Intelligent control method and system for green firing process of porcelain insulator

Abstract

This invention relates to the field of process control technology, specifically to an intelligent control method and system for the green firing process of porcelain insulators. The method includes the following steps: collecting the operating temperatures of the front temperature zone, the current temperature zone, and the rear temperature zone of the kiln; generating the temperature difference between adjacent temperature zones and comparing its direction with historical temperature differences to obtain the heat penetration status identifier of each temperature zone; extracting the temperature and time before and after the inflection point of the firing curve; classifying and generating segmented curve execution commands; generating equivalent input heat based on gas flow rate and low-grade calorific value; calculating the segmented quota of gas calorific value and adjusting the gas valve opening. In this invention, by continuously identifying changes in heat penetration of adjacent temperature zones, the slope of the inflection point of the linkage curve, the loading thickness of the green body, the continuous state of temperature deviation, and the equivalent heat of the gas, the firing curve switching is matched with the gas supply to meet the heating demand of the green body, reducing over-firing or under-firing at the inflection point stage, improving the stability of thermal balance, and reducing ineffective gas consumption.

Description

Technical Field

[0001] This invention relates to the field of process control technology, and in particular to an intelligent control method and system for the green firing process of porcelain insulators. Background Technology

[0002] The field of process control technology involves the collection, judgment, adjustment, and recording of process variables such as temperature, pressure, flow rate, liquid level, composition, position, and time cycle in continuous or intermittent production processes. Its core aspects include the identification of the operating conditions of the controlled object, sensor data acquisition, process parameter setting, actuator action control, process deviation correction, process linkage, and operation data traceability. It typically establishes control relationships around production equipment, material status, energy medium, and operating procedures, and manages key nodes in the production process in segments through participating objects such as temperature sensors, pressure transmitters, flow meters, oxygen content detectors, position detectors, programmable logic controllers, regulating valves, fans, burners, and conveying mechanisms.

[0003] Among them, the traditional intelligent control method for the green firing process of porcelain insulators refers to the process control method for porcelain insulator blanks in the kiln to complete the firing stages such as preheating, dehumidification, heating, sintering, heat preservation, and cooling. The technical issues it addresses mainly include the moisture content of the blank, the kiln car's running position, the furnace temperature of each temperature zone, the gas supply, the amount of combustion air, the oxygen content in the kiln, the kiln pressure, the flue gas temperature, the cooling air volume, and the corresponding relationship of the firing curve. The traditional method usually pre-determines segmented temperature curves according to the porcelain insulator specifications, blank thickness, glaze layer state, and historical firing records, and collects data by thermocouples. The temperature of each zone in the kiln is recorded by the gas flow meter and air flow meter to measure the supply of the combustion medium, the atmosphere inside the kiln is detected by the oxygen probe, and the kiln pressure is detected by the pressure detector. The temperature is then adjusted by the programmable controller according to the set heating rate, holding time, damper opening, gas valve opening, exhaust valve opening, and kiln car advance rhythm. At the same time, the temperature zone settings, gas valve opening, combustion fan frequency, exhaust fan frequency, and cooling fan start-stop sequence are corrected by combining the flame status, body deformation, glaze condition, flue gas color, and kiln exit inspection results recorded by manual inspection.

[0004] Traditional control methods rely on preset segmented temperature curves and manual inspection and correction. The adjustment of each temperature zone mainly revolves around single-point temperature deviation and historical experience. It is difficult to reflect the heat transfer status of adjacent temperature zones in a timely manner. The heating rhythm at the inflection point of the curve is not well matched with the loading thickness of the blank. The gas supply is mostly adjusted based on experience with valve opening. The correspondence between heat input and dynamic heat deficit is unclear, which can easily lead to local heating lag, coarse heat preservation compensation, increased energy consumption, and affect the batch firing stability and glaze consistency. Summary of the Invention

[0005] To address the technical problems existing in the prior art, embodiments of the present invention provide an intelligent control method for the green firing process of porcelain insulators, comprising the following steps: S1: Collect the operating temperature of the front temperature zone of the kiln, the current operating temperature of the kiln, and the operating temperature of the rear temperature zone of the kiln. Calculate the temperature difference between the first and second adjacent temperature zones. Combine the first and second historical adjacent operating temperature differences from the previous sampling period with the direction comparison to generate the first temperature difference change direction feature, the second temperature difference change direction feature, and the temperature zone heat penetration status identifier. S2: Based on the heat penetration status identifier of the temperature zone, extract the starting temperature before the curve inflection point, the ending temperature before the curve inflection point, the duration of the curve inflection point, the starting temperature after the curve inflection point, the target temperature after the curve inflection point, and the set time of the curve inflection point, and calculate the temperature rise slope characteristics before the inflection point, the temperature rise slope characteristics after the inflection point, and the temperature rise slope difference characteristics. S3: Based on the temperature rise slope difference feature, extract the loading thickness of the porcelain insulator blank from the production configuration database, calculate the slope transition judgment threshold parameter, and use the temperature zone heat penetration status identifier, temperature rise slope difference feature and slope transition judgment threshold parameter to classify and generate segmented curve execution instruction features. S4: Configure the green energy-saving setting temperature of each temperature zone according to the segmented curve execution instruction characteristics, and calculate the temperature deviation data and dynamic heat deficit indicator of each temperature zone by combining the actual monitored temperature and the continuous running time of the temperature deviation of each temperature zone. S5: Based on the dynamic heat deficit indicator of the temperature zone, collect the instantaneous operating flow rate of the gas and the low heating value of the gas, calculate the equivalent input heat of the gas, construct the segmented quota parameters of the gas calorific value, adjust the opening parameters of the gas valve of the porcelain insulator kiln, and obtain the green control parameters of the firing process.

[0006] As a further aspect of the present invention, the specific steps of S1 are as follows: S101: Collect the operating temperature of the front temperature zone, the current temperature zone, and the rear temperature zone of the kiln. Perform a subtraction operation between the operating temperature of the front temperature zone and the current temperature zone to obtain the temperature difference between the first adjacent temperature zones. Perform a subtraction operation between the operating temperature of the current temperature zone and the operating temperature of the rear temperature zone to obtain the temperature difference between the second adjacent temperature zones. Combine the values ​​of the temperature difference between the first and second adjacent temperature zones to generate the current sampled temperature difference vector. S102: Call the current sampled temperature difference vector to obtain the first historical adjacent operating temperature difference and the second historical adjacent operating temperature difference in the previous sampling period in the local device. Perform bitwise subtraction on the current sampled temperature difference vector, the first historical adjacent operating temperature difference, and the second historical adjacent operating temperature difference to calculate the time difference variable. Obtain the direction determination benchmark. Compare the time difference variable with the direction determination benchmark to extract the direction sign. Establish a mapping matrix based on the direction sign to generate a set of temperature difference direction change features. S103: Based on the set of temperature difference direction change features, monitor the temperature zone heating interruption Boolean value and the preset cycle convergence threshold, substitute the set of temperature difference direction change features and the temperature zone heating interruption Boolean value into the Boolean operation unit to obtain the merged state bit sequence, perform classification matching retrieval on the merged state bit sequence and the preset cycle convergence threshold to obtain the corresponding condition attribute dictionary item, and generate the temperature zone heat penetration status identifier.

[0007] As a further aspect of the present invention, the process of obtaining the direction determination benchmark specifically involves: collecting the instrument temperature measurement noise amplitude and the base error parameter; performing an addition operation on the instrument temperature measurement noise amplitude and the base error parameter to obtain a non-error stable boundary; configuring the non-error stable boundary as an upward determination threshold; performing a numerical inversion operation on the non-error stable boundary to configure it as a downward determination threshold; and aggregating the upward determination threshold and the downward determination threshold to obtain the direction determination benchmark. The process of extracting the direction symbol by numerically comparing the time difference variable with the direction determination benchmark specifically involves comparing the time difference variable with the upward determination threshold and the downward determination threshold. When the time difference variable is greater than the upward determination threshold, a positive increment character is output; when the time difference variable is less than the downward determination threshold, a negative decrement character is output; and when the time difference variable is within the numerical range of the downward determination threshold and the upward determination threshold, a zero-value stable character is output. The positive increment character, the negative decrement character, and the zero-value stable character are then combined to form the direction symbol.

[0008] As a further aspect of the present invention, the specific steps of S2 are as follows: S201: Obtain the firing curve configuration set in the local storage device, extract the curve inflection point start temperature, curve inflection point end temperature and curve inflection point duration from the firing curve configuration set according to the temperature zone heat penetration status identifier, perform a subtraction operation between the curve inflection point end temperature and the curve inflection point start temperature to obtain the inflection point temperature variable, perform a ratio calculation between the inflection point temperature variable and the curve inflection point duration to generate the inflection point inflection point temperature rise slope feature; S202: Call the temperature zone heat penetration status identifier, retrieve the curve inflection point start temperature, curve inflection point target temperature and curve inflection point set time from the firing curve configuration set, perform a subtraction operation between the curve inflection point target temperature and the curve inflection point start temperature to extract the inflection point temperature increment, perform a ratio operation between the inflection point temperature increment and the curve inflection point set time to generate the inflection point temperature rise slope feature; S203: Call the temperature rise slope feature before the inflection point and the temperature rise slope feature after the inflection point, perform a subtraction operation on the temperature rise slope feature before the inflection point and the temperature rise slope feature after the inflection point to obtain feature difference data, obtain slope benchmark verification parameters, perform numerical comparison between the feature difference data and the slope benchmark verification parameters to extract the absolute value of the difference data, and generate temperature rise slope difference feature.

[0009] As a further aspect of the present invention, the specific steps of S3 are as follows: S301: Based on the temperature rise slope difference feature, retrieve the local production configuration database, extract the loading thickness of the porcelain insulator blank, obtain the preset foundation ratio coefficient, perform a multiplication operation between the loading thickness of the porcelain insulator blank and the preset foundation ratio coefficient to obtain the thickness-related offset, obtain the standard transition slope constant, multiply the thickness-related offset by the unit conversion equivalent coefficient and perform an addition operation with the standard transition slope constant to generate the slope transition judgment threshold parameter; S302: The temperature zone heat penetration status identifier, the temperature rise slope difference feature, and the slope transition judgment threshold parameter are concatenated dimensionally to establish a multi-dimensional state feature vector. A preset hyperplane weight matrix and a classification offset constant are obtained. The multi-dimensional state feature vector and the preset hyperplane weight matrix are used to calculate the inner product to obtain the feature projection value. The feature projection value and the classification offset constant are added to obtain the classification decision distance variable. The classification decision distance variable is compared with the zero boundary judgment benchmark to extract the category mapping label and generate a piecewise curve execution instruction feature.

[0010] As a further aspect of the present invention, the specific steps of S4 are as follows: S401: Call the segmented curve execution instruction feature, obtain the local device's built-in node register address set, perform parsing and extract the expected value of the basic temperature and the compensation coefficient for the segmented curve execution instruction feature, perform a product operation on the expected value of the basic temperature and the compensation coefficient to obtain the corrected temperature reference value, perform mapping and writing of the corrected temperature reference value and the node register address set to generate the green energy-saving setting temperature for each temperature zone. S402: Collect the actual monitored temperature of each temperature zone, perform a dimensionless subtraction operation between the actual monitored temperature of each temperature zone and the green energy-saving set temperature of each temperature zone to extract the temperature difference vector, obtain the probe thermal compensation vector, perform a summation operation between the temperature difference vector and the probe thermal compensation vector to obtain the calibration difference matrix, and generate temperature zone temperature deviation data. S403: The monitoring timer records the corresponding temperature deviation duration. The time dimension integral multiplication operation is performed on the temperature deviation data and the duration of the temperature deviation. The accumulated heat deficit value is obtained by combining the equivalent heat capacity coefficient. The calibration damping coefficient is obtained. The quotient value of the accumulated heat deficit value and the calibration damping coefficient is calculated to obtain the underheating state variable. A corresponding relationship is established to generate a dynamic heat deficit identifier for the temperature zone.

[0011] As a further aspect of the present invention, the process of parsing and extracting the expected base temperature value and compensation coefficient based on the segmented curve execution instruction features specifically involves: disassembling the segmented curve execution instruction features according to the data frame format to extract instruction payload data and instruction header features; matching and retrieving the instruction payload data with the device's built-in mapping table to obtain the expected base temperature value; extracting the baseline ambient temperature deviation data and the preset compensation scaling constant of the operating environment; performing a product operation on the baseline ambient temperature deviation data and the preset compensation scaling constant to set the compensation coefficient; and inputting the instruction header features into the logic parsing stack to extract the compensation coefficient.

[0012] As a further aspect of the present invention, the specific steps of S5 are as follows: S501: Call the dynamic heat deficit indicator of the temperature zone, obtain the instantaneous operating flow rate of the gas and the low heating value of the gas, perform a multiplication operation on the instantaneous operating flow rate of the gas and the low heating value of the gas to obtain the theoretical heat release, obtain the heat loss reduction coefficient, perform a product operation on the theoretical heat release and the heat loss reduction coefficient to generate the equivalent input heat of the gas. S502: Based on the dynamic heat deficit indicator of the temperature zone and the equivalent input heat of the gas, the dynamic heat deficit indicator of the temperature zone and the equivalent input heat of the gas are spliced ​​to establish a multi-dimensional thermal vector, the hidden layer weight matrix and the bias are obtained, the inner product of the multi-dimensional thermal vector and the hidden layer weight matrix is ​​calculated to obtain the intermediate variable, the intermediate variable and the bias are added to obtain the quota value, and the gas calorific value segment quota parameters are generated. S503: Obtain the baseline flow area and the pressure damping constant; divide the gas calorific value segmented quota parameter by the pressure damping constant to obtain the target area requirement; subtract the target area requirement from the baseline flow area and divide by the target area requirement to obtain the area difference ratio; multiply the area difference ratio by the range conversion coefficient to obtain the pulse command variable; adjust the opening parameter of the gas valve of the porcelain insulator kiln to obtain the green control parameters of the firing process.

[0013] As a further aspect of the present invention, the process of obtaining the hidden layer weight matrix and the bias amount specifically involves: reading a preset multidimensional thermal sample set; performing a backward gradient update iteration operation on the multidimensional thermal sample set to obtain a weight array and a set of bias constants; extracting the target weight array and the target bias constant set corresponding to the iteration error being less than a preset convergence constant; assigning the target weight array as the hidden layer weight matrix; and assigning the target bias constant set as the bias amount.

[0014] A smart control system for the green firing process of porcelain insulators, the system comprising: The temperature zone collaborative sensing module collects the operating temperature of the temperature zone in front of the kiln, the current operating temperature of the temperature zone in the kiln, and the operating temperature of the temperature zone behind the kiln. It calculates the temperature difference between the first adjacent temperature zone and the temperature difference between the second adjacent temperature zone, and compares the first historical adjacent operating temperature difference and the second historical adjacent operating temperature difference from the previous sampling period to generate the first temperature difference change direction feature, the second temperature difference change direction feature, and the temperature zone heat penetration status identifier. The curve stage analysis module extracts the starting temperature before the curve inflection point, the ending temperature before the curve inflection point, the duration of the curve inflection point, the starting temperature after the curve inflection point, the target temperature after the curve inflection point, and the set time after the curve inflection point, based on the heat penetration status identifier of the temperature zone. It then calculates the temperature rise slope characteristics before the inflection point, the temperature rise slope characteristics after the inflection point, and the temperature rise slope difference characteristics. The loading condition discrimination module extracts the loading thickness of the porcelain insulator blank from the production configuration database based on the temperature rise slope difference feature, calculates the slope transition judgment threshold parameter, and uses the temperature zone heat penetration status identifier, temperature rise slope difference feature and slope transition judgment threshold parameter to classify and generate segmented curve execution instruction features. The temperature control execution monitoring module configures the green energy-saving set temperature of each temperature zone according to the segmented curve execution command characteristics, and calculates the temperature deviation data and dynamic heat deficit indicator of each temperature zone by combining the actual monitored temperature and temperature deviation duration of each temperature zone. The gas supply regulation module collects the instantaneous operating flow rate and low-level heating value of the gas based on the dynamic heat deficit indicator of the temperature zone, calculates the equivalent input heat of the gas, constructs the gas calorific value segmented quota parameters, and adjusts the opening parameters of the gas valve of the porcelain insulator kiln.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by incorporating the temperature difference between adjacent temperature zones and the direction of historical temperature difference changes into the same judgment chain, a continuous identification of the heat penetration state of the temperature zones is formed. The transition judgment benchmark is established based on the difference in temperature rise slope before and after the curve inflection point and the loading thickness of the billet, so that the segmented curve execution is more in line with the heating response of the billet. The duration of the deviation between the target temperature and the actual temperature is converted into dynamic heat deficit. Combined with the gas flow rate and the low-level heating value, the segmented calorific value quota is determined. The gas valve adjustment shifts from experience correction to heat demand matching, reducing over-burning or under-burning at the inflection point stage, improving the continuity of thermal balance in each temperature zone, reducing ineffective gas consumption, and enhancing batch firing consistency. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the steps of the present invention; Figure 2 This is a detailed schematic diagram of S1 of the present invention; Figure 3 This is a detailed schematic diagram of S2 of the present invention; Figure 4 This is a detailed schematic diagram of S3 of the present invention; Figure 5 This is a detailed schematic diagram of S4 of the present invention; Figure 6 This is a detailed schematic diagram of S5 of the present invention; Figure 7 This is a system module diagram of the present invention. Detailed Implementation

[0018] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0019] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0020] Please see Figure 1 This invention provides an intelligent control method for the green firing process of porcelain insulators, comprising the following steps: S1: Collect the operating temperature of the front temperature zone of the kiln, the current operating temperature zone of the kiln, and the operating temperature zone of the rear temperature zone of the kiln. Calculate the operating temperature of the front temperature zone of the kiln and the current operating temperature zone of the kiln to generate the first adjacent temperature zone temperature difference. Calculate the operating temperature of the current operating temperature zone of the kiln and the operating temperature zone of the rear temperature zone of the kiln to generate the second adjacent temperature zone temperature difference. Collect the first historical adjacent operating temperature difference and the second historical adjacent operating temperature difference from the previous sampling period. Compare the directions of the first adjacent temperature zone temperature difference, the first historical adjacent operating temperature difference, the second adjacent temperature zone temperature difference, and the second historical adjacent operating temperature difference to generate the first temperature difference change direction feature and the second temperature difference change direction feature. Classify and filter the first temperature difference change direction feature and the second temperature difference change direction feature to generate a temperature zone heat penetration status identifier. S2: Extract the starting temperature before the curve inflection point, the ending temperature before the curve inflection point, the duration of the curve inflection point, the starting temperature after the curve inflection point, the target temperature after the curve inflection point, and the set time of the curve inflection point based on the heat penetration status identifier of the temperature zone. Calculate the starting temperature before the curve inflection point, the ending temperature before the curve inflection point, and the duration of the curve inflection point to generate the temperature rise slope feature before the inflection point. Calculate the starting temperature after the curve inflection point, the target temperature after the curve inflection point, and the set time of the curve inflection point to generate the temperature rise slope feature after the inflection point. Calculate the temperature rise slope difference feature between the temperature rise slope feature before the inflection point and the temperature rise slope feature after the inflection point. S3: Extract the loading thickness of the porcelain insulator blank from the production configuration database based on the temperature rise slope difference feature, calculate the loading thickness of the porcelain insulator blank to generate the slope transition judgment threshold parameter, and classify the temperature zone heat penetration status identifier, temperature rise slope difference feature and slope transition judgment threshold parameter to generate segmented curve execution instruction features. S4: Configure the green energy-saving setting temperature of each temperature zone according to the segmented curve execution command characteristics, collect the actual monitored temperature and temperature deviation duration of each temperature zone, calculate the temperature deviation data of each temperature zone by combining the green energy-saving setting temperature of each temperature zone with the actual monitored temperature of each temperature zone, and calculate the temperature deviation data of each temperature zone with the temperature deviation duration to generate the dynamic heat deficit indicator of the temperature zone. S5: Collect the instantaneous operating flow rate and low heating value of the gas according to the dynamic heat deficit indicator of the temperature zone. Calculate the equivalent input heat of the gas using the instantaneous operating flow rate and low heating value of the gas. Calculate the segmented quota parameters of the gas calorific value using the dynamic heat deficit indicator of the temperature zone and the equivalent input heat of the gas. Adjust the opening parameters of the gas valve of the porcelain insulator kiln according to the segmented quota parameters of the gas calorific value to obtain the green control parameters of the firing process.

[0021] Please see Figure 2 The specific steps of S1 are as follows: S101: Collect the operating temperature of the front temperature zone, the current temperature zone, and the rear temperature zone of the kiln. Perform a subtraction operation between the operating temperature of the front temperature zone and the current temperature zone to obtain the temperature difference between the first adjacent temperature zones. Perform a subtraction operation between the operating temperature of the current temperature zone and the operating temperature of the rear temperature zone to obtain the temperature difference between the second adjacent temperature zones. Combine the values ​​of the temperature difference between the first and second adjacent temperature zones to generate the current sampled temperature difference vector. Three sets of K-type thermocouple temperature sensors, arranged on the side wall of the porcelain insulator firing kiln, continuously collect real-time temperature data of each temperature zone within the kiln at a sampling frequency of 10 Hz. The operating temperatures of the front, current, and rear temperature zones at the same timestamp are extracted. After acquiring this temperature data, the front and current temperature zones are subtracted from each other in the processor to obtain the first adjacent temperature zone temperature difference. This calculation involves subtracting the current temperature zone from the front temperature zone temperature. Subsequently, the same subtraction operation is performed between the current and rear temperature zones to obtain the second adjacent temperature zone temperature difference. The extracted first and second adjacent temperature zone differences are then combined through a data concatenation operation to generate a current sampled temperature difference vector composed of two dimensions. In actual industrial kiln operation scenarios, the industrial control system reads the operating temperature of the front temperature zone as 1250.5 degrees Celsius, the current temperature zone as 1265.0 degrees Celsius, and the rear temperature zone as 1282.5 degrees Celsius at the current timestamp. Substituting 1250.5 and 1265.0 into a subtraction operation, 1250.5 - 1265.0 = -14.5, yielding a temperature difference of -14.5 degrees Celsius for the first adjacent temperature zone. Substituting 1265.0 and 1282.5 into a subtraction operation, 1265.0 - 1282.5 = -17.5, yielding a temperature difference of -17.5 degrees Celsius for the second adjacent temperature zone. Combining -14.5 and -17.5 generates the current sampled temperature difference vector [-14.5, -17.5].

[0022] S102: Call the current sampled temperature difference vector, obtain the first historical adjacent operating temperature difference and the second historical adjacent operating temperature difference in the previous sampling period in the local device, perform bitwise subtraction calculation on the current sampled temperature difference vector, the first historical adjacent operating temperature difference and the second historical adjacent operating temperature difference to obtain the time difference variable, obtain the direction determination benchmark, compare the time difference variable with the direction determination benchmark to extract the direction sign, establish a mapping matrix based on the direction sign, and generate a set of temperature difference direction change features; The process of obtaining the direction determination benchmark is as follows: collect the instrument temperature measurement noise amplitude and the base error parameter, perform an addition operation on the instrument temperature measurement noise amplitude and the base error parameter to obtain the error-free stable boundary, configure the error-free stable boundary as the upward determination threshold, perform a numerical inversion operation on the error-free stable boundary to configure it as the downward determination threshold, and aggregate the upward determination threshold and the downward determination threshold to obtain the direction determination benchmark. The process of extracting the direction symbol by comparing the time difference variable with the direction determination benchmark is as follows: the time difference variable is compared with the upward and downward determination thresholds. When the time difference variable is greater than the upward determination threshold, a positive increment character is output; when the time difference variable is less than the downward determination threshold, a negative decrement character is output; when the time difference variable is within the numerical range of the downward and upward determination thresholds, a zero-value stable character is output. The positive increment character, negative decrement character, and zero-value stable character are combined to form the direction symbol. The process of establishing a mapping matrix based on directional symbols is as follows: according to the hardware topology arrangement order of each temperature zone, the directional symbols are sequentially written into the two-dimensional array storage structure to establish a temperature difference feature mapping matrix. The generated current sampled temperature difference vector is invoked, and a retrieval request is sent to the local historical database to obtain the first and second historical adjacent operating temperature differences stored in the previous sampling period on the local device. These two historical adjacent operating temperature differences are then used to construct a historical temperature difference vector. A bitwise subtraction calculation is performed on the current sampled temperature difference vector and the first and second historical adjacent operating temperature differences. The values ​​of each dimension of the current sampled temperature difference vector are subtracted from the corresponding dimensions of the historical temperature difference vector, resulting in a time difference variable containing two time span characteristics. During the process of obtaining the direction determination benchmark, the instrument's temperature measurement noise amplitude and basis error parameters are collected from the instrument's factory calibration parameter table. These parameters are then substituted into an adder to perform an addition operation to obtain the error-free stationary boundary. This operation involves directly adding the two parameters. The value of the obtained error-free stationary boundary is directly configured as the upward judgment threshold. Simultaneously, a value inversion operation is performed on this error-free stationary boundary, converting the positive value to the corresponding negative value, which is then configured as the downward judgment threshold. By combining the upward and downward thresholds through a combination operation, a direction determination benchmark consisting of upper and lower boundaries is obtained. The direction symbol is extracted by comparing the time difference variable with the direction determination benchmark. Specifically, the time difference variable is compared with the upward and downward thresholds. When the time difference variable is greater than the upward threshold, a positive increment character is output to the register; when the time difference variable is less than the downward threshold, a negative decrement character is output to the register; when the time difference variable is within the range of the downward and upward thresholds (including equal to these two boundary values), a zero-value stable character is output. The output positive increment characters, negative decrement characters, and zero-value stable characters are summarized to form the direction symbol. A mapping matrix is ​​established based on the direction symbol. According to the hardware topology arrangement of each temperature zone, the direction symbols are sequentially written into a two-dimensional array storage structure, and a temperature difference feature mapping matrix is ​​established according to the physical position correspondence of rows and columns. For example, the first historical adjacent operating temperature difference is extracted as -14.0, and the second historical adjacent operating temperature difference is -17.0. The current sampled temperature difference vector is [-14.5, -17.5]. Performing bitwise subtraction: -14.5 - (-14.0) = -0.5, -17.5 - (-17.0) = -0.5, yielding the time difference variable [-0.5, -0.5]. The instrument's temperature measurement noise amplitude is 0.2, and the basis error parameter is 0.1. Performing addition: 0.2 + 0.1 = 0.3, yielding the error-free stationary boundary as 0.3. Setting 0.3 as the upward threshold, and inverting it to get -0.3, sets it as the downward threshold. The direction determination benchmark is [-0.3, 0.3]. Comparing the first value of the time difference variable, -0.5, with the boundary, -0.5 is less than -0.3, outputting a negative decrement character; the second value, -0.5, is also less than -0.3, outputting a negative decrement character. The resulting direction symbol is [negative, negative].Following the topological order from the front temperature zone to the back temperature zone, the symbols are combined and written into a two-dimensional array to generate a temperature difference feature mapping matrix, thus completing the construction of the temperature difference direction change feature set.

[0023] S103: Based on the temperature difference direction change feature set, monitor the temperature zone heating interruption Boolean value and the preset cycle convergence threshold, substitute the temperature difference direction change feature set and the temperature zone heating interruption Boolean value into the Boolean operation unit to obtain the merged state bit sequence, perform classification matching retrieval on the merged state bit sequence and the preset cycle convergence threshold to obtain the corresponding condition attribute dictionary item, and generate the temperature zone heat penetration status identifier. Based on the aforementioned set of temperature difference direction change features, the industrial control system reads the temperature rise interruption Boolean value and the preset cycle convergence threshold. The temperature rise interruption Boolean value is returned by the burner status monitoring node via its tag number status, and the preset cycle convergence threshold is issued by the process setting file. The set of temperature difference direction change features and the temperature rise interruption Boolean value are substituted into the Boolean operation node of the logic processor to obtain a merged state bit sequence. The operation process involves performing a logical AND operation on both. If the temperature difference direction indicates no abnormality and the interruption Boolean value is false, the original state bit feature is retained; otherwise, the feature position is forcibly set to the preset abnormal status code, thus generating a multi-bit merged state bit sequence. A classification matching retrieval is performed on this merged state bit sequence and the preset cycle convergence threshold. The abnormal status count value in the sequence is compared with the preset cycle convergence threshold. If the abnormal status count value is less than or equal to the preset cycle convergence threshold, the corresponding condition attribute dictionary item for normal continued heating is retrieved from the relational database; if it is greater than the preset cycle convergence threshold, the corresponding condition attribute dictionary item for abnormal intervention is retrieved. The extracted conditional attribute dictionary entries are used as the temperature zone heat penetration status identifiers and stored. For example, if the read temperature zone heating interruption boolean value is 0 (false), the status code corresponding to the temperature difference direction change feature set is [1, 1]. After performing a logical AND operation, the resulting merged status bit sequence is [1, 1]. The preset loop convergence threshold is configured to 3. The abnormal status count value in the merged status bit sequence is 0. Comparing 0 and 3, since 0 is less than 3, the normal conditional attribute dictionary entry with the number 101 is extracted through classification matching. This 101 is written to the status register to generate the temperature zone heat penetration status identifier.

[0024] Please see Figure 3 The specific steps of S2 are as follows: S201: Obtain the firing curve configuration set in the local storage device, extract the starting temperature before the curve inflection point, the ending temperature before the curve inflection point, and the duration of the curve inflection point from the firing curve configuration set according to the temperature zone heat penetration status identifier, perform a subtraction operation between the ending temperature before the curve inflection point and the starting temperature before the curve inflection point to obtain the front segment temperature variable, perform a ratio calculation between the front segment temperature variable and the duration of the curve inflection point to generate the temperature rise slope feature before the inflection point. The firing curve configuration set is obtained from the local storage device via a data interface. This configuration set contains full-process temperature and time planning data for porcelain insulators of specific specifications. Based on the aforementioned temperature zone heat penetration status identifier, the corresponding process stage in the firing curve configuration set is located, and the starting temperature, ending temperature, and duration of the curve before the inflection point are extracted. A subtraction operation is performed between the extracted ending temperature and starting temperature of the curve before the inflection point, yielding the preceding temperature variable. The preceding temperature variable is then divided by the preceding duration to generate the temperature rise slope feature before the inflection point. For example, if the temperature zone heat penetration status identifier is 101, locating the preheating stage, the extracted starting temperature of the curve before the inflection point is 1000 degrees Celsius, the ending temperature is 1200 degrees Celsius, and the duration is 50 minutes. Perform a subtraction operation: 1200 - 1000 = 200, resulting in a temperature variable of 200 degrees Celsius. Perform a ratio calculation: 200 / 50 = 4, generating a temperature rise slope characteristic of 4 degrees Celsius per minute before the inflection point.

[0025] S202: Call the temperature zone heat penetration status identifier, retrieve the starting temperature after the curve inflection point, the target temperature after the curve inflection point, and the set time after the curve in the firing curve configuration set, perform a subtraction operation on the target temperature after the curve inflection point and the starting temperature after the curve inflection point to extract the temperature increment of the latter segment, and perform a ratio operation on the temperature increment of the latter segment and the set time after the curve inflection point to generate the temperature rise slope feature of the latter segment of the inflection point. The system retrieves the heat penetration status indicator for that temperature zone and continues searching the firing curve configuration set for the starting temperature, target temperature, and set time of the curve's subsequent segment after the inflection point. A subtraction operation is performed on the extracted target temperature and starting temperature of the curve's subsequent segment, subtracting the starting temperature from the target temperature, to extract the temperature increment. This temperature increment is then divided by the set time, generating the temperature rise slope characteristic of the curve's subsequent segment. For example, if the retrieved starting temperature of the curve's subsequent segment is 1200 degrees Celsius, the target temperature is 1350 degrees Celsius, and the set time is 75 minutes, the subtraction operation (1350 - 1200 = 150) yields a temperature increment of 150 degrees Celsius. The ratio operation (150 / 75 = 2) generates a temperature rise slope characteristic of 2 degrees Celsius per minute.

[0026] S203: Call the temperature rise slope features before the inflection point and the temperature rise slope features after the inflection point, perform a subtraction operation on the temperature rise slope features before the inflection point and the temperature rise slope features after the inflection point to obtain feature difference data, obtain slope benchmark verification parameters, perform numerical comparison between feature difference data and slope benchmark verification parameters to extract absolute value data of difference, and generate temperature rise slope difference features. The temperature rise slope characteristics before and after the inflection point, obtained above, are used. A subtraction operation is performed between these two characteristics, subtracting the former from the latter, to obtain the feature difference data. The slope benchmark verification parameter is retrieved from the control strategy configuration file. This feature difference data is then compared with the slope benchmark verification parameter, and the absolute value of the difference is calculated. This absolute difference data is then configured as the temperature rise slope difference feature. For example, if the temperature rise slope characteristic before the inflection point is 4 and the temperature rise slope characteristic after the inflection point is 2, the subtraction operation is performed: 4 - 2 = 2, resulting in a feature difference data of 2. The obtained slope benchmark verification parameter is 1.5. A numerical comparison is performed between 2 and 1.5 to calculate the absolute value of the difference: absolute value |2 - 1.5| = 0.5. Extract 0.5 as the absolute value of the difference, and generate a temperature rise slope difference feature of 0.5.

[0027] Please see Figure 4 The specific steps of S3 are as follows: S301: Based on the temperature rise slope difference feature, retrieve the local production configuration database, extract the loading thickness of the porcelain insulator blank, obtain the preset foundation ratio coefficient, perform a multiplication operation between the loading thickness of the porcelain insulator blank and the preset foundation ratio coefficient to obtain the thickness-related offset, obtain the standard transition slope constant, multiply the thickness-related offset by the unit conversion equivalent coefficient and perform an addition operation with the standard transition slope constant to generate the slope transition judgment threshold parameter; Based on the aforementioned temperature rise slope difference characteristics, a query request is initiated to the local production configuration database to extract the loading thickness of the current batch of porcelain insulator blanks. A preset base ratio coefficient is obtained from the production process specification document. The loading thickness of the porcelain insulator blank is multiplied by this preset base ratio coefficient, resulting in a thickness-related offset. The standard transition slope constant is obtained from the environmental benchmark table. This thickness-related offset is added to the standard transition slope constant, generating a slope transition judgment threshold parameter. For example, if the current porcelain insulator blank loading thickness is extracted to be 120 mm, and the obtained preset base ratio coefficient is 0.05, multiplication is performed: 120 × 0.05 = 6.0, resulting in a thickness-related offset of 6.0. The obtained standard transition slope constant is 2.5, and addition is performed: 6.0 + 2.5 = 8.5, generating a slope transition judgment threshold parameter of 8.5.

[0028] S302: The temperature zone heat penetration status identifier, temperature rise slope difference feature and slope transition judgment threshold parameter are concatenated dimensionally to establish a multi-dimensional state feature vector. The preset hyperplane weight matrix and classification offset constant are obtained. The multi-dimensional state feature vector and the preset hyperplane weight matrix are calculated by performing an inner product to obtain the feature projection value. The feature projection value and the classification offset constant are added to obtain the classification decision distance variable. The classification decision distance variable and the zero boundary judgment benchmark are compared numerically to extract the category mapping label and generate a piecewise curve execution instruction feature. The aforementioned temperature zone heat penetration status identifier, temperature rise slope difference feature, and slope transition judgment threshold parameter are concatenated dimensionally. Specifically, these three values ​​are arranged sequentially according to a fixed data bit order to establish a multi-dimensional state feature vector composed of three dimensions. A preset hyperplane weight matrix and classification offset constant are obtained from the classifier model parameter library. The multi-dimensional state feature vector is then multiplied by the preset hyperplane weight matrix, and the result is summed to obtain the feature projection value. This feature projection value is then added to the classification offset constant to obtain the classification decision distance variable. The classification decision distance variable is compared with the zero boundary judgment benchmark. If the classification decision distance variable is greater than the zero boundary judgment benchmark, a positive category mapping label is extracted; if it is less than or equal to, a negative category mapping label is extracted. Based on the extracted category mapping labels, a piecewise curve execution instruction feature is generated. For example, the temperature zone heat penetration status identifier is configured with a value of 1 (quantized from 101), the temperature rise slope difference feature is 0.5, and the slope transition judgment threshold parameter is 8.5. A multi-dimensional state feature vector [1, 0.5, 8.5] is established by performing dimensional concatenation. The obtained preset hyperplane weight matrix is ​​[0.2, 1.4, -0.1], and the classification offset constant is 0.3. Inner product calculation is performed: 1×0.2+0.5×1.4+8.5×(-0.1)=0.2+0.7-0.85=0.05, resulting in a feature projection value of 0.05. Addition is performed: 0.05+0.3=0.35, resulting in a classification decision distance variable of 0.35. The zero boundary judgment benchmark is 0. Comparing 0.35 with 0, 0.35 is greater than 0, so the positive category mapping label is extracted, and the corresponding piecewise curve execution instruction feature is the instruction to activate the second stage of temperature control logic.

[0029] Please see Figure 5 The specific steps of S4 are as follows: S401: Call the segmented curve execution instruction characteristics, obtain the local device's built-in node register address set, perform parsing on the segmented curve execution instruction characteristics to extract the expected value of the basic temperature and the compensation coefficient, perform a product operation on the expected value of the basic temperature and the compensation coefficient to obtain the corrected temperature reference value, perform mapping and writing on the corrected temperature reference value and the node register address set, and generate the green energy-saving setting temperature for each temperature zone. The process of parsing and extracting the expected base temperature value and compensation coefficient based on the characteristics of the segmented curve execution instructions is as follows: The characteristics of the segmented curve execution instructions are decomposed according to the data frame format to extract instruction payload data and instruction header features; the instruction payload data is matched and retrieved with the device's built-in mapping table to obtain the expected base temperature value; the reference ambient temperature deviation data and preset compensation scaling constant of the operating environment are extracted; the reference ambient temperature deviation data and the preset compensation scaling constant are multiplied to set the compensation coefficient; and the instruction header features are input into the logic parsing stack to extract the compensation coefficient. The above-mentioned piecewise curve execution instruction characteristics are invoked to obtain the address set of built-in node registers of the local device in the industrial control system. The piecewise curve execution instruction characteristics are parsed, and the instruction payload data and instruction header features are extracted according to the data frame format. This instruction payload data is matched and retrieved against the device's built-in mapping table to extract the corresponding expected base temperature value. Simultaneously, the baseline ambient temperature deviation data and the preset compensation scaling constant are obtained. The baseline ambient temperature deviation data and the preset compensation scaling constant are multiplied to obtain the compensation scaling value. The instruction header features are input into the logic parsing stack for decoding and judgment. If the header features indicate that full compensation is required, the above compensation scaling value is directly set as the compensation coefficient. The expected base temperature value and the compensation coefficient are multiplied to obtain the corrected temperature reference value. The corrected temperature reference value is mapped and written to the node register address set, and the corresponding value is written to the corresponding physical address to generate the green energy-saving setting temperature for each temperature zone. For example, the instruction payload data extracted from the data frame is data code 800, and the instruction header feature is identifier bit A. Matching 800 with the mapping table yields an expected base temperature value of 1250. The extracted baseline ambient temperature deviation data is 2.0, and the preset compensation scaling constant is 1.1. A product operation is performed: 2.0 × 1.1 = 2.2. The logic parsing stack decoder flag A confirms that full compensation is required, and the compensation coefficient is set to 1.02 (based on quantization conversion of 2.2). A product operation is performed: 1250 × 1.02 = 1275, resulting in a corrected temperature reference value of 1275 degrees Celsius. 1275 is written into the register corresponding to the address set, generating a green energy-saving setting temperature of 1275 degrees Celsius for each temperature zone.

[0030] S402: Collect the actual monitored temperature of each temperature zone, perform a dimensional subtraction operation between the actual monitored temperature of each temperature zone and the green energy-saving set temperature of each temperature zone to extract the temperature difference vector, obtain the probe thermal compensation vector, perform a summation operation between the temperature difference vector and the probe thermal compensation vector to obtain the calibration difference matrix, and generate temperature zone temperature deviation data. The process of obtaining the probe thermal compensation vector and summing the temperature difference vector with the probe thermal compensation vector to obtain the calibration difference matrix is ​​as follows: extract the zero-point drift error calibration value and nonlinear feedback correction parameter of the physical temperature measuring probe; perform summation calculation on the zero-point drift error calibration value and nonlinear feedback correction parameter to set the probe thermal compensation vector; and perform term-by-term addition algebraic operation on the temperature difference vector and the probe thermal compensation vector to obtain the calibration difference matrix. The actual monitored temperatures of each temperature zone in the kiln are collected. A dimensional subtraction operation is performed between the actual monitored temperature and the green energy-saving set temperature for each zone. The operation involves subtracting the green energy-saving set temperature from the actual monitored temperature, extracting a temperature difference vector composed of multiple temperature zone differences. During the acquisition of the probe thermal compensation vector, the zero-point drift error calibration value and nonlinear feedback correction parameter of the physical temperature measuring probe are extracted from the periodic calibration records. These zero-point drift error calibration values ​​and nonlinear feedback correction parameters are summed and set as the probe thermal compensation vector. The temperature difference vector and the probe thermal compensation vector are then subjected to a term-by-term algebraic addition operation. Each temperature zone difference within the vector is added to the probe thermal compensation vector, resulting in a calibration difference matrix. The data within this matrix is ​​output to generate temperature zone temperature deviation data. For example, if the actual monitored temperature of the front temperature zone is 1270°C and the green energy-saving set temperature is 1275°C, a subtraction operation is performed: 1270 - 1275 = -5, extracting the current dimension value of the temperature difference vector as -5. The zero-point drift error calibration value is extracted to be 0.5, and the nonlinear feedback correction parameter is 0.3. A summation calculation is performed: 0.5 + 0.3 = 0.8, and the probe thermal compensation vector is set to 0.8. Term-by-term algebraic addition is performed: -5 + 0.8 = -4.2, resulting in an element of -4.2 corresponding to the calibration difference matrix, generating a temperature deviation data of -4.2 degrees Celsius.

[0031] Table 1 Sensor Probe Compensation Data Table Temperature probe number Zero drift error calibration value Nonlinear feedback correction parameters Probe thermal compensation vector Probe No. 1 0.5 0.3 0.8 Probe No. 2 0.4 0.2 0.6 Probe No. 3 0.6 0.4 1.0 Table 1 lists the parameters obtained during periodic calibration of multiple probes and the probe thermal compensation vector calculated by summing and aggregating them.

[0032] S403: The monitoring timer records the corresponding temperature deviation continuous running time, performs time dimension integral multiplication operation on the temperature deviation data and the temperature deviation continuous running time, and combines the equivalent heat capacity coefficient to obtain the cumulative heat deficit value, obtain the calibration damping coefficient, calculate the thermal under-temperature state variable by performing quotient calculation on the cumulative heat deficit value and the calibration damping coefficient, establish the corresponding relationship, and generate the temperature zone dynamic thermal under-temperature indicator. The controller's internal timer module monitors and records the duration of temperature deviation corresponding to the aforementioned temperature deviation data, calculated from the moment the deviation exceeds the limit to the current processing cycle. A time-dimensional integral multiplication operation is performed on the temperature deviation data and its duration. This operation multiplies the absolute value of the temperature deviation data by the duration to obtain the accumulated heat deficit value. The calibration damping coefficient is obtained from the material's thermophysical property configuration file. The cumulative heat deficit value and the calibration damping coefficient are then divided to calculate the quotient, resulting in a thermal under-depression state variable. A correspondence is established between this thermal under-depression state variable and a preset state code, generating a dynamic thermal under-depression identifier for the temperature zone. For example, if the absolute value of the extracted temperature deviation data is 4.2 degrees Celsius, and the timer records a duration of 10 minutes, the integral multiplication operation is performed: 4.2 × 10 = 42.0, yielding a cumulative heat deficit value of 42.0. The obtained calibration damping coefficient is 2.0. The quotient value is calculated as follows: 42.0 / 2.0 = 21.0, resulting in a thermal under-condition variable of 21.0. This 21.0 is then converted into the corresponding quantization code to generate a dynamic thermal under-condition identifier for the temperature zone.

[0033] Please see Figure 6 The specific steps of S5 are as follows: S501: Call the dynamic heat deficit indicator of the temperature zone, obtain the instantaneous operating flow rate of the gas and the low heating value of the gas, perform a multiplication operation on the instantaneous operating flow rate of the gas and the low heating value of the gas to obtain the theoretical heat release, obtain the heat loss reduction coefficient, perform a product operation on the theoretical heat release and the heat loss reduction coefficient to generate the equivalent input heat of the gas. The system invokes the aforementioned dynamic heat deficit indicator for the temperature zone and sends read commands to the flow meter and calorimeter of the main gas supply pipeline to obtain the instantaneous operating flow rate and the lower heating value of the gas at the current moment. A multiplication operation is performed on the instantaneous operating flow rate and the lower heating value of the gas, directly multiplying the flow rate data by the heating value to obtain the theoretically releaseable heat at that flow rate, i.e., the theoretical heat release. The heat loss reduction factor corresponding to the current furnace insulation state is obtained from the kiln thermal evaluation data table. The theoretical heat release is multiplied by the heat loss reduction factor to generate the equivalent input heat of the gas. For example, if the instantaneous operating flow rate of the gas is 50 cubic meters per hour and the lower heating value of the gas is 35 megajoules per cubic meter, the multiplication operation is performed: 50 × 35 = 1750, resulting in a theoretical heat release of 1750 megajoules per hour. The heat loss reduction factor is obtained as 0.8. Perform the product operation: 1750 × 0.8 = 1400, the equivalent input heat of the generated gas is 1400 megajoules per hour.

[0034] S502: Based on the dynamic heat deficit indicator of the temperature zone and the equivalent heat input of the gas, the dynamic heat deficit indicator of the temperature zone and the equivalent heat input of the gas are spliced ​​to establish a multi-dimensional thermal vector, the hidden layer weight matrix and the bias are obtained, the inner product of the multi-dimensional thermal vector and the hidden layer weight matrix is ​​calculated to obtain the intermediate variable, the intermediate variable and the bias are added to obtain the quota value, and the segmented quota parameters of the gas calorific value are generated. The process of obtaining the hidden layer weight matrix and bias is as follows: read a preset multidimensional thermal sample set; perform a backward gradient update iteration operation on the multidimensional thermal sample set to obtain a weight array and a set of bias constants; extract the target weight array and target bias constant set corresponding to the iteration error being less than a preset convergence constant; assign the target weight array as the hidden layer weight matrix; and assign the target bias constant set as the bias. Based on the aforementioned dynamic heat deficit indicator for the temperature zone and the equivalent input heat of the gas, these two data points are concatenated to establish a multidimensional thermodynamic vector composed of the heat deficit state and the input heat. During the acquisition of the hidden layer weight matrix and bias, multiple multidimensional thermodynamic sample sets containing input features and target quotas are read from the historical database. For this multidimensional thermodynamic sample set, a multilayer perceptron network is used to perform inverse gradient update iterations. The network structure includes an input layer, a single-layer hidden layer with 64 neurons, and an output layer, with rectified linear units as the activation function. During the training phase, mean squared error is used as the loss function to calculate the difference between the predicted value and the true target value, and an adaptive moment estimation optimization algorithm is used to adjust the network parameters. After multiple iterations, a weight array and a set of bias constants are obtained. When the iteration error is less than a preset convergence constant, the corresponding target weight array and target bias constant set are extracted. The target weight array is assigned as the hidden layer weight matrix, and the target bias constant set is assigned as the bias. The multidimensional thermal vector is multiplied by the hidden layer weight matrix, and the values ​​of each dimension are multiplied by their corresponding weights and summed to obtain intermediate variables. These intermediate variables are then added to the bias to obtain specific quota values, generating segmented quota parameters for the gas calorific value. For example, the dynamic heat deficit indicator for a temperature zone has a quantified value of 21.0, and the equivalent input heat of the gas is 1400. A multidimensional thermal vector [21.0, 1400] is constructed. After model training, the hidden layer weight matrix is ​​[0.5, 0.01], and the bias is 10. The inner product calculation is performed: 21.0 × 0.5 + 1400 × 0.01 = 10.5 + 14.0 = 24.5, resulting in an intermediate variable of 24.5. The summation operation is performed: 24.5 + 10 = 34.5, resulting in a quota value of 34.5, and generating segmented quota parameters for the gas calorific value of 34.5.

[0035] S503: Obtain the baseline flow area and gas pressure damping constant, divide the gas calorific value segment quota parameter with the gas pressure damping constant to obtain the target area requirement, subtract the target area requirement from the baseline flow area and divide by the target area requirement to obtain the area difference ratio, multiply the area difference ratio with the range conversion coefficient to obtain the pulse command variable, adjust the opening parameter of the gas valve of the porcelain insulator kiln to obtain the green control parameters of the firing process; The process of multiplying the area difference ratio and the range conversion coefficient to obtain the pulse command variable is as follows: extract the full-load stroke parameter and full-load signal parameter from the preset hardware configuration table; divide the full-load signal parameter and the full-load stroke parameter to obtain the unit signal ratio constant; configure the unit signal ratio constant as the range conversion coefficient; multiply the area difference ratio and the range conversion coefficient to obtain the continuous pulse value; round down the continuous pulse value to extract the integer part; and set the integer part value as the pulse command variable. Obtain the reference flow area of ​​the burner nozzle and the corresponding pressure damping constant of the pipeline environment. Divide the above-mentioned gas calorific value segmented quota parameters by the pressure damping constant. The calculation process is to divide the quota parameter by the damping constant to obtain the target area requirement. Subtract the target area requirement from the reference flow area. The calculation process is to subtract the reference flow area from the target area requirement and divide the difference by the reference flow area to extract the area difference ratio. Extract the valve actuator full-load stroke parameters and full-load signal parameters from the preset hardware configuration table. Divide the full-load signal parameters by the full-load stroke parameters to obtain the unit signal ratio constant, and configure this unit signal ratio constant as the range conversion coefficient. Multiply the area difference ratio by the range conversion coefficient to obtain the continuous pulse value. Round down the continuous pulse value, discard the decimal part, and extract the integer part. Set the integer part as the pulse command variable, and generate the opening parameters of the porcelain insulator kiln gas valve to be sent to the actuator through the output port. For example, the gas calorific value tiered quota parameter is 34.5, and the gas pressure damping constant is 1.5. A division calculation is performed: 34.5 / 1.5 = 23.0, resulting in a target area requirement of 23.0 square centimeters. The baseline flow area is 20.0 square centimeters. A subtraction calculation is performed to extract the ratio: (23.0-20.0) / 20.0 = 0.15, resulting in an area difference ratio of 0.15. The full-load stroke parameter is extracted as 100 mm, and the full-load signal parameter is 4000 pulses. A division calculation is performed: 4000 / 100 = 40, configuring the range conversion coefficient as 40. A multiplication operation is performed: 0.15 × 40 = 6.0, resulting in a continuous pulse value of 6.0. A round-down operation is performed to extract the integer digit value as 6. 6 is set as the pulse command variable, ultimately generating the gas valve opening parameters for the porcelain insulator kiln, controlling the valve's action, and obtaining the green control parameters for the firing process.

[0036] Please see Figure 7 A smart control system for the green firing process of porcelain insulators, comprising: The temperature zone collaborative sensing module collects the operating temperature of the temperature zone in front of the kiln, the current operating temperature of the temperature zone in the kiln, and the operating temperature of the temperature zone behind the kiln. It calculates the temperature difference between the first adjacent temperature zone and the temperature difference between the second adjacent temperature zone, and compares the first historical adjacent operating temperature difference and the second historical adjacent operating temperature difference from the previous sampling period to generate the first temperature difference change direction feature, the second temperature difference change direction feature, and the temperature zone heat penetration status identifier. The curve stage analysis module extracts the starting temperature before the curve inflection point, the ending temperature before the curve inflection point, the duration of the curve inflection point, the starting temperature after the curve inflection point, the target temperature after the curve inflection point, and the set time of the curve inflection point, based on the heat penetration status indicator of the temperature zone. It also calculates the temperature rise slope characteristics before the inflection point, the temperature rise slope characteristics after the inflection point, and the temperature rise slope difference characteristics. The loading condition discrimination module extracts the loading thickness of the porcelain insulator blank from the production configuration database based on the temperature rise slope difference feature, calculates the slope transition judgment threshold parameter, and uses the temperature zone heat penetration status identifier, temperature rise slope difference feature and slope transition judgment threshold parameter to classify and generate segmented curve execution instruction features. The temperature control execution monitoring module configures the green energy-saving set temperature of each temperature zone according to the segmented curve execution command characteristics, and calculates the temperature deviation data and dynamic heat deficit indicator of each temperature zone by combining the actual monitored temperature and the continuous running time of temperature deviation of each temperature zone. The gas supply regulation module collects the instantaneous operating flow rate and low heating value of gas based on the dynamic heat deficit indicator of the temperature zone, calculates the equivalent input heat of gas, constructs the gas calorific value segmented quota parameters, adjusts the opening parameters of the gas valve of the porcelain insulator kiln, and obtains the green control parameters of the firing process.

[0037] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of protection of the described technical solutions.

Claims

1. A method for intelligent control of green firing process of porcelain insulator, characterized in that, Includes the following steps: S1: Collect the operating temperature of the front temperature zone of the kiln, the current operating temperature of the kiln, and the operating temperature of the rear temperature zone of the kiln. Calculate the temperature difference between the first and second adjacent temperature zones. Combine the first and second historical adjacent operating temperature differences from the previous sampling period with the direction comparison to generate the first temperature difference change direction feature, the second temperature difference change direction feature, and the temperature zone heat penetration status identifier. S2: Based on the heat penetration status identifier of the temperature zone, extract the starting temperature before the curve inflection point, the ending temperature before the curve inflection point, the duration of the curve inflection point, the starting temperature after the curve inflection point, the target temperature after the curve inflection point, and the set time of the curve inflection point, and calculate the temperature rise slope characteristics before the inflection point, the temperature rise slope characteristics after the inflection point, and the temperature rise slope difference characteristics. S3: Based on the temperature rise slope difference feature, extract the loading thickness of the porcelain insulator blank from the production configuration database, calculate the slope transition judgment threshold parameter, and use the temperature zone heat penetration status identifier, temperature rise slope difference feature and slope transition judgment threshold parameter to classify and generate segmented curve execution instruction features. S4: Configure the green energy-saving setting temperature of each temperature zone according to the segmented curve execution instruction characteristics, and calculate the temperature deviation data and dynamic heat deficit indicator of each temperature zone by combining the actual monitored temperature and the continuous running time of the temperature deviation of each temperature zone. S5: Based on the dynamic heat deficit indicator of the temperature zone, collect the instantaneous operating flow rate of the gas and the low heating value of the gas, calculate the equivalent input heat of the gas, construct the segmented quota parameters of the gas calorific value, adjust the opening parameters of the gas valve of the porcelain insulator kiln, and obtain the green control parameters of the firing process.

2. The intelligent control method for green firing process of porcelain insulator according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Collect the operating temperature of the front temperature zone, the current temperature zone, and the rear temperature zone of the kiln. Perform a subtraction operation between the operating temperature of the front temperature zone and the current temperature zone to obtain the temperature difference between the first adjacent temperature zones. Perform a subtraction operation between the operating temperature of the current temperature zone and the operating temperature of the rear temperature zone to obtain the temperature difference between the second adjacent temperature zones. Combine the values ​​of the temperature difference between the first and second adjacent temperature zones to generate the current sampled temperature difference vector. S102: Call the current sampled temperature difference vector to obtain the first historical adjacent operating temperature difference and the second historical adjacent operating temperature difference in the previous sampling period in the local device. Perform bitwise subtraction on the current sampled temperature difference vector, the first historical adjacent operating temperature difference, and the second historical adjacent operating temperature difference to calculate the time difference variable. Obtain the direction determination benchmark. Compare the time difference variable with the direction determination benchmark to extract the direction sign. Establish a mapping matrix based on the direction sign to generate a set of temperature difference direction change features. S103: Based on the set of temperature difference direction change features, monitor the temperature zone heating interruption Boolean value and the preset cycle convergence threshold, substitute the set of temperature difference direction change features and the temperature zone heating interruption Boolean value into the Boolean operation unit to obtain the merged state bit sequence, perform classification matching retrieval on the merged state bit sequence and the preset cycle convergence threshold to obtain the corresponding condition attribute dictionary item, and generate the temperature zone heat penetration status identifier.

3. The intelligent control method for green firing process of porcelain insulator according to claim 2, characterized in that, The process of obtaining the direction determination benchmark is as follows: collecting the instrument temperature measurement noise amplitude and the base error parameter; performing an addition operation on the instrument temperature measurement noise amplitude and the base error parameter to obtain the error-free stable boundary; configuring the error-free stable boundary as the upward determination threshold; performing a numerical inversion operation on the error-free stable boundary to configure it as the downward determination threshold; and aggregating the upward determination threshold and the downward determination threshold to obtain the direction determination benchmark. The process of extracting the direction symbol by numerically comparing the time difference variable with the direction determination benchmark specifically involves comparing the time difference variable with the upward determination threshold and the downward determination threshold. When the time difference variable is greater than the upward determination threshold, a positive increment character is output; when the time difference variable is less than the downward determination threshold, a negative decrement character is output; and when the time difference variable is within the numerical range of the downward determination threshold and the upward determination threshold, a zero-value stable character is output. The positive increment character, the negative decrement character, and the zero-value stable character are then combined to form the direction symbol.

4. The intelligent control method for the green firing process of porcelain insulators according to claim 1, characterized in that, The specific steps of S2 are as follows: S201: Obtain the firing curve configuration set in the local storage device, extract the curve inflection point start temperature, curve inflection point end temperature and curve inflection point duration from the firing curve configuration set according to the temperature zone heat penetration status identifier, perform a subtraction operation between the curve inflection point end temperature and the curve inflection point start temperature to obtain the inflection point temperature variable, perform a ratio calculation between the inflection point temperature variable and the curve inflection point duration to generate the inflection point inflection point temperature rise slope feature; S202: Call the temperature zone heat penetration status identifier, retrieve the curve inflection point start temperature, curve inflection point target temperature and curve inflection point set time from the firing curve configuration set, perform a subtraction operation between the curve inflection point target temperature and the curve inflection point start temperature to extract the inflection point temperature increment, perform a ratio operation between the inflection point temperature increment and the curve inflection point set time to generate the inflection point temperature rise slope feature; S203: Call the temperature rise slope feature before the inflection point and the temperature rise slope feature after the inflection point, perform a subtraction operation on the temperature rise slope feature before the inflection point and the temperature rise slope feature after the inflection point to obtain feature difference data, obtain slope benchmark verification parameters, perform numerical comparison between the feature difference data and the slope benchmark verification parameters to extract the absolute value of the difference data, and generate temperature rise slope difference feature.

5. The intelligent control method for the green firing process of porcelain insulators according to claim 1, characterized in that, The specific steps for S3 are as follows: S301: Based on the temperature rise slope difference feature, retrieve the local production configuration database, extract the loading thickness of the porcelain insulator blank, obtain the preset foundation ratio coefficient, perform a multiplication operation between the loading thickness of the porcelain insulator blank and the preset foundation ratio coefficient to obtain the thickness-related offset, obtain the standard transition slope constant, multiply the thickness-related offset by the unit conversion equivalent coefficient and perform an addition operation with the standard transition slope constant to generate the slope transition judgment threshold parameter; S302: The temperature zone heat penetration status identifier, the temperature rise slope difference feature, and the slope transition judgment threshold parameter are concatenated dimensionally to establish a multi-dimensional state feature vector. A preset hyperplane weight matrix and a classification offset constant are obtained. The multi-dimensional state feature vector and the preset hyperplane weight matrix are used to calculate the inner product to obtain the feature projection value. The feature projection value and the classification offset constant are added to obtain the classification decision distance variable. The classification decision distance variable is compared with the zero boundary judgment benchmark to extract the category mapping label and generate a piecewise curve execution instruction feature.

6. The intelligent control method for the green firing process of porcelain insulators according to claim 1, characterized in that, The specific steps of S4 are as follows: S401: Call the segmented curve execution instruction feature, obtain the local device's built-in node register address set, perform parsing and extract the expected value of the basic temperature and the compensation coefficient for the segmented curve execution instruction feature, perform a product operation on the expected value of the basic temperature and the compensation coefficient to obtain the corrected temperature reference value, perform mapping and writing of the corrected temperature reference value and the node register address set to generate the green energy-saving setting temperature for each temperature zone. S402: Collect the actual monitored temperature of each temperature zone, perform a dimensionless subtraction operation between the actual monitored temperature of each temperature zone and the green energy-saving set temperature of each temperature zone to extract the temperature difference vector, obtain the probe thermal compensation vector, perform a summation operation between the temperature difference vector and the probe thermal compensation vector to obtain the calibration difference matrix, and generate temperature zone temperature deviation data. S403: The monitoring timer records the corresponding temperature deviation duration. The time dimension integral multiplication operation is performed on the temperature deviation data and the duration of the temperature deviation. The accumulated heat deficit value is obtained by combining the equivalent heat capacity coefficient. The calibration damping coefficient is obtained. The quotient value of the accumulated heat deficit value and the calibration damping coefficient is calculated to obtain the underheating state variable. A corresponding relationship is established to generate a dynamic heat deficit identifier for the temperature zone.

7. The intelligent control method for the green firing process of porcelain insulators according to claim 6, characterized in that, The process of parsing and extracting the expected base temperature value and compensation coefficient based on the piecewise curve execution instruction features specifically involves: disassembling the piecewise curve execution instruction features according to the data frame format to extract instruction payload data and instruction header features; matching the instruction payload data with the device's built-in mapping table to obtain the expected base temperature value; extracting the reference ambient temperature deviation data and the preset compensation scaling constant of the operating environment; performing a product operation on the reference ambient temperature deviation data and the preset compensation scaling constant to set the compensation coefficient; and inputting the instruction header features into the logic parsing stack to extract the compensation coefficient.

8. The intelligent control method for the green firing process of porcelain insulators according to claim 1, wherein step S5 is as follows: S501: Call the dynamic heat deficit indicator of the temperature zone, obtain the instantaneous operating flow rate of the gas and the low heating value of the gas, perform a multiplication operation on the instantaneous operating flow rate of the gas and the low heating value of the gas to obtain the theoretical heat release, obtain the heat loss reduction coefficient, perform a product operation on the theoretical heat release and the heat loss reduction coefficient to generate the equivalent input heat of the gas. S502: Based on the dynamic heat deficit indicator of the temperature zone and the equivalent input heat of the gas, the dynamic heat deficit indicator of the temperature zone and the equivalent input heat of the gas are spliced ​​to establish a multi-dimensional thermal vector, the hidden layer weight matrix and the bias are obtained, the inner product of the multi-dimensional thermal vector and the hidden layer weight matrix is ​​calculated to obtain the intermediate variable, the intermediate variable and the bias are added to obtain the quota value, and the gas calorific value segment quota parameters are generated. S503: Obtain the baseline flow area and the pressure damping constant; divide the gas calorific value segmented quota parameter by the pressure damping constant to obtain the target area requirement; subtract the target area requirement from the baseline flow area and divide by the target area requirement to obtain the area difference ratio; multiply the area difference ratio by the range conversion coefficient to obtain the pulse command variable; adjust the opening parameter of the gas valve of the porcelain insulator kiln to obtain the green control parameters of the firing process.

9. The intelligent control method for the green firing process of porcelain insulators according to claim 8, characterized in that, The process of obtaining the hidden layer weight matrix and bias is specifically as follows: reading a preset multidimensional thermal sample set; performing a backward gradient update iteration operation on the multidimensional thermal sample set to obtain a weight array and a set of bias constants; extracting the target weight array and target bias constant set corresponding to the iteration error being less than a preset convergence constant; assigning the target weight array as the hidden layer weight matrix; and assigning the target bias constant set as the bias.

10. An intelligent control system for the green firing process of porcelain insulators, characterized in that, The system is used to implement the intelligent control method for the green firing process of porcelain insulators according to any one of claims 1-9, the system comprising: The temperature zone collaborative sensing module collects the operating temperature of the temperature zone in front of the kiln, the current operating temperature of the temperature zone in the kiln, and the operating temperature of the temperature zone behind the kiln. It calculates the temperature difference between the first adjacent temperature zone and the temperature difference between the second adjacent temperature zone, and compares the first historical adjacent operating temperature difference and the second historical adjacent operating temperature difference from the previous sampling period to generate the first temperature difference change direction feature, the second temperature difference change direction feature, and the temperature zone heat penetration status identifier. The curve stage analysis module extracts the starting temperature before the curve inflection point, the ending temperature before the curve inflection point, the duration of the curve inflection point, the starting temperature after the curve inflection point, the target temperature after the curve inflection point, and the set time after the curve inflection point, based on the heat penetration status identifier of the temperature zone. It then calculates the temperature rise slope characteristics before the inflection point, the temperature rise slope characteristics after the inflection point, and the temperature rise slope difference characteristics. The loading condition discrimination module extracts the loading thickness of the porcelain insulator blank from the production configuration database based on the temperature rise slope difference feature, calculates the slope transition judgment threshold parameter, and uses the temperature zone heat penetration status identifier, temperature rise slope difference feature and slope transition judgment threshold parameter to classify and generate segmented curve execution instruction features. The temperature control execution monitoring module configures the green energy-saving set temperature of each temperature zone according to the segmented curve execution command characteristics, and calculates the temperature deviation data and dynamic heat deficit indicator of each temperature zone by combining the actual monitored temperature and temperature deviation duration of each temperature zone. The gas supply regulation module collects the instantaneous operating flow rate and low-level heating value of the gas based on the dynamic heat deficit indicator of the temperature zone, calculates the equivalent input heat of the gas, constructs the gas calorific value segmented quota parameters, adjusts the gas valve opening parameters of the porcelain insulator kiln, and obtains the green control parameters for the firing process.

Images