A multivariable coordinated control method for continuous production of an environment-friendly asphalt waterproof coating

Abstract

The application provides an environment-friendly asphalt waterproof coating continuous production multivariable cooperative control method, which comprises the following steps: firstly, determining a target pipe section prone to forming a wall adhesion layer, collecting main flow process data and wall response data, and generating main flow process data packets, wall response data packets and wall response deviation amounts; secondly, extracting pressure recovery duration, temperature response lag amount, acoustic echo attenuation amplitude and flow slip deviation from an edge computing node to generate effective state records; then, comparing the effective state records with preset interference threshold values to generate executable control instructions; a process control system sequentially executes low-amplitude pulse flow, temperature step, short-time reverse backflow and interface flushing, and controls a recovery branch to receive separated materials within a separation window; finally, generating updated wall active layer state records through review and collection, and updating control boundaries.

Description

Technical Field

[0001] This invention relates to the field of continuous production process control technology for coatings, and in particular to a multivariate collaborative control method for the continuous production of environmentally friendly asphalt waterproof coatings. Background Technology

[0002] The preparation of environmentally friendly asphalt waterproof coatings is gradually shifting from intermittent to continuous production. Current continuous production control technologies largely rely on online instruments to monitor macroscopic process variables such as material temperature, pipeline pressure, discharge viscosity, and equipment load, and then passively adjust the feed rate and process parameters accordingly.

[0003] However, existing control strategies primarily focus on the overall state of the main materials. When black spots, particles, or batch quality fluctuations occur in the product, conventional technologies typically employ hysteresis compensation measures such as increasing shear strength, adding filtration, or shutting down for flushing. This default control logic, which assumes that "normal online indicators represent stable production," makes it difficult to detect and address hidden sources of interference induced by local anomalies in the flow field.

[0004] Specifically, in flow field areas with abrupt changes in flow direction, shear state, or concentrated temperature gradients, such as pipe bends, colloid mill outlets, and static mixer outlets, asphalt heavy components and polymer modifiers readily deposit on the inner walls of the equipment and evolve into a wall adhesion layer. In the early stages of its formation, this adhesion layer often does not cause channel blockage, nor is it sufficient to trigger significant jumps in pipeline pressure or discharge viscosity, resulting in severe blind spots and lags in conventional detection methods.

[0005] Therefore, this invention proposes a multivariate collaborative control method for the continuous production of environmentally friendly asphalt waterproof coatings. The information disclosed in the background section is only for enhancing understanding of the background of this disclosure and may therefore contain prior art information that is not common knowledge to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of existing technologies by providing a multivariate collaborative control method for the continuous production of environmentally friendly asphalt waterproof coatings, thereby solving the technical problems mentioned in the background section.

[0007] To achieve the above objectives, the present invention provides the following technical solution: A multivariate collaborative control method for the continuous production of environmentally friendly asphalt waterproof coatings includes the following steps: S1. The process control system reads equipment operation records, pipeline layout records and abnormal records, identifies the production channel that is prone to forming a wall adhesion layer as the target pipe section, and collects the mainstream process data and wall response data of the target pipe section by the edge computing node, generating mainstream process data package, wall response data package and wall response deviation. S2. The edge computing node confirms that the target pipe section is in a stable state according to the mainstream process data packet, and extracts the pressure recovery time, temperature response hysteresis, acoustic attenuation amplitude and flow slip deviation according to the wall response data packet to generate an effective state record. S3. The edge computing node compares the effective status record with the preset interference threshold to generate an interference status judgment record. The process control system determines the interference type, generates and verifies the wall self-peeling control command, and outputs an executable control command. S4. The process control system executes low-amplitude pulse flow, temperature step, short-time reverse reflux and interface flushing according to executable control commands, controls the recovery branch to receive the separated material within the separation window, and generates an online self-peeling process control record. S5. The edge computing node generates an updated wall active layer status record based on the online self-peeling process control record. The process control system restores the finished product branch discharge or generates a secondary correction record and updates the control boundary.

[0008] S1 specifically includes: the process control system reads equipment operation records, pipeline layout records, historical cleaning records, and historical finished product anomaly records, identifies continuous production channels with abrupt changes in flow direction, shear state, concentrated temperature gradients, or high frequency of historical anomalies as target pipe sections, and generates target pipe section acquisition and configuration records; the edge computing node collects mainstream process data based on the target pipe section acquisition and configuration records, and generates mainstream process data packages after time alignment, field verification, and invalid data deletion; the process control system applies pressure pulse excitation, thermal pulse excitation, and acoustic echo detection to the target pipe section under the mainstream stable state, and the edge computing node collects the corresponding curves and generates wall response data packages and wall response deviations.

[0009] S2 specifically includes: the edge computing node reads the mainstream process data packet, wall response data packet, and wall response deviation; when the target pipe section is in a stable mainstream state, it extracts the pressure recovery time, temperature response hysteresis, acoustic echo attenuation amplitude, and flow slip deviation, and generates a wall feature set; the edge computing node identifies the wall adhesion layer formation state based on the wall feature set, calculates the wall active layer state value, and generates a wall active layer state record containing the wall adhesion layer thickness, wall adhesion strength, wall adsorption activity, and wall detachment risk; the process control system performs consistency verification on the wall active layer state record and marks the records that pass the verification as valid state records.

[0010] S3 specifically includes: the edge computing node reads the valid state record, compares the wall adhesion layer thickness, wall adhesion strength, wall adsorption activity, and wall detachment risk with the corresponding preset interference thresholds, and generates a stable state judgment record or an interference state judgment record; the process control system determines the interference type (thickness-dominated, strength-dominated, adsorption-dominated, detachment-dominated, or combined) based on the interference state judgment record, and generates a wall self-peeling control command; the edge computing node performs on-site safety verification of the wall self-peeling control command, outputs an executable control command after the verification passes, and downgrades parameters or generates a prohibition on self-peeling mark when the verification fails.

[0011] S4 specifically includes: the process control system reads executable control commands and sequentially executes low-amplitude pulse flow, temperature step, short-term reverse reflux, and interface flushing on the target pipe section to control the detachment of the wall adhesion layer and generate a self-peeling execution record; the edge computing node identifies the detachment window based on the target pipe section pressure, acoustic echo signal, and recovery branch flow rate; within the detachment window, the process control system controls the recovery branch to prioritize receiving the detached material and generates a branch diversion record; the edge computing node verifies the pressure, viscosity, recovery capacity, and finished product branch risk based on the self-peeling execution record and branch diversion record, and generates an online self-peeling process control record.

[0012] S5 specifically includes: edge computing nodes reading self-stripping execution records, branch diversion records, and online self-stripping process control records; after the mainstream process data is restored to the allowable disturbance range, they perform verification and collection, generating updated wall active layer status records and wall status recovery amplitude; the process control system compares the updated wall adhesion layer thickness, wall adhesion strength, wall adsorption activity, and wall detachment risk with the preset stability range, generating production recovery records or secondary correction records; and the edge computing nodes updating the preset interference threshold, preset stability range, target pipe section acquisition configuration records, and safety verification boundaries based on the stability state judgment records, production recovery records, secondary correction records, and anomaly handling records.

[0013] The beneficial effects of this invention are as follows: This invention utilizes the collaborative acquisition of mainstream process data and wall response data to identify the formation state of wall adhesion layers at pipeline bends, colloid mill outlets, static mixer outlets, and buffer tank discharge sections. This avoids relying solely on discharge viscosity, temperature, pressure, and flow rate to determine production stability, thereby enabling the early detection of hidden wall interference sources that are difficult to capture with conventional online detection. Wall response data is acquired through pressure pulse excitation, thermal pulse excitation, and acoustic echo detection. Combined with edge computing nodes to extract pressure recovery time, temperature response hysteresis, acoustic echo attenuation amplitude, and flow slip deviation, this invention can distinguish between ordinary residual states and wall adhesion layers with risks of adsorption, deterioration, or detachment, improving the accuracy of wall anomaly identification.

[0014] This invention generates an interference state judgment record based on the wall adhesion layer thickness, wall adhesion strength, wall adsorption activity, and wall detachment risk. After on-site safety verification, executable control commands are output, ensuring that the online self-peeling action is constrained by pressure, temperature, load, recovery capacity, and valve response. This avoids production fluctuations caused by blind flushing or forced peeling. Within the detachment window, the recovery branch prioritizes receiving detached material, and risk identification in the finished product branch prevents material flow containing detached material from entering the finished product branch. This reduces the probability of black spots, hard particles, gel clusters, and incompatible micro-clusters entering the finished product, improving the batch stability of environmentally friendly asphalt waterproof coatings.

[0015] This invention utilizes online self-peeling followed by verification data collection. Based on the updated wall surface active layer status record, it generates a production recovery record or a secondary correction record, and updates preset interference thresholds, preset stability ranges, target pipe section acquisition configuration records, and safety verification boundaries. This creates a closed-loop process control for wall surface adhesion layer identification, online self-peeling, branch diversion, and verification updates. It enables the identification, reduction, and diversion protection of the wall surface adhesion layer with minimal or no downtime, reducing reliance on downtime cleaning, manual inspection, and post-process filtration, resulting in wastewater, waste materials, and production line interruptions. This is beneficial for improving continuous production efficiency and achieving environmentally friendly production goals. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of a multivariate collaborative control method for the continuous production of environmentally friendly asphalt waterproof coatings according to the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example: Figure 1 As shown in the figure, this embodiment provides a multivariate collaborative control method for the continuous production of environmentally friendly asphalt waterproof coatings, including the following steps: S1. The process control system reads equipment operation records, pipeline layout records and abnormal records, identifies the production channel that is prone to forming a wall adhesion layer as the target pipe section, and collects the mainstream process data and wall response data of the target pipe section by the edge computing node, generating mainstream process data package, wall response data package and wall response deviation. S2. The edge computing node confirms that the target pipe section is in a stable state according to the mainstream process data packet, and extracts the pressure recovery time, temperature response hysteresis, acoustic attenuation amplitude and flow slip deviation according to the wall response data packet to generate an effective state record. S3. The edge computing node compares the effective status record with the preset interference threshold to generate an interference status judgment record. The process control system determines the interference type, generates and verifies the wall self-peeling control command, and outputs an executable control command. S4. The process control system executes low-amplitude pulse flow, temperature step, short-time reverse reflux and interface flushing according to executable control commands, controls the recovery branch to receive the separated material within the separation window, and generates an online self-peeling process control record. S5. The edge computing node generates an updated wall active layer status record based on the online self-peeling process control record. The process control system restores the finished product branch discharge or generates a secondary correction record and updates the control boundary.

[0019] S1 specifically includes the following sub-steps: S110 The process control system reads the equipment operation records, pipeline layout records, historical cleaning records, and historical finished product anomaly records of the environmentally friendly asphalt waterproof coating continuous production line. First, it identifies continuous production channels with sudden changes in flow direction, sudden changes in shear state, concentrated temperature gradients, prolonged material residence time, or high frequency of historical anomalies as candidate pipe sections. Then, it sorts the candidate pipe sections according to the number of historical pressure fluctuations, the number of historical cleaning residues, the number of historical black spots or particle anomalies, and whether the conditions for applying excitation and data acquisition are met. The candidate pipe sections with the highest ranking are identified as target pipe sections.

[0020] The target pipe section refers to a continuous production channel where materials are in long-term contact and easily form a wall adhesion layer, including at least one of the following: colloid mill outlet section, static mixer outlet section, pipeline bend section, or buffer tank discharge section.

[0021] The process control system establishes a target pipe segment acquisition configuration record for each target pipe segment. The target pipe segment acquisition configuration record includes the target pipe segment number, target pipe segment location, acquisition cycle, pressure pulse excitation upper limit, thermal pulse excitation upper limit, acoustic echo acquisition frequency, allowable disturbance range, supplementary acquisition rules, and sensor number.

[0022] To prevent detection stimuli from altering the normal continuous production state, the process control system determines the upper limits for pressure pulse excitation and thermal pulse excitation based on the disturbance ratio:

[0023] Where D represents the proportion of disturbance. This refers to the change in pipeline pressure caused by a pressure pulse excitation. The maximum allowable pressure change for the target pipe section. This represents the change in material temperature caused by thermal pulse excitation. The maximum allowable temperature change for the target pipe section. The change in material flow rate caused by the excitation process. The maximum allowable flow rate change for the target pipe section; when the disturbance proportion is not greater than the preset disturbance threshold, the process control system determines that the excitation intensity has not disrupted the stable state of continuous production.

[0024] The target pipe section acquisition configuration record is written to the local configuration area via the edge computing node, and serves as the unified basis for subsequent mainstream process data acquisition, wall response data acquisition, and supplementary acquisition processing.

[0025] S120, the edge computing node collects configuration records based on the target pipe section, and synchronously reads the mainstream process data of the target pipe section in each collection cycle. The mainstream process data includes material temperature, pipeline pressure, material flow rate, discharge viscosity, colloid mill load, metering pump speed, finished product branch valve opening and recovery branch valve opening.

[0026] Edge computing nodes perform time alignment and field verification on mainstream process data. Time alignment refers to unifying the sampled values ​​of different sensors within the same target pipe section and the same acquisition cycle into the same time window. Field verification refers to verifying whether the acquisition time, target pipe section number, equipment number, sensor number, data unit, and sampled value range are complete and consistent.

[0027] For data with missing timestamps, mismatched target pipe segment numbers, unregistered sensor numbers, inconsistent data units, or sampled values ​​exceeding the sensor range or exceeding the allowable disturbance range, the edge computing node marks them as invalid data and deletes them from the mainstream process data packet. The deleted data will not participate in subsequent wall state recognition.

[0028] Edge computing nodes determine whether the current data collection period meets the conditions for subsequent computation based on the proportion of valid data:

[0029] Where R is the proportion of valid data. This refers to the number of data points that pass field verification and range verification within a single data collection period. This represents the total number of data to be collected within the current collection period. When the proportion of valid data is not lower than the preset valid proportion threshold, the edge computing node generates the main process data packet. When the proportion of valid data is lower than the preset valid proportion threshold, the edge computing node generates a supplementary collection marker and records the data packet of the corresponding target pipe segment in the current collection period as the data packet to be retested.

[0030] The mainstream process data package is used to determine whether the mainstream material is in a stable state and serves as background data for subsequent wall response data packages. It is used to eliminate response deviations caused by natural fluctuations in material temperature, material flow rate, pipeline pressure, or colloid mill load.

[0031] S130. When the mainstream process data package meets the effective data ratio requirement and continuous production does not experience flow interruption, pump pressure jump, or abnormal valve switching, the process control system applies pressure pulse excitation, thermal pulse excitation, and acoustic echo detection to the target pipe section according to the target pipe section acquisition and configuration record.

[0032] Pressure pulse excitation is generated by briefly changing the inlet pressure of the target pipe section. The edge computing node collects the pressure attenuation curve to characterize the degree of flow obstruction of materials near the wall of the target pipe section. Thermal pulse excitation is generated by briefly changing the heating or cooling intensity of the outer wall of the target pipe section. The edge computing node collects the temperature response hysteresis curve to characterize the degree of heat transfer obstruction by the wall adhesion layer. Acoustic echo detection is generated by sending a fixed frequency sound wave to the target pipe section and receiving the reflected signal. The edge computing node collects the acoustic echo attenuation curve to characterize the degree of absorption of sound wave propagation by the adhering material on the inner wall of the target pipe section.

[0033] In a specific industrial implementation scenario, the target pipe section is a jacketed insulated pipe with an inner diameter of DN80-DN150. During normal production, the material temperature is controlled between 160℃ and 190℃, and the pipeline base pressure is 0.2MPa-0.6MPa. Pressure pulse excitation is specifically achieved by adjusting the instantaneous speed of the variable frequency feed pump or triggering a high-frequency pressure relief valve connected in parallel to the pipeline. The pulse peak value does not exceed 120% of the pipeline base pressure, and the frequency is controlled between 0.5Hz and 5Hz. The thermal pulse excitation is achieved by instantaneously increasing the flow rate of high-temperature heat transfer oil in the jacket of the target pipe section or activating the auxiliary electromagnetic induction heating coil wound around the outside of the pipe wall, so that the local pipe wall temperature increases by 5℃-15℃ within 10-30 seconds; the acoustic echo detection is achieved by installing a high-temperature resistant ultrasonic probe close to the outer wall of the target pipe section without insulation layer. The probe's emission frequency is preferably 0.5MHz-2MHz to penetrate the metal pipe wall and remain sensitive to changes in the thickness of the asphalt inner wall adhesion layer.

[0034] Edge computing nodes use the target pipe segment number and acquisition cycle number as the main index to write the pressure attenuation curve, temperature response hysteresis curve, acoustic echo attenuation curve, and mainstream process data packets within the same acquisition cycle into the same wall response data packet. When any curve is missing, the acquisition time window is inconsistent with the mainstream process data packet, or the excitation intensity exceeds the allowable disturbance range, the wall response data packet is marked as a data packet to be retested and will not be included in subsequent wall feature extraction.

[0035] Edge computing nodes calculate wall response deviation:

[0036] Where E is the wall response deviation. This represents the deviation of the pressure decay curve from the non-adhesive reference curve. This represents the deviation of the temperature response hysteresis curve relative to the non-adhesive reference curve. This represents the deviation of the acoustic echo attenuation curve from the non-adhesive reference curve. For stress response weights, As temperature response weight, This represents the acoustic response weights. The wall response bias is output along with the wall response data packet for subsequent extraction of the wall feature set.

[0037] S2 specifically includes the following sub-steps: S210 and the edge computing node read the mainstream process data packet, wall response data packet and wall response deviation output by S130, and match them using the target pipe segment number and the acquisition cycle number as indexes. Only data from the same target pipe segment, the same acquisition cycle and not marked as a data packet to be retested are included in the wall feature extraction.

[0038] The edge computing node first determines whether the target pipe section is in a stable mainstream state based on the mainstream process data packets. The judgment fields include the pipeline pressure change rate, material flow rate change rate, discharge viscosity change rate, colloid mill load change rate, finished product branch valve opening change, and recovery branch valve opening change. When all of the above fields are within the allowable disturbance range defined by the target pipe section acquisition configuration record, the target pipe section is determined to be in a stable mainstream state. When there is a sudden pump pressure jump, flow interruption, sudden increase in colloid mill load, abnormal valve switching, or the effective data ratio is lower than the preset effective ratio threshold, the edge computing node marks the current wall response data packet as a data packet affected by mainstream abnormal interference, does not extract the wall feature set, and generates a retest mark to return to S130 for re-acquisition.

[0039] For data determined through the mainstream steady state, the edge computing node uses the pressure decay curve, temperature response hysteresis curve, and acoustic echo decay curve formed by the target pipe segment in the clean state or the previous steady cycle as the reference curve; it generates the pressure recovery time based on the time required for the pressure to recover from the excitation peak to the stable pressure range in the current pressure decay curve; it generates the temperature response hysteresis based on the time difference between the temperature change of the outer wall of the target pipe segment after thermal pulse excitation and the temperature change of the material; it generates the acoustic echo decay amplitude based on the attenuation ratio of the current acoustic echo amplitude relative to the reference curve; and it calculates the deviation of the actual flow velocity from the theoretical flow velocity based on the material flow velocity, pipeline pressure, and valve opening in the mainstream process data package, generating the flow slip deviation.

[0040] Edge computing nodes normalize the pressure recovery time, temperature response hysteresis, acoustic echo attenuation amplitude, and flow slip deviation respectively:

[0041] in, For the normalized value of the i-th type of wall feature, Let be the measured value of the i-th type of wall feature in the current acquisition period. This is the baseline value for the i-th type of wall feature in the target pipe section under the condition of no adhesion or in the previous stable cycle. Let represent the upper limit of the warning value corresponding to the i-th type of wall feature. The i-th type of wall feature includes pressure recovery time, temperature response hysteresis, acoustic echo attenuation amplitude, and flow slip deviation. The corresponding subscripts i are denoted as P, T, S, and V, respectively.

[0042] The edge computing node writes the above normalization results, wall response deviation, target pipe segment number and acquisition cycle number into the wall feature set. The wall feature set is used by S220 to identify the formation state of the wall adhesion layer.

[0043] S220: The edge computing node reads the wall feature set and identifies the formation state of the adhesion layer on the inner wall of the target pipe section based on the normalized values ​​of pressure recovery time, temperature response hysteresis, acoustic echo attenuation amplitude, flow slip deviation, and wall response deviation.

[0044] The wall adhesion layer refers to the thin layer structure formed by the long-term retention of asphalt heavy components, filler fine powder, emulsifier and polymer modifier on the inner wall of the target pipe section. The difference between this thin layer structure and ordinary residue is that it not only adheres to the wall surface, but also continuously adsorbs emulsifier, water, softening oil or light components in newly introduced materials, and there is a risk of shedding and release when the flow rate, temperature or pressure changes.

[0045] Edge computing nodes calculate the wall active layer state value based on the wall feature set. The wall active layer state value is a calculated value used to characterize the activity and interference level of the wall adhesion layer, and does not refer to an independent physical layer structure distinct from the wall adhesion layer.

[0046] Where L is the state value of the active layer on the wall. This is the normalized value for pressure recovery time. This is the normalized value of the temperature response hysteresis. This is the normalized value of the acoustic echo attenuation amplitude. The normalized value of flow slip deviation is denoted as a, where a is the pressure characteristic weight, b is the temperature characteristic weight, c is the acoustic characteristic weight, and d is the flow characteristic weight. Each characteristic weight is preset based on the target pipe section acquisition and configuration records, and the sum of each characteristic weight is 1.

[0047] The setting of each feature weight is highly related to the local flow field characteristics of the target pipe section. For example, when the target pipe section is the outlet section of a colloid mill, due to its high flow velocity and large shear force, the pressure recovery time and flow slip deviation are more sensitive to changes in the adhesion layer. In this case, the pressure feature weight a is set to a range of 0.35-0.45, the flow feature weight d is set to a range of 0.35-0.45, and the temperature feature weight b and acoustic feature weight c are set to 0.05-0.15. When the target pipe section is a low-speed advection region such as the discharge section of a buffer tank, wall deposition is more likely to form and has a significant effect on temperature and sound wave isolation. In this case, the temperature feature weight b is preferentially increased to 0.3-0.4, and the acoustic feature weight c is increased to 0.4-0.5.

[0048] The edge computing node uses the normalized values ​​of temperature response hysteresis and acoustic echo attenuation amplitude to generate a characterization value of wall adhesion layer thickness, the normalized values ​​of pressure recovery time and flow slip deviation to generate a characterization value of wall adhesion strength, the growth of wall response deviation within a continuous acquisition period to generate a characterization value of wall adsorption activity, and the combination of a sudden drop in acoustic echo attenuation amplitude, a shortened pressure recovery time, and an increased flow slip deviation to generate a characterization value of wall detachment risk.

[0049] The wall adsorption activity characterization value is determined according to the wall response deviation growth rate:

[0050] Where G is the wall response deviation growth rate. Let be the wall response deviation in the kth acquisition cycle. For the first The wall response deviation per acquisition cycle The time interval between two adjacent acquisition cycles; when the wall response deviation growth rate exceeds the preset growth threshold, the edge computing node determines that the wall adsorption activity is enhanced.

[0051] If the acoustic echo attenuation increases but the wall response deviation no longer increases during the continuous acquisition period, and the pressure recovery time and flow slip deviation remain stable, it is identified as a normal residual state. If the acoustic echo attenuation, temperature response hysteresis, and wall response deviation continue to increase during the continuous acquisition period, accompanied by an increase in pressure recovery time or flow slip deviation, it is identified as a wall adhesion layer formation state.

[0052] The edge computing node writes the identification results into the wall active layer status record. The wall active layer status record includes the target pipe section number, the acquisition cycle number, the wall active layer status value, the wall adhesion layer thickness characterization value, the wall adhesion strength characterization value, the wall adsorption activity characterization value, and the wall detachment risk characterization value, and serves as the input for S230 consistency verification.

[0053] S230: The process control system reads the wall active layer status record, target pipe section acquisition configuration record, mainstream process data package and wall response data package, and performs consistency verification on the wall active layer status record to avoid incorrect identification results caused by acquisition anomalies, mainstream fluctuations or excitation out-of-bounds leading to subsequent interference status determination.

[0054] Consistency verification includes four categories: First, verify whether the target pipe section number, acquisition cycle number, sensor number, and wall response data packet source number are consistent; Secondly, verify whether the acquisition cycle is continuous, and whether the pressure pulse excitation, thermal pulse excitation, and acoustic echo detection are all within the allowable disturbance range; Third, verify whether the mainstream process data package is still in a stable state and whether there are any sudden pump pressure jumps, flow interruptions, abnormal valve switching, or sudden increases in colloid mill load. Fourth, verify whether the characterization values ​​of wall adhesion layer thickness, wall adhesion strength, wall adsorption activity, and wall detachment risk are complete, valid in range, and consistent with the changing trends of adjacent collection periods.

[0055] When the target pipe segment number is inconsistent, the acquisition cycle is discontinuous, any state quantity is missing, the excitation intensity exceeds the allowable disturbance range, the mainstream process data packet is unstable, or the state quantity jump of the same target pipe segment exceeds the allowable jump range for two consecutive acquisition cycles, the process control system will mark the corresponding wall active layer state record as a record to be retested, and write the reason for retesting, the data source number, and the retest target pipe segment number into the retest record; the record to be retested will not enter the S310 interference state judgment, but will be retained for abnormal cause tracing, and the process control system will return to S130 to re-acquire in the next acquisition cycle.

[0056] For wall active layer status records that pass the consistency verification, the process control system marks them as valid status records. Valid status records include at least the target pipe section number, acquisition cycle number, mainstream stable state marker, wall feature set source number, wall response data packet source number, wall adhesion layer thickness characterization value, wall adhesion strength characterization value, wall adsorption activity characterization value, wall detachment risk characterization value, and retest status marker.

[0057] The valid status record serves as the sole status input for comparing S310 with the preset interference threshold. Target pipe sections that do not generate a valid status record will not trigger wall self-peeling control.

[0058] S3 specifically includes the following sub-steps: S310, the edge computing node reads the valid status record output by S230 and retrieves the corresponding preset interference threshold according to the target pipe segment number. The preset interference threshold is not a uniform fixed value, but is established by the process control system based on the target pipe segment type, historical stable production records, historical abnormal finished product records, and target pipe segment acquisition configuration records. It is used to determine whether the wall adhesion layer has changed from an acceptable adhesion state to an interference state that will affect the quality of the finished product.

[0059] The preset interference thresholds include at least the wall adhesion layer thickness threshold, the wall adhesion strength threshold, the wall adsorption activity threshold, and the wall detachment risk threshold; among them, the verification weight of the wall adhesion strength threshold is increased in the colloid mill outlet section, and the verification weight of the wall adsorption activity threshold and the wall detachment risk threshold is increased in the buffer tank outlet section.

[0060] The edge computing node compares the wall adhesion layer thickness, wall adhesion strength, wall adsorption activity, and wall detachment risk values ​​in the effective state record with the corresponding preset interference thresholds. When none of the four values ​​reach the corresponding preset interference threshold, a stable state determination record is generated. The process control system maintains the current continuous production control parameters, and the stable state determination record is written into the historical stable production record for subsequent updates of the preset interference threshold, preset stability range, and target pipe section acquisition configuration record.

[0061] To avoid false triggering of online self-stripping actions by single sampling fluctuations, the edge computing node first generates an interference flag to be confirmed when any characterization value exceeds the corresponding preset interference threshold for the first time; the edge computing node only generates an interference status judgment record when the same target pipe segment has characterization values ​​exceeding the corresponding preset interference threshold in two consecutive sampling cycles, or when the interference level value exceeds the severe interference threshold.

[0062] The interference level is calculated using the following formula:

[0063] Where C represents the level of interference. This represents the current value of the j-th wall condition index. The wall condition index includes the wall adhesion layer thickness, wall adhesion strength, wall adsorption activity, and wall detachment risk. The preset interference threshold is the value corresponding to the j-th wall state index. Let j be the disturbance weight for the j-th wall state index. This is used to calculate only the portion exceeding a preset interference threshold.

[0064] The interference status determination record includes the target pipe section number, acquisition cycle number, name of the out-of-limit index, current value of the out-of-limit index, corresponding preset interference threshold, interference degree value, interference level, and pending processing mark, and serves as the input for the S320 to generate the wall self-peeling control command.

[0065] S320: The process control system reads the interference status judgment record, determines the interference type of the target pipe section based on the name of the out-of-limit index, the interference degree value, and the interference level, and generates a wall self-peeling control command. The wall self-peeling control command is a process control command used to reduce the wall adhesion layer and prevent detached material from entering the finished product branch. Detached material refers to material that detaches from the inner wall of the target pipe section during the online self-peeling process and may carry components of the wall adhesion layer. The wall self-peeling control command includes low-amplitude pulse flow intensity, short-term reverse backflow duration, temperature step amplitude, interface flushing medium volume, recovery branch early opening time, finished product branch valve opening adjustment amount, and buffer backflow valve opening adjustment amount.

[0066] When the wall adhesion layer thickness exceeds the limit and other characterization values ​​do not reach the corresponding preset interference threshold, the process control system determines the interference type as thickness-dominant interference and prioritizes the generation of low-amplitude pulse flow parameters and temperature step parameters; when the wall adhesion strength exceeds the limit, the interference type is determined as strength-dominant interference and prioritizes the generation of short-time reverse flow parameters; when the wall adsorption activity exceeds the limit, the interference type is determined as adsorption-dominant interference and prioritizes the generation of interface flushing parameters; when the wall detachment risk exceeds the limit, the interference type is determined as detachment-dominant interference and prioritizes the generation of parameters for early opening of the recovery branch and reduction of the opening degree of the finished product branch valve.

[0067] If multiple characterization values ​​exceed the limit simultaneously, the process control system prioritizes the characterization value of wall detachment risk, followed by the characterization value of wall adsorption activity. The characterization values ​​of wall adhesion strength and wall adhesion layer thickness are used as the basis for adjusting the peel strength, generating a combined wall self-peeling control command.

[0068] At the execution level, in order to achieve controlled stripping without interrupting normal continuous production, the target values ​​of various self-stripping control parameters must be limited within physical boundaries that do not compromise the overall compatibility of the materials. Specifically: the velocity variation of low-amplitude pulse flow intensity is controlled between ±10% and ±20% of the normal target pipe section velocity; the duration of short-term reverse backflow is typically controlled within a very short period of 3 to 10 seconds to avoid large-scale backflow of materials; the transient increase in temperature step amplitude is controlled between 8℃ and 15℃, and the continuous heating time does not exceed 2 minutes to prevent the polymer modifier in the asphalt system from undergoing thermo-oxidative aging and chain breakage.

[0069] Various self-stripping control parameters are corrected according to the disturbance level:

[0070] in, The target values ​​for the q-th type of self-peeling control parameters include the low-amplitude pulse current intensity, short-time reverse backflow duration, temperature step amplitude, and interface flushing medium volume. Let be the baseline value for the q-th type of self-peeling control parameter. C is the interference amplification factor, and C is the interference level value obtained from S310.

[0071] The process control system writes the target pipe segment number, interference type, target value of each stripping control parameter, branch switching conditions and execution order into the wall self-stripping control command, and sends the command to the edge computing node for on-site safety verification.

[0072] The S330 edge computing node reads the wall self-peeling control command, the main process data package, the target pipe section acquisition configuration record, and the recovery branch status record to perform on-site safety verification of the wall self-peeling control command. On-site safety verification includes the margin between the current pressure and the upper pressure limit of the target pipe section, the margin between the current material temperature and the upper temperature limit, the colloid mill load margin, the allowable reverse running time of the reflux pump, the remaining capacity of the recovery branch, the closing response time of the finished product branch valve, the adjustable opening degree of the buffer reflux valve, and the allowable addition amount of interface flushing medium. If any verification item fails to meet the requirements, the wall self-peeling control command must not be directly issued to the actuator.

[0073] For self-stripping control parameters that fail on-site safety verification, the edge computing nodes are corrected according to the degradation factor:

[0074] in, The corrected value for the q-th type self-peeling control parameter is generated after the security check fails. The target value of the q-th type self-peeling control parameter before safety verification. This is the downgrade coefficient, which is greater than 0 and less than 1.

[0075] The edge computing node re-performs on-site safety verification using the corrected self-stripping control parameters. If the corrected parameters meet the requirements for pressure, temperature, load, valve response, and recovery capacity, an executable control command is generated. The executable control command includes the target pipe segment number, the low-amplitude pulse flow intensity after safety verification, the short-term reverse backflow duration, the temperature step amplitude, the interface flushing medium volume, the recovery branch early opening time, the finished product branch valve opening adjustment amount, the buffer backflow valve opening adjustment amount, the execution start condition, the execution termination condition, the branch switching condition, the degradation record, and the safety verification result. These commands serve as the direct inputs for S410 to execute online self-stripping and S420 to execute branch diversion.

[0076] If the on-site safety verification requirements cannot be met after continuous downgrading, the edge computing node generates a prohibition self-peeling mark and a conservative diversion instruction, and the finished product branch enters a restricted release state. The restricted release state means that the valve opening of the finished product branch is not higher than the preset safety opening, and the material discharged during this period is marked as material to be inspected or transitional material, and is not directly released as qualified finished product.

[0077] Self-stripping markers and conservative diversion instructions are prohibited from being written into the anomaly handling record. The anomaly handling record is used by S530 to update the safety verification boundary of the target pipe section and to determine in the next control cycle whether to reduce the excitation intensity or extend the verification acquisition cycle.

[0078] S4 specifically includes the following sub-steps: S410: The process control system reads the executable control commands output by S330 and confirms the execution object, execution sequence, and execution boundaries according to the target pipe section number. The execution boundaries include the pressure limit, temperature limit, reflux speed limit, colloid mill load limit, remaining capacity of the recovery branch, and valve closing response time of the finished product branch.

[0079] The process control system first adjusts the recovery branch valve to a pre-opening position, ensuring that the target pipe section has a channel to receive the detached material before the wall adhesion layer detaches. Then, it executes a low-amplitude pulse flow to create a periodic shear difference between the mainstream material in the target pipe section and the wall adhesion layer. Next, it executes a temperature step to reduce the adhesion strength between the wall adhesion layer and the inner wall of the target pipe section. After that, it executes a short-term reverse backflow to cause the loosened wall adhesion layer to detach towards the recovery branch. Finally, it executes interface flushing to reduce the thin layer of material remaining on the inner wall of the target pipe section to the wall state corresponding to the preset stable range.

[0080] Interface flushing refers to the input of a small amount of interface flushing medium compatible with the current coating system into the target pipe section, and the flushed material is guided into the recovery branch through a buffer reflux valve. This part of the material is not allowed to directly enter the finished product branch.

[0081] For environmentally friendly asphalt waterproof coating systems, in order to ensure the efficiency of the flushing action and the subsequent recyclability of waste, the preferred interface flushing medium is a high-temperature softening oil (such as naphthenic oil or aromatic oil) with good miscibility with asphalt, or a high-temperature aqueous soap solution containing a complex surfactant with a concentration 2-5 times higher than that of conventional formulations; the amount of interface flushing medium introduced in a single run is strictly limited to 0.5 to 2.0 times the internal volume of the target pipe section.

[0082] The process control system calculates the percentage of safe operation in real time during execution.

[0083] Where M represents the percentage of execution security. The current target pipe section pressure, The upper limit of the target pipe section pressure. The current material temperature, This is the upper limit of the material temperature. The current reflux rate, This is the upper limit of the reflux speed. For the current colloid mill load, The upper limit of the load on the colloid mill is set. When the execution safety ratio is not greater than the preset safety ratio threshold, the process control system continues to perform online self-stripping. When the execution safety ratio is greater than the preset safety ratio threshold, the process control system reduces the intensity of the low-amplitude pulse current, shortens the duration of the short-term reverse backflow, or suspends interface rinsing.

[0084] After execution, the process control system generates a self-stripping execution record, which includes the target pipe section number, executable control command number, execution start time, execution end time, low-amplitude pulse current intensity, low-amplitude pulse current frequency, short-term reverse backflow duration, reverse backflow direction, temperature step amplitude, temperature step duration, interface flushing medium volume, execution safety percentage, execution interruption flag, and execution completion flag, and serves as the input for S430 process review.

[0085] During the online self-stripping process in S410, edge computing nodes continuously read the target pipe section pressure, acoustic echo signal, and recovery branch flow rate, and identify the wall adhesion layer detachment window based on the above data. The detachment window refers to the time period during which the wall adhesion layer has loosened from the inner wall of the target pipe section and may enter the flow channel with the material. The identification result of the detachment window is used to determine the start and end times of the priority diversion of the recovery branch.

[0086] Edge computing nodes calculate the out-of-window identification value using the following formula:

[0087] Where H is the out-of-window recognition value, This is to measure the change in branch pressure during online self-stripping. The upper limit of the target pipe section pressure. The change in acoustic echo signal. The acoustic echo reference amplitude, To recover real-time traffic from branch lines, To recover the maximum allowable flow rate of the branch, Weighting for pressure changes, As acoustic variation weight, To recover the flow weight; when the detachment window identification value reaches the preset detachment threshold, the process control system determines that the wall adhesion layer has entered the detachment window.

[0088] Before the departure window begins, the process control system keeps the recovery branch valve in a pre-open state and reduces the opening of the finished product branch valve. During the departure window, the system maintains the opening of the recovery branch valve and the corresponding equivalent flow area at a level not lower than the opening of the finished product branch valve and the corresponding equivalent flow area, so that the material flow containing the departed material preferentially enters the recovery branch. After the departure window ends, the process control system first maintains the recovery branch open for a delay, then gradually restores the opening of the finished product branch valve, and finally reduces the opening of the recovery branch valve or closes the recovery branch valve.

[0089] Edge computing nodes synchronously collect pressure fluctuations, turbidity changes, viscosity changes, and acoustic particle signals from the finished product branch. When the finished product branch experiences a sudden pressure jump, a sudden increase in turbidity, a sudden increase in viscosity, or an acoustic particle signal exceeding the allowable threshold for the finished product branch, the process control system determines that there is a risk of erroneous release of the finished product branch and immediately increases the opening of the recovery branch valve, decreases the opening of the finished product branch valve, or switches the finished product branch to a restricted release state.

[0090] After the branch switching is completed, the process control system generates a branch diversion record. The branch diversion record includes the start time of the departure window, the end time of the departure window, the opening degree of the recovery branch valve, the opening degree of the finished product branch valve, the opening degree of the buffer reflux valve, the flow rate of the recovery branch, the risk mark of the finished product branch and the handling action of mis-release, and serves as the basis for S430 anomaly verification and S510 wall status review.

[0091] S430 and edge computing nodes read the self-peeling execution record output by S410 and the branch diversion record output by S420, and continue to collect mainstream process data and wall response data during online self-peeling to determine whether the self-peeling action is completed according to the executable control command and whether the detached material is blocked outside the finished product branch.

[0092] Edge computing nodes focus on verifying whether the target pipe section has risks of sudden pressure spikes, sudden viscosity increases, insufficient capacity of the recovery branch, sudden load increase of the colloid mill, or accidental release of the finished product branch. When the capacity of the recovery branch is insufficient, the process control system prioritizes reducing the self-peeling strength and suspending interface flushing. When the pressure of the target pipe section spikes, it prioritizes stopping short-term reverse reflux and reducing the intensity of low-amplitude pulse flow. When there is a risk of accidental release of the finished product branch, it prioritizes increasing the opening of the recovery branch valve and decreasing the opening of the finished product branch valve. When the viscosity of the material suddenly increases, it prioritizes extending the opening time of the recovery branch and reducing the temperature step amplitude. When the load of the colloid mill suddenly increases, it prioritizes reducing the reflux rate and maintaining sampling verification of the target pipe section.

[0093] Edge computing nodes calculate the self-peeling completion degree based on the wall active layer state values ​​before and after self-peeling:

[0094] Where K represents the degree of self-peeling completion. This represents the state value of the active layer on the wall of the target pipe segment after online self-stripping. This represents the state value of the active layer on the wall of the target pipe segment before online self-peeling is performed.

[0095] The self-peeling completion rate in S430 is the preliminary completion rate obtained based on real-time data acquisition during the online execution phase. The wall state recovery range in S510 is the verification recovery range obtained after the mainstream process data has stabilized during the verification acquisition phase. These two correspond to the online preliminary judgment and the verification recovery judgment, respectively. When the self-peeling completion rate reaches the preset completion rate threshold and the finished product branch risk is marked as risk-free, the process control system preliminarily determines that the current online self-peeling action has achieved the expected reduction effect and writes this preliminary judgment result into the online self-peeling process control record for reconfirmation after the S510 verification acquisition.

[0096] The edge computing node writes the target pipe segment number, executable control command number, self-peeling execution record number, branch diversion record number, execution safety percentage, departure window identification value, recovery branch flow rate, finished product branch risk marker, pressure jump marker, viscosity surge marker, insufficient recovery capacity marker, self-peeling strength correction amount, extended recovery branch opening time, and self-peeling completion rate into the online self-peeling process control record. The online self-peeling process control record serves as input for the S510 to recalculate the wall active layer state record and is used to determine whether the target pipe segment needs further correction of self-peeling parameters.

[0097] S5 specifically includes the following sub-steps: S510, the edge computing node reads the self-stripping execution record output by S410, the branch shunting record output by S420, and the online self-stripping process control record output by S430, and starts the verification collection after the online self-stripping action is completed.

[0098] The verification data collection is not executed immediately, but only after the target pipe section pressure, material flow rate, discharge viscosity, colloid mill load, and finished product branch valve opening have all returned to the allowable disturbance range defined by the target pipe section data collection configuration record. For data still in the pressure drop, reflux switching, delayed opening of the recovery branch, or restricted release stage of the finished product branch, the edge computing node is marked as transitional data. Transitional data only retains the data source number and does not participate in the production recovery determination. The verification data collection objects include the mainstream process data package, wall response data package, and wall response deviation of the target pipe section; the verification basis includes self-stripping execution records, branch diversion records, and online self-stripping process control records.

[0099] The edge computing nodes normalize the pressure recovery time, temperature response hysteresis, acoustic echo attenuation amplitude, and flow slip deviation obtained from the verification and acquisition according to the same wall feature extraction rules in S210 and S220, and recalculate the wall active layer state value, wall adhesion layer thickness characterization value, wall adhesion strength characterization value, wall adsorption activity characterization value, and wall detachment risk characterization value, thereby generating an updated wall active layer state record.

[0100] The updated wall active layer status record includes at least the target pipe section number, the verification and collection cycle number, the wall active layer status value, the wall adhesion layer thickness characterization value, the wall adhesion strength characterization value, the wall adsorption activity characterization value, the wall detachment risk characterization value, the self-peeling completion degree, the finished branch risk mark, and the transition data deletion mark.

[0101] Edge computing nodes further calculate the extent of wall state recovery:

[0102] Where B represents the wall surface recovery amplitude. This represents the active layer status value of the target pipe section wall before online self-peeling execution. The value of the active layer on the wall is obtained after the online self-peeling action is performed. When the recovery range of the wall state reaches the preset recovery range threshold, the edge computing node determines that the online self-peeling action has effectively reduced the wall adhesion layer.

[0103] The updated wall surface active layer status record and the wall surface status recovery range are used together as inputs for the S520 production recovery judgment.

[0104] S520: The process control system reads the updated wall active layer status record, wall status recovery amplitude, branch diversion record, and online self-peeling process control record output by S510, and compares the updated wall adhesion layer thickness characterization value, wall adhesion strength characterization value, wall adsorption activity characterization value, and wall detachment risk characterization value with the preset stability range, respectively.

[0105] The preset stability range is a low-risk range used to determine whether continuous discharge of finished product branches is allowed after the online self-peeling action ends. Its boundary value is stricter than the preset interference threshold in S310. The preset interference threshold is used to trigger wall self-peeling control, and the preset stability range is used to confirm that the interference of the wall adhesion layer on the quality of finished product has been eliminated. The two should not be used interchangeably.

[0106] The process control system generates a production recovery record only when all four characterization values ​​are within the preset stable range, the mainstream process data package is in a mainstream stable state, the finished product branch risk is marked as risk-free, the delayed opening time of the recovery branch meets the preset requirements, and the wall state recovery amplitude reaches the preset recovery amplitude threshold. The production recovery record includes the target pipe section number, the verification and acquisition cycle number, the recovery judgment result, the finished product branch recovery time, the recovery branch closure time, and the pre-recovery state confirmation field. The process control system gradually restores the continuous discharge of the finished product branch and closes or reduces the opening degree of the recovery branch valve based on the production recovery record.

[0107] When any characteristic value fails to recover to the preset stable range, or when the risk marker of the finished product branch remains at risk, the process control system does not generate a production recovery record, but instead generates a secondary correction record. The secondary correction record includes at least the target pipe section number, the name of the unrecovered indicator, the current value of the unrecovered indicator, the corresponding preset stable range, the extent of unrecovered state, the self-stripping completion rate, the execution safety percentage, the exit window identification value, the finished product branch risk marker, the suggested correction action, the correction amount for low-amplitude pulse flow intensity, the correction amount for short-term reverse backflow duration, the correction amount for temperature step amplitude, the correction amount for interface flushing medium volume, and the correction amount for the delayed opening of the recovery branch.

[0108] The secondary correction amount is determined according to the following formula:

[0109] in, This is the second correction value for the qth type of self-peeling control parameters, which include low-amplitude pulse current intensity, short-time reverse backflow duration, temperature step amplitude, and interface flushing medium volume. For the qth type of self-peeling control parameter, the correction coefficient is... The current value of the unrecovered indicator. This is the upper limit of the preset stable range corresponding to the unrecovered indicator.

[0110] The secondary correction record serves as the basis for S530 to update the control boundary and for the next round of S320 to generate the wall self-peeling control command. The corresponding correction action still needs to be re-verified on-site in the next round of S330.

[0111] The S530 edge computing node reads stable state determination records, production recovery records, secondary correction records, anomaly handling records, online self-peeling process control records, and updated wall active layer status records, and performs closed-loop updates on the preset interference threshold, preset stable range, target pipe section acquisition configuration records, and safety verification boundaries for the target pipe section.

[0112] During updates, only verified data that has passed consistency checks is used; data awaiting retesting, transitional data, data with missing fields, and data marked as risky by finished branch lines are not used. For data with incomplete sources or inconsistent with the target pipe section number, the edge computing node directly marks it as unusable for updates and does not participate in control boundary correction. Control boundaries include preset interference thresholds, preset stability range upper limits, acquisition cycle, pressure pulse excitation upper limits, thermal pulse excitation upper limits, allowable disturbance range, and on-site safety verification boundaries.

[0113] To prevent a single anomaly from causing a sudden drift in the control boundary, edge computing nodes use a smooth update method to correct the control boundary:

[0114] in, For the updated control boundary values, The control boundary values ​​before the update. These are the recommended control boundary values ​​obtained based on the data from this review. The update coefficient is greater than 0 and less than 1.

[0115] To prevent large oscillations in the control boundary caused by a single abnormal measurement, the updated coefficients... The value is dynamically adjusted based on the signal-to-noise ratio of the collected data. Its preferred conventional value range is defined between 0.05 and 0.20, thereby ensuring that the control boundary can both track the long-term trend of equipment aging and have robustness against short-term disturbances.

[0116] The process control system will use the updated target pipe section acquisition configuration record for the S120 mainstream process data acquisition and S130 wall response data acquisition in the next control cycle, use the updated preset interference threshold for the S310 interference status determination in the next control cycle, use the updated preset stability range for the S520 production recovery determination in the next control cycle, and use the updated safety verification boundary for the S330 field safety verification in the next control cycle. For target pipe sections that generate secondary correction records, the process control system will send the secondary correction records to the next S320 and re-perform the field safety verification in the next S330.

[0117] Thus, the wall adhesion layer identification, interference status determination, online self-peeling, branch diversion, verification and recovery, and control boundary update form a continuous closed loop.

[0118] All the above formulas are performed using dimensionless numerical calculations; the relevant formulas are based on empirical models that approximate the real situation, obtained through extensive data collection and software simulation fitting. The preset parameters and thresholds involved in the formulas can be conventionally set and adjusted by those skilled in the art according to the physical constraints of the actual application scenario.

[0119] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0120] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0121] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-variable collaborative control method for the continuous production of environmentally friendly asphalt waterproof coatings, characterized in that, Includes the following steps: S1. The process control system reads equipment operation records, pipeline layout records and abnormal records, identifies the production channel that is prone to forming a wall adhesion layer as the target pipe section, and collects the mainstream process data and wall response data of the target pipe section by the edge computing node, generating mainstream process data package, wall response data package and wall response deviation. S2. The edge computing node confirms that the target pipe section is in a stable state according to the mainstream process data packet, and extracts the pressure recovery time, temperature response hysteresis, acoustic attenuation amplitude and flow slip deviation according to the wall response data packet to generate an effective state record. S3. The edge computing node compares the effective status record with the preset interference threshold to generate an interference status judgment record. The process control system determines the interference type, generates and verifies the wall self-peeling control command, and outputs an executable control command. S4. The process control system executes low-amplitude pulse flow, temperature step, short-time reverse reflux and interface flushing according to executable control commands, controls the recovery branch to receive the separated material within the separation window, and generates an online self-peeling process control record.

2. The multivariate collaborative control method for continuous production of environmentally friendly asphalt waterproof coating according to claim 1, characterized in that, Also includes: S5. The edge computing node generates an updated wall active layer status record based on the online self-peeling process control record. The process control system restores the finished product branch discharge or generates a secondary correction record and updates the control boundary.

3. The multi-variable collaborative control method for continuous production of environmentally friendly asphalt waterproof coating according to claim 1, characterized in that, S1 specifically includes: The process control system reads equipment operation records, pipeline layout records, historical cleaning records, and historical finished product anomaly records. It identifies continuous production channels with sudden changes in flow direction, sudden changes in shear state, concentrated temperature gradients, or high frequency of historical anomalies as target pipe sections and generates target pipe section acquisition and configuration records. Edge computing nodes collect mainstream process data based on the target pipeline segment collection configuration record. After time alignment, field verification, and deletion of invalid data, mainstream process data packets are generated.

4. The multivariate collaborative control method for continuous production of environmentally friendly asphalt waterproof coating according to claim 3, characterized in that, Also includes: The process control system applies pressure pulse excitation, thermal pulse excitation and acoustic echo detection to the target pipe section under the mainstream steady state. The edge computing node collects the corresponding curves and generates wall response data packets and wall response deviation.

5. The multivariate collaborative control method for continuous production of environmentally friendly asphalt waterproof coating according to claim 1, characterized in that, S2 specifically includes: Edge computing nodes read mainstream process data packets, wall response data packets, and wall response deviations. When the target pipe section is in a stable mainstream state, they extract pressure recovery time, temperature response hysteresis, acoustic echo attenuation amplitude, and flow slip deviation, and generate a set of wall features. Edge computing nodes identify the formation state of the wall adhesion layer based on the wall feature set, calculate the wall active layer state value, and generate a wall active layer state record that includes the wall adhesion layer thickness, wall adhesion strength, wall adsorption activity, and wall detachment risk. The process control system performs consistency verification on the status records of the active layer on the wall and marks the records that pass the verification as valid status records.

6. The multivariable collaborative control method for continuous production of environmentally friendly asphalt waterproof coating according to claim 1, characterized in that, S3 specifically includes: Edge computing nodes read valid state records and compare the wall adhesion layer thickness, wall adhesion strength, wall adsorption activity, and wall detachment risk with the corresponding preset interference thresholds to generate stable state determination records or interference state determination records. The process control system determines the type of interference—thickness-dominant, intensity-dominant, adsorption-dominant, shedding-dominant, or a combination thereof—based on the interference status judgment record, and generates wall self-peeling control commands.

7. The multivariable collaborative control method for continuous production of environmentally friendly asphalt waterproof coating according to claim 6, characterized in that, It also includes: edge computing nodes perform on-site safety verification of the wall self-peeling control command. If the verification is successful, the executable control command is output. If the verification fails, the parameters are downgraded or a self-peeling prohibition mark is generated.

8. The multivariate collaborative control method for continuous production of environmentally friendly asphalt waterproof coating according to claim 1, characterized in that, S4 specifically includes: The process control system reads executable control commands and sequentially executes low-amplitude pulse flow, temperature step, short-time reverse reflux and interface flushing on the target pipe section to control the detachment of the wall adhesion layer and generate a self-peeling execution record. The edge computing node identifies the detachment window based on the target pipe section pressure, acoustic echo signal, and recovery branch flow. The process control system controls the recovery branch to prioritize receiving the detached material within the detachment window and generates a branch diversion record. Edge computing nodes verify pressure, viscosity, recovery capacity, and finished product branch risk based on self-peeling execution records and branch diversion records, and generate online self-peeling process control records.

9. The multivariate collaborative control method for continuous production of environmentally friendly asphalt waterproof coating according to claim 2, characterized in that, S5 specifically includes: Edge computing nodes read self-stripping execution records, branch shunting records, and online self-stripping process control records. After the mainstream process data is restored to the allowable disturbance range, they perform verification and collection to generate updated wall active layer status records and wall status recovery amplitude. The process control system compares the updated wall adhesion layer thickness, wall adhesion strength, wall adsorption activity, and wall detachment risk with the preset stability range, and generates a production recovery record or a secondary correction record.

10. The multivariable collaborative control method for continuous production of environmentally friendly asphalt waterproof coating according to claim 9, characterized in that, It also includes: edge computing nodes updating preset interference thresholds, preset stability ranges, target pipe segment acquisition configuration records, and security verification boundaries based on stable state determination records, production recovery records, secondary correction records, and anomaly handling records.

Images