Intelligent temperature control production system for environment-friendly asphalt cold patch additive

Abstract

The invention discloses an intelligent temperature control production system for an environment-friendly asphalt cold patch additive, and relates to the technical field of asphalt cold patch additive production, and the system comprises a controller initial modeling module which is used for recording the currently set proportional-integral-differential controller parameters, and constructing a reference control feature vector used for representing the initial response behavior of a controller; and the thermal load response identification module is used for applying micro-amplitude step type temperature disturbance to the to-be-heated environment-friendly asphalt cold patch additive and extracting a thermal load characteristic vector representing the thermal response characteristic of the additive. According to the method, intelligent matching and full-time-domain response evaluation are realized by constructing control and thermal load feature vectors and combining cross-correlation and grey correlation analysis; a sliding spectrum entropy and self-adaptive parameter correction mechanism is introduced, a closed-loop regulation system is constructed, the suitability, stability and additive dispersing performance of a temperature control system are effectively improved, and the problems of control lag and fluctuation are solved.

Description

Technical Field

[0001] This invention relates to the field of asphalt cold patch additive production technology, specifically to an environmentally friendly intelligent temperature control production system for asphalt cold patch additives. Background Technology

[0002] The intelligent temperature-controlled production system for environmentally friendly cold-patch asphalt additives refers to an automated production device for preparing environmentally friendly cold-patch asphalt materials. It integrates intelligent temperature control, multi-component quantitative proportioning, and dynamic stirring and dispersion technology to achieve precise heating, feeding, and uniform mixing of additives (such as polymer modifiers, tackifiers, and environmentally friendly emulsifiers) under different working conditions. The system automatically adjusts the heating temperature and stirring rate based on the external ambient temperature, additive properties, and production batch settings, ensuring that the additives are fully dispersed in the asphalt masterbatch without thermal degradation, thereby improving the adhesion and workability of the cold-patch material in low-temperature and humid environments. Simultaneously, the system integrates data acquisition and anomaly early warning modules, achieving green and controllable production throughout the entire process, reducing energy consumption and exhaust emissions, and meeting the dual requirements of environmental friendliness and ready-to-use in rapid road repair.

[0003] The existing technology has the following shortcomings: In the intelligent temperature-controlled production process of existing environmentally friendly asphalt cold patch additives, adaptive PID control algorithms are often used to perform closed-loop temperature regulation of the heating unit. However, in practical applications, if the initial control parameters of the system fail to adequately match the thermal load response characteristics of the target additive, continuous oscillations can easily occur during the temperature control process. This means the temperature fluctuates frequently around the set target value, making it difficult to maintain a stable constant temperature range. Such temperature fluctuations cause the additive to repeatedly undergo nonlinear heating and cooling processes during the heating phase, leading to molecular fatigue, instability of the phase interface energy state, and degradation of its physicochemical properties. This, in turn, reduces its dispersion uniformity and interfacial adhesion performance in the asphalt matrix. These problems not only affect the mechanical consistency of the cold patch but also easily lead to early failure phenomena such as low-temperature cracking and material peeling after actual paving, severely restricting the reliability and service life of this type of environmentally friendly cold patch material in road rapid repair scenarios.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent temperature-controlled production system for environmentally friendly asphalt cold patch additives. This invention constructs control and thermal response feature vectors and introduces time-domain cross-correlation and grey relational analysis to achieve dynamic adaptation modeling of the temperature control system for different environmentally friendly asphalt cold patch additives. This effectively improves the matching accuracy of control parameters and solves the problems of adjustment lag and oscillation in traditional temperature control. Simultaneously, by combining sliding spectral entropy analysis and an adaptive parameter correction mechanism, a closed-loop intelligent adjustment system is formed. This system can identify changes in control stability in real time and automatically optimize control parameters, ensuring consistent temperature response and stable heating process. This significantly improves the additive dispersion effect and the bonding performance and service life of the cold patch, thus solving the problems mentioned in the background technology.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an environmentally friendly intelligent temperature control production system for asphalt cold patching additives, comprising a controller initial modeling module, a thermal load response identification module, a thermal load coupling analysis module, a grey relational dynamic evaluation module, a control response monitoring module, and an adaptive parameter adjustment module; The controller initial modeling module records the currently set proportional-integral-derivative controller parameters during the no-load preheating stage of the heating system, and synchronously collects the heating power output data within the corresponding time series of the preheating stage to construct a benchmark control feature vector to characterize the initial response behavior of the controller. The thermal load response identification module applies a small step temperature disturbance to the environmentally friendly asphalt cold patch additive to be heated within the same time window as the no-load preheating stage, and monitors the temperature change curve of the key area of ​​the additive in real time. Based on the temperature rise rate and its hysteresis response behavior, it extracts the thermal load feature vector characterizing the thermal response characteristics of the additive. The thermal load coupling analysis module performs time-domain cross-correlation analysis on the reference control eigenvector and the thermal load eigenvector, extracts the response lag time and amplitude gain deviation between the two, and constructs an initial matching degree matrix to describe the dynamic coupling relationship between the proportional-integral-derivative controller and the thermal load. The grey relational dynamic evaluation module inputs the initial matching degree matrix into the improved grey relational analysis model with a time-series weight dynamic adjustment mechanism, and outputs a normalized multidimensional matching coefficient sequence that characterizes the consistency of the control response across the entire time domain. The control response monitoring module performs sliding spectral entropy analysis on the matching coefficient sequence, calculates the degree of nonlinear fluctuation and stability boundary of the temperature control production system response in the time domain, and generates a dynamic matching degree deviation index for quantifying the quality of control adaptation. The adaptive parameter adjustment module compares the dynamic matching degree deviation index with the preset matching degree threshold in real time. If the matching degree deviation exceeds the preset matching degree threshold, the adaptive control correction mechanism is triggered to dynamically adjust the proportional, integral and derivative parameters in the proportional-integral-derivative controller to adapt to the thermal load response characteristics of different environmentally friendly asphalt cold patch additives, thereby improving the control stability of the temperature control process and the consistency of the additive's thermal response behavior.

[0007] Preferably, constructing a baseline control feature vector to characterize the initial response behavior of the controller includes the following sub-steps: Initialize the heating system and ensure it operates under no-material-load conditions. Read and record the currently set proportional, integral, and derivative parameters of the proportional-integral-derivative controller through the control terminal as the initial controller parameters. During the process of the heating system entering stable no-load heating, the output power data of the heating unit within the preset time window is collected in real time, and the data is timestamped and recorded synchronously to form a complete heating power time series. The parameters of the above proportional-integral-derivative controller and the collected heating power time series are normalized in terms of feature dimensions to extract key dynamic behavior features that reflect the response rate, settling time, and power fluctuation range of the temperature control production system. Based on the feature extraction results, a joint multidimensional vector set containing control parameter dimension and power response dimension is constructed as a benchmark control feature vector to characterize the initial dynamic behavior of the proportional-integral-derivative controller, which can then be used for subsequent matching degree analysis and controller adaptation evaluation with thermal load response behavior.

[0008] Preferably, the extraction of the thermal load feature vector characterizing the thermal response properties of environmentally friendly asphalt cold patch additives includes the following sub-steps: After the heating system reaches a stable state, within the same time window as the no-load preheating stage, the control unit applies a step-type temperature disturbance signal within a set amplitude range to the heating unit to simulate the instantaneous thermal change process under actual working conditions, ensuring that the disturbance amplitude does not cause material degradation. During the application of a step disturbance, high-precision temperature sensors deployed in the key thermal response area of ​​the environmentally friendly asphalt cold patch additive are used to record the temperature change curve in real time and capture the instantaneous response characteristics with a high sampling rate to form a complete disturbance response temperature dataset. Based on the collected temperature change data, multiple dynamic indicators such as temperature rise rate, response onset delay time, peak response time, and settling time are calculated. Several response characteristic parameters that can reflect the sensitivity and temporal characteristics of the additive to thermal input are extracted, including temperature rise onset delay time, temperature rise rate per unit time, maximum temperature response amplitude, time required to reach steady state, fluctuation amplitude during the response process, and slope change rate of the temperature response curve, which are used to comprehensively characterize the dynamic thermal response capability and time response characteristics of the additive under thermal disturbance. After normalizing the feature dimensions of the above response feature parameters (refer to the specific content of the feature dimension normalization process above), a thermal load feature vector is constructed. This vector is used to perform coupling and matching analysis with the controller's reference control feature vector to evaluate whether the current proportional-integral-derivative controller setting is suitable for the thermal response characteristics of the additive to be heated, thereby providing a basis for subsequent parameter tuning and adaptive control.

[0009] Preferably, performing time-domain cross-correlation analysis on the baseline control eigenvector and the thermal load eigenvector to construct an initial matching degree matrix includes the following sub-steps: The constructed baseline control feature vector and thermal load feature vector are aligned on the time axis to ensure that they have the same time resolution and sampling interval, so as to guarantee the accuracy and consistency of the cross-correlation calculation process. The time-domain cross-correlation between the baseline control feature vector and the thermal load feature vector is performed using a sliding time window method. The correlation coefficient sequence reflecting the similarity of their response processes is extracted, and the time lag point corresponding to the maximum cross-correlation coefficient is identified. Based on the time lag point corresponding to the maximum cross-correlation coefficient, the response lag time and amplitude gain deviation between the reference control output and the thermal load response are calculated and used as the measurement indicators of control output advance or lag and energy coupling efficiency, respectively. The extracted response lag time and amplitude gain deviation are converted into normalized metrics to construct an initial matching degree matrix, which is used for subsequent grey relational analysis and adaptability optimization judgment of controller-thermal load coupling performance.

[0010] Preferably, when analyzing the dynamic relationship between the reference control characteristic vector and the thermal load characteristic vector, a sliding time window approach is used. One vector (usually the thermal load characteristic vector) is progressively shifted forward or backward along the time axis, and compared point-by-point with the other vector (the reference control characteristic vector). By calculating the cross-correlation coefficient at each sliding position, a sequence of correlation coefficients reflecting the degree of similarity between the two vectors as they change with time offset is obtained. This cross-correlation analysis method is essentially a time-series matching technique that reveals the degree of synchronization between two time series at different time offsets. After obtaining this series of correlation coefficients, the system identifies the time offset point corresponding to the largest cross-correlation coefficient. This time point is the time delay at which the thermal load response is most strongly coupled with the control signal, i.e., the so-called "time lag point." This lag time quantifies whether the controller's response to the thermal load is timely and whether there is a serious delay or advance. It is an important indicator for evaluating the degree of dynamic matching between control and load, and helps to subsequently make targeted corrections to the control parameters.

[0011] Preferably, the initial matching degree matrix is ​​input into an improved grey relational analysis model with a dynamic adjustment mechanism for time-series weights to output a normalized multidimensional matching coefficient sequence, including the following sub-steps: The system receives an initial matching degree matrix consisting of response lag time and amplitude gain deviation, and performs normalization transformation on each element in the initial matching degree matrix to keep all parameters within a uniform numerical range, ensuring numerical comparability and computational stability during the analysis process. To address the importance of matching between controller output and thermal load response within different time windows, a dynamic adjustment mechanism for time weights is introduced. Different time weight factors are assigned based on the degree of impact of the early, middle, and steady-state response on the overall system performance, thereby achieving differentiated emphasis on the matching degree analysis in the time dimension. Using an improved grey relational analysis model, the grey relational degree of the normalized initial matrix is ​​calculated with the ideal response benchmark sequence at each time segment, and weighted fusion is performed by combining the weights of each time series to generate a multidimensional matching coefficient sequence that characterizes the dynamic consistency of the entire temperature control production system. The obtained matching coefficient sequence is used as an intermediate result to measure the adaptability of the control strategy, and is used for subsequent sliding spectral entropy analysis and adaptive parameter correction to ensure that the proportional-integral-derivative controller can achieve a comprehensive evaluation and self-optimization of response consistency and steady-state control performance under different thermal load conditions.

[0012] Preferably, the specific steps for performing sliding spectral entropy analysis on the matching coefficient sequence to generate a dynamic matching degree deviation index are as follows: Obtain the normalized multidimensional matching coefficient sequence output by the improved grey relational analysis model, and divide the entire sequence into several continuous but overlapping local subsequences based on a fixed-length time sliding window in order to extract local statistical features. Discrete distribution modeling is performed on the matching coefficients in each sliding subsequence, and its frequency distribution histogram is calculated. Based on this, the information entropy calculation formula is applied to obtain the spectral entropy value of the matching coefficients within the corresponding time window, thereby quantifying the disorder and fluctuation amplitude of the system response. For each spectral entropy value, a time weighting factor is further introduced to emphasize the different sensitivities of the temperature-controlled production system to the control quality in the initial and steady-state stages of the response. A weighted spectral entropy change curve that evolves over time is constructed to reveal the unstable sections that may exist in the dynamic response process of the control system. By analyzing the deviation between the spectral entropy values ​​at each time point in the weighted spectral entropy curve and the minimum spectral entropy value under ideal steady-state conditions, a dynamic matching degree deviation index sequence is formed. This sequence can be used to judge the response consistency and stability of the temperature control production system at different stages in real time, providing high-resolution feedback data support for the adaptive parameter correction module.

[0013] Preferably, the dynamic matching degree deviation index is compared with the preset matching degree threshold in real time and an adaptive control correction mechanism is triggered. The specific steps are as follows: During the operation of the temperature-controlled production system, the dynamic matching degree deviation index sequence generated in the previous stage is received and updated in real time, and the dynamic matching degree deviation value corresponding to each time node is compared one-to-one with the preset matching degree threshold to determine whether the current control strategy has the risk of adaptation failure or response fluctuation exceeding the limit. When the dynamic matching degree deviation value at any time exceeds the set matching degree threshold, the embedded adaptive control correction mechanism is immediately triggered, the control parameters are updated, and the current temperature control production system status is used as a feedback input signal to guide the direction of parameter correction. Based on historical data of control deviation characteristics and thermal load response behavior, optimization algorithms (such as fuzzy adjustment, genetic algorithm or gradient update mechanism) are called to jointly optimize and adjust the proportional parameters, integral parameters and derivative parameters to ensure that the parameter adjustment is targeted and convergent, while avoiding new unstable factors in the adjustment process. The updated proportional-integral-derivative controller parameters are written to the control unit in real time, and the response behavior of the updated temperature control production system is continuously monitored to ensure that it falls back into the allowable range of the matching degree threshold, forming a closed-loop dynamic adjustment and performance verification mechanism. This enables rapid adaptation to the thermal load characteristics of different environmentally friendly asphalt cold patching additives and continuous improvement of the consistency of response throughout the entire process.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention constructs a baseline control feature vector and a thermal load feature vector, and introduces time-domain cross-correlation analysis and an improved grey relational analysis model to model the dynamic coupling relationship between the output behavior of a proportional-integral-derivative controller and the thermal response of environmentally friendly asphalt cold patching additives. This results in a full-time-domain, multi-dimensional control matching coefficient sequence. This data-driven feature vector coupling and matching analysis method overcomes the dependence of traditional empirical temperature control strategies on parameter tuning. It can automatically identify the response characteristics of different thermal load materials under actual working conditions, thereby significantly improving the adaptation accuracy of the control strategy and the targeting of the adjustment objective. It effectively solves the problems of control lag, temperature oscillation, and large adjustment errors existing in the prior art.

[0015] This invention introduces sliding spectral entropy analysis and dynamic matching degree deviation evaluation methods into the control feedback mechanism, combined with an adaptive parameter correction mechanism, forming a closed-loop learning and autonomous adjustment intelligent control system. By real-time monitoring of control response fluctuations and stability boundaries, and feeding back deviation exceedance information to the controller parameter layer for dynamic adjustment, the proportional-integral-derivative controller can continuously adapt to the thermal response changes of different types or batches of environmentally friendly asphalt cold patching additives, ensuring stable temperature distribution, rapid response, and controllable fluctuations during the heating process. This mechanism significantly improves the robustness and consistency of the temperature control system under varying heat load conditions, thereby improving the dispersion effect of additives in asphalt masterbatch and enhancing the adhesion performance and long-term service stability of cold patching materials. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0017] Figure 1 This is a schematic diagram of a module of an intelligent temperature-controlled production system for an environmentally friendly asphalt cold patching additive according to the present invention. Detailed Implementation

[0018] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0019] This invention provides, for example Figure 1The environmentally friendly asphalt cold patching additive intelligent temperature control production system shown includes a controller initial modeling module, a thermal load response identification module, a thermal load coupling analysis module, a grey relational dynamic evaluation module, a control response monitoring module, and an adaptive parameter adjustment module. The controller initial modeling module records the currently set proportional-integral-derivative controller parameters during the no-load preheating stage of the heating system, and synchronously collects the heating power output data within the corresponding time series of the preheating stage to construct a benchmark control feature vector to characterize the initial response behavior of the controller. Constructing a baseline control feature vector to characterize the initial response behavior of the controller includes the following sub-steps: Initialize the heating system and ensure it operates under no-material-load conditions. Read and record the currently set proportional, integral, and derivative parameters of the proportional-integral-derivative controller through the control terminal as the initial controller parameters. During the process of the heating system entering stable no-load heating, the output power data of the heating unit within the preset time window is collected in real time, and the data is timestamped and recorded synchronously to form a complete heating power time series. The parameters of the above proportional-integral-derivative controller and the collected heating power time series are normalized in terms of feature dimensions to extract key dynamic behavior features that reflect the response rate, settling time, and power fluctuation range of the temperature control production system. Based on the feature extraction results, a joint multidimensional vector set containing control parameter dimension and power response dimension is constructed as a benchmark control feature vector to characterize the initial dynamic behavior of the proportional-integral-derivative controller, which can then be used for subsequent matching degree analysis and controller adaptation evaluation with thermal load response behavior.

[0020] Feature dimension normalization refers to the process of extracting numerical features from the proportional-integral-derivative (PID) controller parameters and the collected heating power time series. To eliminate differences in numerical scale, unit, or dimension among the various parameters and power data, a unified mathematical mapping method (such as min-max normalization, Z-score standardization, or fractional scaling) is used to transform all feature values ​​to the same order of magnitude or a normalized interval (such as [0, 1] or a normal distribution with a mean of 0 and a variance of 1). This avoids the unbalanced impact of certain dimensions on the overall results in subsequent multidimensional vector analysis due to excessively large or small values, ensuring that parameters from different sources have a unified basis for comparison and statistical weight when calculating feature similarity, distance measurement, or correlation. This process is a crucial prerequisite for achieving the comparability and computability of high-dimensional control vectors.

[0021] During the no-load preheating phase of the heating system, the currently set proportional-integral-derivative (PID) controller parameters are recorded, and heating power output data within the corresponding time series of the preheating phase are simultaneously collected. A baseline control feature vector is constructed to characterize the controller's initial response behavior. This vector provides a standardized and quantifiable control behavior reference template for subsequent adaptation analysis between the control system and the thermal load. Specifically, under no-load conditions, the system is not affected by any external material thermal load. The controller's output behavior is primarily driven by its internal parameter settings. Therefore, the heating power output data collected during this phase accurately reflects the dynamic response characteristics of the PID controller to the temperature setpoint, such as response speed, power output waveform, settling time, and fluctuations during the adjustment process.

[0022] By jointly extracting features from the controller's initial proportional, integral, and derivative parameters and their control output behavior, and constructing a baseline control feature vector after normalization, not only can the controller's characteristics be digitally expressed in multiple dimensions, but it can also serve as a reference vector in the matching degree calculation during subsequent time-domain correlation analysis with thermal load response behavior. This step enables the entire temperature-controlled production system to possess the fundamental capability for self-sensing and response quality evaluation, and is a core prerequisite for realizing dynamic coupling modeling, adaptive parameter adjustment, and response consistency optimization of the "controller-thermal load" in intelligent temperature-controlled production systems. Therefore, this step plays a fundamental and crucial role in the entire intelligent temperature-controlled production method, and is the technical foundation for ensuring subsequent accurate temperature control and additive thermal response matching capabilities.

[0023] The thermal load response identification module applies a small step temperature disturbance to the environmentally friendly asphalt cold patch additive to be heated within the same time window as the no-load preheating stage, and monitors the temperature change curve of the key area of ​​the additive in real time. Based on the temperature rise rate and its hysteresis response behavior, it extracts the thermal load feature vector characterizing the thermal response characteristics of the additive. Extracting the thermal load feature vector characterizing the thermal response properties of environmentally friendly asphalt cold patching additives includes the following sub-steps: After the heating system reaches a stable state, within the same time window as the no-load preheating stage, the control unit applies a step-type temperature disturbance signal within a set amplitude range to the heating unit to simulate the instantaneous thermal change process under actual working conditions, ensuring that the disturbance amplitude does not cause material degradation. During the application of a step disturbance, high-precision temperature sensors deployed in the key thermal response area of ​​the environmentally friendly asphalt cold patch additive are used to record the temperature change curve in real time and capture the instantaneous response characteristics with a high sampling rate to form a complete disturbance response temperature dataset. Based on the collected temperature change data, multiple dynamic indicators such as temperature rise rate, response onset delay time, peak response time, and settling time are calculated. Several response characteristic parameters that can reflect the sensitivity and temporal characteristics of the additive to thermal input are extracted, including temperature rise onset delay time, temperature rise rate per unit time, maximum temperature response amplitude, time required to reach steady state, fluctuation amplitude during the response process, and slope change rate of the temperature response curve, which are used to comprehensively characterize the dynamic thermal response capability and time response characteristics of the additive under thermal disturbance. After normalizing the feature dimensions of the above response feature parameters (refer to the specific content of the feature dimension normalization process above), a thermal load feature vector is constructed. This vector is used to perform coupling and matching analysis with the controller's reference control feature vector to evaluate whether the current proportional-integral-derivative controller setting is suitable for the thermal response characteristics of the additive to be heated, thereby providing a basis for subsequent parameter tuning and adaptive control.

[0024] Within the same time window as the no-load preheating stage, a small-amplitude step-type temperature perturbation is applied to the environmentally friendly asphalt cold patch additive to be heated, and the temperature change curves of key areas of the additive are monitored in real time. Based on the temperature rise rate and its hysteresis response behavior, a heat load feature vector characterizing the thermal response characteristics of the additive is extracted. Its purpose is to obtain the dynamic response characteristics of the additive to changes in heat input, thereby laying an accurate and personalized thermal response foundation for subsequent adaptability analysis with controller output characteristics. Due to the significant differences in chemical structure, heat capacity, thermal conductivity, and phase change behavior among different types of environmentally friendly additives, their temperature response process to the input power of the heating unit has highly individualized nonlinear characteristics, manifested in differences in temperature rise rate, hysteresis time, steady-state time, and response fluctuations.

[0025] By applying a small-amplitude step-type temperature perturbation, the original response curve of the additive under localized heating perturbation can be extracted without inducing thermal degradation of the material. This curve reflects key characteristics such as its sensitivity to temperature regulation, thermal inertia, and temperature control response hysteresis. Using temperature change data recorded by a high-precision temperature sensor over time, multiple parameters such as temperature rise onset delay, temperature rise rate per unit time, peak response time, and settling time are further extracted and constructed into a unified thermal load feature vector.

[0026] The role of this eigenvector is not only to quantify the thermal response capability of the material itself, but also to serve as a fundamental parameter for determining whether the current proportional-integral-derivative (PID) controller strategy matches the material's thermal characteristics. If there is a deviation in the time domain or amplitude domain between the control output characteristics and the thermal response characteristics, it may cause problems such as oscillations, delays, or overcompensation in the actual operation of the temperature-controlled production system, affecting the stability of temperature control. Therefore, this step is the core link in realizing the coupling and matching analysis of "controller-material thermal response" and is a key technical support for subsequent dynamic parameter optimization and intelligent adaptive adjustment.

[0027] The thermal load coupling analysis module performs time-domain cross-correlation analysis on the reference control eigenvector and the thermal load eigenvector, extracts the response lag time and amplitude gain deviation between the two, and constructs an initial matching degree matrix to describe the dynamic coupling relationship between the proportional-integral-derivative controller and the thermal load. Perform time-domain cross-correlation analysis on the baseline control eigenvector and the thermal load eigenvector to construct an initial matching degree matrix, including the following sub-steps: The constructed baseline control feature vector and thermal load feature vector are aligned on the time axis to ensure that they have the same time resolution and sampling interval, so as to guarantee the accuracy and consistency of the cross-correlation calculation process. The time-domain cross-correlation between the baseline control feature vector and the thermal load feature vector is performed using a sliding time window method. The correlation coefficient sequence reflecting the similarity of their response processes is extracted, and the time lag point corresponding to the maximum cross-correlation coefficient is identified. The specific process is as follows: When analyzing the dynamic relationship between the reference control eigenvector and the thermal load eigenvector, a sliding time window approach is used. One vector (usually the thermal load eigenvector) is progressively shifted forward or backward along the time axis, and compared point-by-point with the other vector (the reference control eigenvector). By calculating the cross-correlation coefficient at each sliding position, a sequence of correlation coefficients reflecting the degree of similarity between the two vectors as they change with time offset is obtained. This cross-correlation analysis method is essentially a time-series matching technique that reveals the degree of synchronization between two time series at different time offsets. After obtaining this series of correlation coefficients, the system identifies the time offset point corresponding to the maximum cross-correlation coefficient. This time point represents the time delay in which the thermal load response is most strongly coupled with the control signal, i.e., the so-called "time lag point." This lag time quantifies whether the controller's response to the thermal load is timely and whether there is a serious delay or advance. It is an important indicator for evaluating the degree of dynamic matching between control and load, and helps in subsequent targeted adjustments to control parameters.

[0028] Based on the time lag point corresponding to the maximum cross-correlation coefficient, the response lag time and amplitude gain deviation between the reference control output and the thermal load response are calculated and used as the measurement indicators of control output advance or lag and energy coupling efficiency, respectively. The extracted response lag time and amplitude gain deviation are converted into normalized metrics to construct an initial matching degree matrix, which is used for subsequent grey relational analysis and adaptability optimization judgment of controller-thermal load coupling performance.

[0029] A time-domain cross-correlation analysis is performed on the reference control eigenvector and the thermal load eigenvector to extract the response lag time and amplitude gain deviation between them. An initial matching degree matrix is ​​constructed to describe the dynamic coupling relationship between the proportional-integral-derivative controller and the thermal load. Its core function is to quantitatively evaluate the dynamic matching degree between the controller output behavior and the thermal response characteristics of the additive, thereby providing an accurate and calculable basis for the adaptability analysis of the control strategy and the adaptive parameter adjustment.

[0030] In practical temperature-controlled production systems, the controller outputs adjustment signals to the heating unit, aiming to quickly and stably bring the controlled object (i.e., the additive) to the set temperature range. However, due to differences in heat capacity, thermal conductivity, molecular structure, and physical state among different types of additives, their response to heat input exhibits unique dynamic characteristics, potentially leading to complex phenomena such as response hysteresis, overshoot, and nonlinear rise. Therefore, relying solely on the controller's set parameters is insufficient to determine the overall actual temperature control effect of the system; a data-driven approach is necessary to model and measure the relationship between control behavior and thermal response.

[0031] In this step, a time-domain cross-correlation analysis of the baseline control eigenvector and the thermal load eigenvector is first performed using a sliding time window method. This systematically reveals the degree of similarity between the two under different time offset conditions. By extracting the time lag point corresponding to the maximum cross-correlation coefficient, the degree of timing misalignment between the control signal and the response signal can be determined. Simultaneously, the amplitude gain deviation of the thermal load during the response process reflects the actual response intensity of the additive to a unit control power. Together, these two aspects constitute the two core dimensions of "timely response" and "sufficient response."

[0032] Furthermore, the two quantitative indicators mentioned above are normalized and integrated into an initial matching degree matrix, enabling the system to evaluate the coupling and adaptation relationship between different control strategies and thermal load responses within a unified mathematical space. This matrix not only provides an objective basis for determining whether the current control settings are suitable, but also provides key inputs for subsequent grey relational analysis, dynamic matching optimization, and adaptive parameter adjustment.

[0033] The grey relational dynamic evaluation module inputs the initial matching degree matrix into the improved grey relational analysis model with a time-series weight dynamic adjustment mechanism, and outputs a normalized multidimensional matching coefficient sequence that characterizes the consistency of the control response across the entire time domain. The initial matching degree matrix is ​​input into an improved grey relational analysis model with a dynamic adjustment mechanism for time-series weights to output a normalized multidimensional matching coefficient sequence, including the following sub-steps: The system receives an initial matching degree matrix consisting of response lag time and amplitude gain deviation, and performs normalization transformation on each element in the initial matching degree matrix to keep all parameters within a uniform numerical range, ensuring numerical comparability and computational stability during the analysis process. To address the importance of matching between controller output and thermal load response within different time windows, a dynamic adjustment mechanism for time weights is introduced. Different time weight factors are assigned based on the degree of impact of the early, middle, and steady-state response on the overall system performance, thereby achieving differentiated emphasis on the matching degree analysis in the time dimension. Using an improved grey relational analysis model, the grey relational degree of the normalized initial matrix is ​​calculated with the ideal response benchmark sequence at each time segment, and weighted fusion is performed by combining the weights of each time series to generate a multidimensional matching coefficient sequence that characterizes the dynamic consistency of the entire temperature control production system. The obtained matching coefficient sequence is used as an intermediate result to measure the adaptability of the control strategy, and is used for subsequent sliding spectral entropy analysis and adaptive parameter correction to ensure that the proportional-integral-derivative controller can achieve a comprehensive evaluation and self-optimization of response consistency and steady-state control performance under different thermal load conditions.

[0034] A multidimensional matching coefficient sequence refers to a set of values ​​reflecting the similarity between the controller output and the thermal load response across multiple time dimensions and performance characteristic dimensions throughout the heating process. This sequence is obtained by inputting the initial matching degree matrix into an improved grey relational analysis model and dynamically adjusting it with time-series weights. It is not merely a single numerical value, but a series of correlation indices that evolve over time. Each index dimension corresponds to a specific response characteristic, such as response hysteresis, amplitude consistency, stability trend, and response rate. By calculating the grey relational degree between the control output and the thermal response at different time points and across multiple performance dimensions, and forming a time-sorted multidimensional array, the changes in the system's control adaptability at different stages can be dynamically tracked. In short, this multidimensional matching coefficient sequence is essentially a time-evolving, structurally complex control adaptability curve used to systematically evaluate the dynamic consistency and coupling between the control strategy and the thermal load response throughout their entire lifecycle, providing a high-resolution data foundation for subsequent judgments on whether adaptive adjustments to the control are necessary.

[0035] The initial matching degree matrix is ​​input into an improved grey relational analysis model with a dynamic adjustment mechanism for time-series weights, and a normalized multidimensional matching coefficient sequence representing the consistency of the control response across the entire time domain is output. Its core function is to systematically, quantitatively, and multidimensionally evaluate the dynamic adaptability between the control system and the thermal load, thereby providing scientific and accurate data for the adaptive adjustment of controller parameters. This step uses grey system theory to model and analyze the response consistency between input and output at different time stages, effectively compensating for the shortcomings of traditional control evaluation methods in comprehensively capturing changes in time-series response.

[0036] Specifically, the initial matching matrix constructed in the previous stage only contains key but static parameters such as response lag time and amplitude gain deviation. However, the thermal load response behavior is essentially a nonlinear, time-varying, and process-dependent dynamic system. Judging solely by a single moment or indicator cannot fully reflect the stability and consistency of the system's control performance. By introducing this initial matrix into an improved grey relational analysis model and combining it with a dynamic adjustment mechanism for time-series weights, we can differentiate the importance of the controller output and thermal load response at different stages. For example, we can emphasize the rapid response in the initial heating phase, the stable adjustment in the middle phase, and the steady-state maintenance in the later phase, thus better aligning with the comprehensive requirements for control performance in actual production.

[0037] The normalized multidimensional matching coefficient sequence output by this model can represent the changes in control adaptability throughout the heating process in the form of a time series, and form a structured matching degree expression for each response characteristic dimension (such as response speed, amplitude stability, regulation efficiency, etc.). This not only helps to discover the time nodes of control performance degradation, but also provides time-feature joint input data for subsequent sliding spectral entropy analysis, further revealing the control stability boundary of the system throughout its entire life cycle.

[0038] The control response monitoring module performs sliding spectral entropy analysis on the matching coefficient sequence, calculates the degree of nonlinear fluctuation and stability boundary of the temperature control production system response in the time domain, and generates a dynamic matching degree deviation index for quantifying the quality of control adaptation. The specific steps for performing sliding spectral entropy analysis on the matching coefficient sequence to generate a dynamic matching degree deviation index are as follows: Obtain the normalized multidimensional matching coefficient sequence output by the improved grey relational analysis model, and divide the entire sequence into several continuous but overlapping local subsequences based on a fixed-length time sliding window in order to extract local statistical features. Discrete distribution modeling is performed on the matching coefficients in each sliding subsequence, and its frequency distribution histogram is calculated. Based on this, the information entropy calculation formula is applied to obtain the spectral entropy value of the matching coefficients within the corresponding time window, thereby quantifying the disorder and fluctuation amplitude of the system response. For each spectral entropy value, a time weighting factor is further introduced to emphasize the different sensitivities of the temperature-controlled production system to the control quality in the initial and steady-state stages of the response. A weighted spectral entropy change curve that evolves over time is constructed to reveal the unstable sections that may exist in the dynamic response process of the control system. By analyzing the deviation between the spectral entropy values ​​at each time point in the weighted spectral entropy curve and the minimum spectral entropy value under ideal steady-state conditions, a dynamic matching degree deviation index sequence is formed. This sequence can be used to judge the response consistency and stability of the temperature control production system at different stages in real time, providing high-resolution feedback data support for the adaptive parameter correction module.

[0039] Sliding spectral entropy analysis is performed on the matching coefficient sequence to calculate the degree of nonlinear fluctuation and stability boundary of the temperature-controlled production system response in the time domain, and to generate a dynamic matching degree deviation index for quantifying the quality of control adaptation. Its core function is to use information entropy theory to refine the identification and quantitative evaluation of the dynamic response stability in the control process, solve the problem that traditional evaluation methods are difficult to capture, such as nonlinear fluctuations and time mismatch, and improve the ability of intelligent temperature-controlled production systems to accurately perceive the control adaptability under different additive heat load conditions.

[0040] This step is based on the normalized multidimensional matching coefficient sequence obtained in the previous stage. This sequence reflects the degree of consistency between the controller output and the thermal load response throughout the heating process. Spectral entropy analysis is performed on this sequence using a sliding time window approach. Within each local time window, the distribution of the matching coefficients is treated as a probability event, and its information entropy value is calculated to reveal the "orderliness" or "fluctuation intensity" of the system response during that period. Higher spectral entropy indicates a more chaotic and volatile temperature control response; lower spectral entropy indicates higher control stability and a clearer response trend.

[0041] Furthermore, by continuously tracking the spectral entropy change curve, it is possible not only to pinpoint when temperature-controlled production experiences abnormal responses, regulatory imbalances, or parameter mismatches, but also to extract the stability boundary, i.e., to identify the limiting threshold at which the control strategy can maintain stable operation under the current thermal load conditions. Based on this, by comparing the deviation with the reference spectral entropy value under ideal steady-state conditions, a dynamic matching degree deviation index sequence reflecting the fluctuation of control quality is generated.

[0042] The generation of this dynamic matching degree deviation index is of great significance: it not only provides a real-time quantitative basis for whether the control system needs to perform adaptive parameter adjustment, but also serves as a criterion for online health judgment, fault tolerance strategy triggering and early warning mechanism activation. It is a key feedback mechanism for realizing closed-loop self-learning and active optimization of the control system.

[0043] The adaptive parameter adjustment module compares the dynamic matching degree deviation index with the preset matching degree threshold in real time. If the matching degree deviation exceeds the preset matching degree threshold, the adaptive control correction mechanism is triggered to dynamically adjust the proportional, integral and derivative parameters in the proportional-integral-derivative controller to adapt to the thermal load response characteristics of different environmentally friendly asphalt cold patch additives, thereby improving the control stability of the temperature control process and the consistency of the additive's thermal response behavior. The dynamic matching degree deviation index is compared with the preset matching degree threshold in real time, and the adaptive control correction mechanism is triggered. The specific steps are as follows: During the operation of the temperature-controlled production system, the dynamic matching degree deviation index sequence generated in the previous stage is received and updated in real time, and the dynamic matching degree deviation value corresponding to each time node is compared one-to-one with the preset matching degree threshold to determine whether the current control strategy has the risk of adaptation failure or response fluctuation exceeding the limit. When the dynamic matching degree deviation value at any time exceeds the preset matching degree threshold, the embedded adaptive control correction mechanism is immediately triggered, the control parameters are updated, and the current temperature control production system status is used as a feedback input signal to guide the direction of parameter correction. Based on historical data of control deviation characteristics and thermal load response behavior, optimization algorithms (such as fuzzy adjustment, genetic algorithm or gradient update mechanism) are called to jointly optimize and adjust the proportional parameters, integral parameters and derivative parameters to ensure that the parameter adjustment is targeted and convergent, while avoiding new unstable factors in the adjustment process. To achieve optimized adjustments based on historical data of control deviation characteristics and thermal load response behavior, a parameter optimization model must first be constructed. This model takes control deviation indicators (such as dynamic matching degree deviation) as input and temperature control performance evaluation functions (such as minimum error, shortest response time, and highest stability) as objectives. In this model, the temperature control production system invokes a class of optimization algorithms, such as fuzzy adjustment, genetic algorithms, or gradient update mechanisms, each corresponding to different optimization strategies: Fuzzy adjustment, based on an expert rule base and membership functions, implements fuzzy logic adjustments to proportional, integral, and derivative parameters, suitable for handling nonlinear fluctuations; genetic algorithms, through population initialization, fitness function evaluation, and crossover and mutation operations, search for the optimal PID parameter combination globally, improving robustness; while gradient update mechanisms can calculate the direction of the partial derivative of the temperature control error with respect to the PID parameters in real time, performing gradual adjustments, suitable for rapid optimization of continuous systems. The algorithm described above constructs constraints or penalty terms by combining historical fluctuation characteristics of thermal load response, hysteresis parameters, and frequency change trends. This effectively prevents parameter adjustment from introducing new oscillations or overshoot problems, ensuring that the final adjustment result achieves the optimal balance between rapid response and system stability. This enables the temperature-controlled production system to achieve precise adaptation and efficient convergence under different additive thermal load conditions.

[0044] The updated proportional-integral-derivative controller parameters are written to the control unit in real time, and the response behavior of the updated temperature control production system is continuously monitored to ensure that it falls back into the allowable range of the matching degree threshold, forming a closed-loop dynamic adjustment and performance verification mechanism. This enables rapid adaptation to the thermal load characteristics of different environmentally friendly asphalt cold patching additives and continuous improvement of the consistency of response throughout the entire process.

[0045] The dynamic matching degree deviation index is compared with the preset matching degree threshold in real time. If the matching degree deviation exceeds the preset matching degree threshold, the adaptive control correction mechanism is triggered, and the proportional, integral and derivative parameters in the proportional-integral-derivative controller are dynamically adjusted. Its function is to enable the temperature control production system to quickly adapt to the changes in the thermal response characteristics of different additives, and ensure that the temperature control behavior is maintained with high stability and high consistency under dynamic working conditions, thereby improving the thermal control accuracy and material performance consistency of the overall production process.

[0046] In intelligent temperature-controlled production systems, the performance of the proportional-integral-derivative (PID) controller is highly dependent on whether its parameter settings match the current thermal load characteristics. Different types or batches of environmentally friendly additives have differences in heat capacity, thermal conductivity, phase change characteristics, and molecular structure, resulting in variations in response time, amplitude, and hysteresis to the same control input. If the controller parameters remain fixed, when the thermal load characteristics shift, the system may experience problems such as control overshoot, temperature oscillation, response delay, or under-regulation, severely affecting the dispersion effect of the additive in asphalt and the final material properties.

[0047] Therefore, the dynamic matching degree deviation index, calculated using methods such as sliding spectral entropy, can reflect the coupling and adaptation between the current control strategy and the thermal response in real time, and compare it with a set threshold. If the deviation exceeds the limit, it indicates that the existing control parameters are no longer applicable, and the system will immediately trigger an adaptive control correction mechanism to enter the parameter adjustment phase. This mechanism, based on historical response data, deviation change trends, and the current system state, uses optimization algorithms (such as fuzzy logic, genetic algorithms, or gradient updates) to jointly optimize the proportional, integral, and derivative parameters, ensuring that parameter adjustments are directional, targeted, and stable.

[0048] Ultimately, the new parameters are loaded into the control system in real time, and its adjustment effect continues to be monitored, forming a closed-loop feedback to ensure that the entire temperature control process remains in an optimal state of rapid response, controllable fluctuations, and stable temperature. This step not only improves the system's adaptability to complex and variable heat loads but also significantly reduces the need for human intervention and parameter adjustments. It is a key step in building a high-precision, self-learning intelligent temperature control production system and is of great significance for ensuring the quality of cold-mixed materials production and their adaptability to construction.

[0049] This invention constructs a baseline control feature vector and a thermal load feature vector, and introduces time-domain cross-correlation analysis and an improved grey relational analysis model to model the dynamic coupling relationship between the output behavior of a proportional-integral-derivative controller and the thermal response of environmentally friendly asphalt cold patching additives. This results in a full-time-domain, multi-dimensional control matching coefficient sequence. This data-driven feature vector coupling and matching analysis method overcomes the dependence of traditional empirical temperature control strategies on parameter tuning. It can automatically identify the response characteristics of different thermal load materials under actual working conditions, thereby significantly improving the adaptation accuracy of the control strategy and the targeting of the adjustment objective. It effectively solves the problems of control lag, temperature oscillation, and large adjustment errors existing in the prior art.

[0050] This invention introduces sliding spectral entropy analysis and dynamic matching degree deviation evaluation methods into the control feedback mechanism, combined with an adaptive parameter correction mechanism, forming a closed-loop learning and autonomous adjustment intelligent control system. By real-time monitoring of control response fluctuations and stability boundaries, and feeding back deviation exceedance information to the controller parameter layer for dynamic adjustment, the proportional-integral-derivative controller can continuously adapt to the thermal response changes of different types or batches of environmentally friendly asphalt cold patching additives, ensuring stable temperature distribution, rapid response, and controllable fluctuations during the heating process. This mechanism significantly improves the robustness and consistency of the temperature control system under varying heat load conditions, thereby improving the dispersion effect of additives in asphalt masterbatch and enhancing the adhesion performance and long-term service stability of cold patching materials.

[0051] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. An environmentally friendly asphalt cold patch additive intelligent temperature control production system, characterized in that, The controller initial modeling module, the thermal load response identification module, the thermal load coupling analysis module, the grey correlation dynamic evaluation module, the control response monitoring module and the adaptive parameter adjustment module are included. The controller initial modeling module records the current set proportional-integral-derivative controller parameters, and constructs a benchmark control feature vector for representing the initial response behavior of the controller. The thermal load response identification module applies a micro-step temperature disturbance to the environmental protection type asphalt cold patch additive to be heated, and extracts a thermal load feature vector representing the thermal response characteristics of the additive. The thermal load coupling analysis module performs time domain cross-correlation analysis on the benchmark control feature vector and the thermal load feature vector, extracts the response lag time and amplitude gain deviation therebetween, and constructs an initial matching degree matrix for describing the dynamic coupling relationship between the proportional-integral-derivative controller and the thermal load. The grey correlation dynamic evaluation module inputs the initial matching degree matrix into an improved grey correlation analysis model with a time sequence weight dynamic adjustment mechanism, and outputs a normalized multi-dimensional matching coefficient sequence. The control response monitoring module performs sliding spectrum entropy analysis on the matching coefficient sequence, and generates a dynamic matching degree deviation index for quantifying the control adaptation quality. The adaptive parameter adjustment module compares the dynamic matching degree deviation index with a preset matching degree threshold in real time, and if it is detected that the matching degree deviation exceeds the preset matching degree threshold, the proportional parameter, the integral parameter and the derivative parameter in the proportional-integral-derivative controller are dynamically adjusted to adapt to the thermal load response characteristics of different environmental protection type asphalt cold patch additives.

2. The intelligent temperature control production system for an environmentally friendly asphalt cold patch additive according to claim 1, characterized in that, The benchmark control feature vector for representing the initial response behavior of the controller is constructed, including the following sub-steps: Initialize the heating system and ensure that it runs without material load, read and record the current set proportional parameter, integral parameter and derivative parameter values of the proportional-integral-derivative controller through the control terminal as the initial controller parameters; During the stable empty load heating process of the heating system, the output power data of the heating unit in the preset time window is collected in real time, and the data is recorded with time stamp synchronization to form a heating power time sequence; The proportional-integral-derivative controller parameters and the collected heating power time sequence are subjected to feature dimension normalization processing to extract key dynamic behavior characteristics reflecting the temperature control production system; Based on the feature extraction results, a joint multi-dimensional vector set containing the control parameter dimension and the power response dimension is constructed as the benchmark control feature vector.

3. The intelligent temperature control production system for an environmentally friendly asphalt cold patch additive according to claim 1, characterized in that, The thermal load feature vector representing the thermal response characteristics of the environmental protection type asphalt cold patch additive is extracted, including the following sub-steps: After the heating system runs to a stable state, a step-type temperature disturbance signal with a set amplitude range is applied to the heating unit through the control unit in the same time window during the empty preheating stage; During the step disturbance application process, the temperature change curve is recorded in real time, and the instantaneous response characteristics are captured to form a disturbance response temperature data set; Based on the collected temperature change data, multiple dynamic indicators are calculated, and multiple response characteristic parameters reflecting the reaction sensitivity and time sequence characteristics of the additive to heat input are extracted from the dynamic indicators; The response characteristic parameters are normalized in feature dimension to construct a thermal load feature vector.

4. The intelligent temperature control production system for an environmentally friendly asphalt cold patch additive according to claim 1, characterized in that, Time-domain cross-correlation analysis is performed on the reference control feature vector and the thermal load feature vector to construct an initial matching degree matrix, including the following sub-steps: The constructed reference control feature vector and the thermal load feature vector are respectively subjected to time axis uniform alignment processing; The time-domain cross-correlation operation between the reference control feature vector and the thermal load feature vector is performed in a sliding time window manner, the correlation coefficient sequence reflecting the similarity of the response processes of the two is extracted, and the time lag point corresponding to the maximum cross-correlation coefficient is identified; Based on the time lag point corresponding to the maximum cross-correlation coefficient, the response lag time and the amplitude gain deviation between the reference control output and the thermal load response are calculated, which are respectively used as the measurement indexes of the control output advancement or lag and the energy coupling efficiency; The extracted response lag time and amplitude gain deviation are converted into normalized measurement values to construct the initial matching degree matrix.

5. The environmentally friendly asphalt cold patch additive intelligent temperature control production system according to claim 4, characterized in that, The specific acquisition process of the correlation coefficient sequence is as follows: In a sliding time window manner, the thermal load feature vector is gradually shifted forward or backward on the time axis, compared with the reference control feature vector point by point, and a group of correlation coefficient sequences reflecting the similarity degree of the two with time offset change are obtained by calculating the cross-correlation coefficients of the two at each sliding position.

6. The environmentally friendly asphalt cold patch additive intelligent temperature control production system according to claim 1, characterized in that, The initial matching degree matrix is input into an improved grey correlation analysis model with a time sequence weight dynamic adjustment mechanism to output a normalized multi-dimensional matching coefficient sequence, including the following sub-steps: Receive the initial matching degree matrix, and perform normalization transformation processing on each element in the initial matching degree matrix; Introduce a time sequence weight dynamic adjustment mechanism to give different time sequence weight factors; Perform grey correlation degree calculation on the normalized initial matrix and the ideal response reference sequence on each time segment, and combine the time sequence weights for weighted fusion to generate a multi-dimensional matching coefficient sequence.

7. The environmentally friendly asphalt cold patch additive intelligent temperature control production system according to claim 1, characterized in that, The specific steps of performing sliding spectral entropy analysis on the matching coefficient sequence to generate a dynamic matching degree deviation index are as follows: Obtain the multi-dimensional matching coefficient sequence, and divide the entire sequence into several continuous but overlapping local subsequences based on a fixed-length time sliding window; Model the matching coefficients in each sliding subsequence, calculate the frequency distribution histogram, and apply the information entropy calculation formula to obtain the spectral entropy value of the matching coefficients in the corresponding time window; For each spectral entropy value, introduce a time weighting factor to construct a weighted spectral entropy change curve that evolves over time; Perform deviation analysis on the spectral entropy values at each time in the weighted spectral entropy curve and the minimum spectral entropy value in the ideal stable state to form a dynamic matching degree deviation index sequence.

8. The environmentally friendly asphalt cold patch additive intelligent temperature control production system according to claim 1, characterized in that, Real-time comparison is performed between the dynamic matching degree deviation index and the preset matching degree threshold to trigger an adaptive control correction mechanism, and the specific steps are as follows: The dynamic matching degree deviation value corresponding to each time node is compared with the preset matching degree threshold one by one, and when the dynamic matching degree deviation value at any time exceeds the set matching degree threshold, the embedded adaptive control correction mechanism is immediately triggered, and the current temperature control production system state is taken as a feedback input signal; Based on the control deviation characteristics and the thermal load response behavior history data, an optimization algorithm is called to jointly optimize and adjust the proportional parameter, integral parameter and differential parameter; The updated proportional-integral-differential controller parameters are written into the control unit in real time, and the response behavior of the updated temperature control production system is continuously monitored to determine whether it falls within the matching threshold range, forming a closed-loop dynamic adjustment and performance verification mechanism.

Images