the variable universe fuzzy control and the adaptive PID algorithm deeply. By dynamically adjusting the input / output universe range and membership function of the fuzzy controller, it realizes the rapid response in the large deviation stage and the fine adjustment in the small deviation stage; combines the particle swarm optimization algorithm to globally optimize the PID parameters, breaks through the limitation of traditional fuzzy rules relying on expert experience, and improves the parameter convergence speed and control accuracy; at the same time, integrates the embedded high-precision sensing, RTOS real-time operating system and PWM drive module to build a millisecond-level closed-loop control architecture, and designs the output scale factor based on the Lyapunov stability theory to effectively suppress overshoot and oscillation.

[0185] Experiments show that this method reduces the temperature fluctuation range to ±0.5% in the high-temperature enclosed environment, reduces the overshoot by more than 40%, shortens the adjustment time by 30%, and significantly improves the robustness of the system to nonlinear hysteresis and thermal disturbances.

[0186] The above shows and describes 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 by the above embodiments. The above embodiments and the descriptions in the specification only illustrate the principles of the present invention. Without departing from the spirit and scope of the present invention, any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A fuzzy adaptive PID control method for high-temperature environment in a small enclosed space, characterized in that, The method includes: Step 1, build an adaptive PID control system for adjusting the ambient temperature based on the fuzzy PID control principle, and the system supports a closed-loop control period of milliseconds; Step 2, determine the fuzzy PID parameter reference value; Step 3, build a variable universe fuzzy control layer to adjust the fuzzy PID parameters online; the variable universe fuzzy control layer adjusts the universe of discourse range of the input variable / output variable to adjust the fuzzy PID parameters online; Step 4, optimize the PID parameters through the particle swarm optimization algorithm; Step 5, realize the fuzzy adaptive PID control of the ambient temperature through the optimized adaptive PID control system.

2. The fuzzy adaptive PID control method for a high-temperature environment in a small enclosed space according to claim 1, characterized in that, In step 2, the fuzzy PID parameter reference value is determined by the Ziegler-Nichols method, specifically including: Step 1-1, initialize the integral gain and the derivative gain as: and ; Start from a preset minimum value and gradually increase the proportional gain , while observing the output response of the adaptive PID control system. When the adaptive PID control system starts to exhibit equal-amplitude oscillations, record the proportional gain at this time as , and the oscillation period as ; Step 1-2, according to the Ziegler-Nichols method, calculate the parameter reference value of the PID controller: ； ； ； Step 1-3, analyze the thermodynamic model of the controlled object, and the transfer function is represented by as follows: ； In the formula, is the gain of the adaptive PID control system, which is the ratio of the steady-state output to the input of the system; is the time constant, is the lag time, the pure time-delay effect; s represents the complex variable; Steps 1 - 4, when is higher than its preset threshold, decrease the proportional gain in Steps 1 - 2 by a preset step size until it is lower than its preset threshold; when is higher than its preset threshold, increase the integral gain in Steps 1 - 2 by a preset step size until it is lower than its preset threshold; when is higher than its preset threshold, increase the derivative gain by a preset step size until it is lower than its preset threshold.

3. The fuzzy adaptive PID control method for a small enclosed space in a high-temperature environment according to claim 1, wherein Step 3 specifically includes: Step 3-1, according to the characteristics of the adaptive PID control system, select the input variables as the error and the error change rate , and the output variable is the voltage ; the error is the difference between the current ambient temperature and the desired temperature; Step 3-2, set the physical ranges for the input variables and output variables: Set the initial universe of discourse of the error as , set the initial universe of discourse of the error change rate as , and set the initial universe of discourse of the output variable as ; Step 3-3, for the error of the initial universe of discourse and the error change rate of the initial universe of discourse, perform normalization processing so that the ranges of the two initial universes of discourse are the same; Step 3-4, the Gaussian membership function is used to divide the domain of the input / output variables into 7 fuzzy subsets from negative large to positive large: , and the output variable is mapped to the actual control quantity through anti-normalization; among them respectively represent negative large, negative medium, negative small, zero, positive small, positive medium, and positive large; Step 3-5, build the input universe scaling factor: ； ； Among them, are all adjustment functions used to achieve error to expand the universe of discourse when the error is large, and the error to contract the universe of discourse when it is small, and the rate of change of the error to expand the universe of discourse when it is large; and are respectively the scaling factors corresponding to the universes of discourse of the error and the rate of change of the error ; Based on the input universe scaling factor, adjust the universe of the error to , and adjust the universe of the error change rate to ; Step 3-6, build the output universe scaling factor: ； Among them, is the attenuation coefficient, is the scaling factor corresponding to the universe of discourse of the output variable u; Based on the output universe scaling factor, adjust the universe of the output variable to ; Step 3-7, build a fuzzy rule base, including the corresponding relationship between the fuzzy subsets of the input variables and the fuzzy subsets of the output variables; the fuzzy rule base is in the form of a list. In this list: the 7 fuzzy subsets of the two input variables are used as the 7 values of the rows and columns respectively, and the corresponding fuzzy subsets of the output variables are output at the intersection of each row and each column; Step 3-8, perform defuzzification by the centroid method to solve the integral gain , differential gain and proportional gain respectively, and the calculation formula is as follows: ； Among them, is the exact value after defuzzification of the output variable u of the controller in the adaptive PID control system; is the i-th membership value of the output variable, which is represented by a fuzzy subset. n + 1 indicates that there are n + 1 membership values for the output variable, and one membership value corresponds to one fuzzy subset; is that the output variable belongs to membership degree.

4. The fuzzy adaptive PID control method for a high-temperature environment in a small enclosed space according to claim 3, characterized in that, Step 3 also includes: Step 3-9, combine the output universe scaling factor to correct the output range of the output variable u of the controller in real time: ； In the formula, represents the corrected u.

5. The fuzzy adaptive PID control method for a small enclosed space in a high-temperature environment according to claim 3, characterized in that In Step 3-7, the fuzzy rule base is constructed according to the following principle: when the error is large, i.e., , the proportional gain effect is enhanced; when the error is small, including , the integral gain effect is enhanced; when the rate of change of the error is large, i.e., , the derivative gain effect is enhanced.

6. The fuzzy adaptive PID control method for a small enclosed space in a high-temperature environment according to claim 3, characterized in that, In step 3-8, defuzzification is performed by the centroid method based on the membership degree values of the output variables within the m×m neighborhood of the fuzzy rule base, specifically including: Suppose there are multiple membership degree values of the output variables within the m×m neighborhood; For each membership value , solve using the following formula : = ； Among them, q represents the total number of membership values within the m×m neighborhood range, and represents the membership of the j-th membership value within the m×m neighborhood range.

7. The fuzzy adaptive PID control method for a small enclosed space in a high-temperature environment according to claim 6, characterized in that, The j-th membership value of the membership degree is calculated by the formula: = ； wherein, and are respectively the membership degrees of the membership degrees of the two input variables corresponding to the j-th membership degree value ; the specific numerical values of the membership degrees of the membership degrees of each input variable are set by user-defined.

8. The fuzzy adaptive PID control method for a small enclosed space in a high-temperature environment according to claim 7, characterized in that In step 3-8, defuzzification is performed using the centroid method based on the 2×2 neighborhood range of the fuzzy rule base. Each input variable includes two membership values. Set the membership of one of the membership values to , and the membership of the other membership value to 1-a, where .

9. The fuzzy adaptive PID control method for a high-temperature environment in a small enclosed space according to claim 1, wherein Step 4 specifically includes: Step 4-1: Encode the parameters as the position vectors of particles , corresponding respectively to ; randomly generate a particle swarm and initialize the positions and velocities of each particle; Step 4-2, comprehensively consider the dynamic response index of the adaptive PID control system and design a fitness function; Step 4-3, Particle swarm iterative optimization: Substitute the parameters of each particle into the simulation model, calculate the step response and extract the performance indicators; Update the individual and group optima, and record the position corresponding to the historical best fitness value of each particle as ; Record the position corresponding to the best fitness value in the entire particle swarm, i.e., the global optimal position, as ; Adjust the particle motion direction according to the formula: ； ； Among them, are the particle velocities corresponding to the (t + 1)-th iteration and the t-th iteration respectively, is the inertia weight, is the learning factor; is a random number; are the particle positions corresponding to the (t + 1)-th iteration and the t-th iteration respectively; Step 4-4, continuously iterate according to step 4-3, and stop the iteration when the fitness value has no obvious improvement in the continuous N iteration results, that is, the change rate is lower than the preset threshold, and N is set by the user; Step 4-5, take the parameter combination corresponding to the globally optimal position as the optimization result of the PID parameters.

10. The fuzzy adaptive PID control method for a high-temperature environment in a small enclosed space according to claim 9, characterized in that The fitness function in step 4-2 is: ； In the formula, is the integral of squared error, which is used to measure the steady-state error; is the settling time, which is used to evaluate the response speed; is the overshoot, which is used to reflect the stability; is the weight, which is adaptively adjusted according to requirements.

Images