A pressure-sensitive self-repairing asphalt cold patch mixing intelligent control system and method

Abstract

The present application belongs to the technical field of intelligent manufacturing control, and particularly relates to a pressure-sensitive self-repairing asphalt cold patch material mixing intelligent control system and method, which comprises the following steps: collecting torque time series data of a mixing motor; constructing a dynamic feature vector of the torque time series data at each time point based on the torque time series data, which is used to represent the asymmetry and transient peak degree of the torque time series data fluctuation; performing unsupervised analysis on a feature space formed by the dynamic feature vector by using a local anomaly factor algorithm to calculate a local anomaly factor corresponding to each time point; and executing a closed-loop control strategy according to the local anomaly factor and the feature representing asymmetry in the dynamic feature vector to realize intelligent control of the mixing process. The present application can accurately distinguish different mixing conditions, can timely stop production to protect product activity, and can reasonably judge mixing uniformity through the asymmetry feature to improve production efficiency.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing control technology. More specifically, this invention relates to a pressure-sensitive, self-healing asphalt cold patch mixing intelligent control system and method. Background Technology

[0002] Pressure-sensitive self-healing asphalt cold patch is a new type of high-performance road maintenance material. Unlike traditional cold patch materials, its core components mainly include pressure-sensitive binders, aggregates coated with self-healing gel coats, and active fillers. The self-healing gel coat encapsulates the repair agent; during actual road use, the pressure of vehicle loads causes the gel coat to rupture, releasing the internal repair agent, thereby achieving dynamic self-healing of road damage and pressure-enhanced bonding.

[0003] Currently, a crucial process control point in the mixing and production stage of cold asphalt patching is protecting the integrity of the gel coat microcapsules to prevent them from prematurely breaking under the shearing force of the mixer. Premature damage to the gel coat leads to the premature release and reaction of the repair agent, resulting in the final product losing its self-healing function. Existing technologies generally use monitoring the torque or power of the mixing motor to indirectly determine the uniformity of the mixture. Typically, mixing is considered complete when the mixture reaches a certain viscosity and the torque value stabilizes.

[0004] However, the mixing process of pressure-sensitive cold-mixed materials is complex and variable. The motor torque curve may include three distinct operating conditions: normal mixing, where the torque value fluctuates symmetrically around the mean; component agglomeration, where aggregate or active filler clumping causes a periodic increase in mixing resistance, resulting in sustained high-amplitude torque oscillations; and gel coat rupture, where gel coat microcapsules break under shear force, releasing internal fluid and causing a sudden, sharp drop in the mixture viscosity. Existing technologies generate high-intensity abnormal signals for both component agglomeration and gel coat rupture because both exhibit high volatility, making it difficult for the control system to distinguish between the two anomalies. This confusion leads to erroneous control decisions, potentially causing the entire batch of product to be scrapped due to premature gel coat failure, resulting in significant economic losses. Summary of the Invention

[0005] To address the technical problem that existing technologies struggle to distinguish between component agglomeration and gel coat breakage, the present invention provides solutions in several aspects.

[0006] In a first aspect, the present invention provides a pressure-sensitive self-healing asphalt cold patch mixing intelligent control method, comprising: acquiring torque time-series data of a mixing motor; constructing a dynamic feature vector of the torque time-series data at each time point based on the torque time-series data, wherein the dynamic feature vector is used to characterize the asymmetry and transient kurtosis of the torque time-series data fluctuation; performing unsupervised analysis on the feature space formed by the dynamic feature vector using a local anomaly factor algorithm to calculate the local anomaly factor corresponding to each time point; and performing closed-loop control based on the local anomaly factor and the dynamic feature vector to achieve intelligent control of the mixing process.

[0007] This invention constructs asymmetric fluctuation and transient peak dynamic feature vectors, combined with the LOF algorithm, to accurately distinguish between component agglomeration and gel coat breakage. Therefore, it can instantly detect gel coat breakage and immediately stop the machine, protecting the core self-healing activity of the product. At the same time, it can reasonably determine the mixing endpoint and achieve intelligent closed-loop control.

[0008] Preferably, the dynamic feature vector is a two-dimensional feature vector, which includes: an asymmetric fluctuation index for evaluating the directionality of the torque time series data fluctuation; and a transient spike index for evaluating the persistence of the torque time series data fluctuation.

[0009] Preferably, the asymmetric fluctuation index satisfies the following relationship: ;in, Let be the asymmetric fluctuation index at time t. For a point in time, This is the sum of squares of all fluctuations above the local average torque within the sliding window. This is the sum of squares of all fluctuations below the local average torque within the sliding window. To prevent extremely small positive numbers with a denominator of zero.

[0010] Preferably, the transient spike index satisfies the following relationship: ;in, The transient peak exponent at time t, For a point in time, For the current transient fluctuation, The length of the background window. For time indexing, To prevent extremely small positive numbers with a denominator of zero.

[0011] Preferably, the transient fluctuations satisfy the following relationship: ;in, for The torque value at that moment. and These are the torque values ​​at time t+1 and t-1, respectively.

[0012] By employing second-order difference to capture the abrupt changes in a signal, it is possible to detect sharp changes in the signal curve with extreme sensitivity. Compared to first-order difference or fluctuation amplitude, second-order difference can better amplify such instantaneous changes, improving the instantaneity and accuracy of detection.

[0013] Preferably, the unsupervised analysis of the feature space formed by the dynamic feature vectors using the local anomaly factor algorithm includes: taking all dynamic feature vectors as input datasets; standardizing the input datasets; and calculating the local anomaly factor corresponding to each time point based on the standardized datasets.

[0014] Preferably, the unsupervised analysis of the feature space formed by the dynamic feature vector using the local anomaly factor algorithm further includes: identifying two high-density clusters in the feature space that represent normal mixing and component agglomeration conditions, respectively; and identifying sparse anomalies in the feature space that represent gel coat cracking conditions, wherein the asymmetric feature of the sparse anomalies is close to -1 and the transient kurtosis feature is much greater than 1.

[0015] By performing density clustering analysis in a two-dimensional feature space, the fundamental differences in data distribution under different working conditions are utilized. Gelcoat breakage, as an occasional event, will inevitably result in sparse outliers, while agglomeration and normal mixing, as continuous working conditions, will inevitably form high-density clusters. This density-based differentiation method can effectively distinguish between two highly volatile signals.

[0016] Preferably, the closed-loop control includes: when the local abnormal factor at time t is greater than the gelcoat breakage alarm threshold, a gelcoat breakage is diagnosed, and mixing is immediately stopped; when the local abnormal factor is not greater than the gelcoat breakage alarm threshold and the asymmetric feature is greater than the component agglomeration diagnosis threshold, component agglomeration is diagnosed, and the mixing time is extended; when the mixing time is greater than the standard mixing time and the local abnormal factor is not greater than the gelcoat breakage alarm threshold, and the corresponding asymmetric feature is not greater than the mixing uniformity judgment threshold, mixing is diagnosed as complete, and mixing is stopped.

[0017] A clear priority system is established through hierarchical control, with the highest priority given to the diagnosis of gel coat breakage. This ensures that the machine can be stopped immediately upon detection of this catastrophic anomaly, maximizing the protection of product activity. Simultaneously, the diagnosis of agglomeration and completion enables dynamic intelligent adjustment of mixing time, avoiding ineffective or insufficient mixing and improving production efficiency.

[0018] Preferably, the specific process of collecting the torque time series data of the mixing motor is as follows: torque time series data is collected in real time at a fixed sampling period by means of a torque sensor installed on the drive motor of the mixer, or by means of feedback data from the frequency converter of the drive motor.

[0019] High-frequency sampling ensures the precision of torque data, providing the necessary data foundation for subsequent calculations of second-order differences and the capture of instantaneous changes, thus guaranteeing the sensitivity of the diagnostic algorithm.

[0020] Secondly, the present invention provides a pressure-sensitive self-healing asphalt cold patch mixing intelligent control system, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned pressure-sensitive self-healing asphalt cold patch mixing intelligent control method is implemented.

[0021] By adopting the above technical solution, a computer program is generated from the above-mentioned intelligent control method for mixing pressure-sensitive self-healing asphalt cold patch material, and stored in the memory so that it can be loaded and executed by the processor. In this way, a terminal device can be made based on the memory and the processor for convenient use.

[0022] This invention overcomes the shortcomings of traditional torque monitoring in distinguishing different working conditions. By constructing an asymmetric fluctuation index that characterizes the directionality of fluctuations and a transient spike index that characterizes the duration of fluctuations, the three states of normal mixing, component agglomeration, and gel coat rupture are clearly separated in the feature space, thus achieving accurate differentiation of mixing conditions.

[0023] Furthermore, this invention utilizes the local anomaly factor algorithm, which is highly sensitive to sparse anomaly points in the feature space representing gelcoat breakage. It can trigger the highest priority shutdown the instant the breakage occurs, effectively protecting the product's self-healing core function. Attached Figure Description

[0024] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent upon reading the following detailed description with reference to the accompanying drawings. In the drawings, several embodiments of the invention are illustrated by way of example and not limitation, and like or corresponding reference numerals denote like or corresponding parts, wherein:

[0025] Figure 1 This is a schematic flowchart illustrating an intelligent control method for mixing pressure-sensitive self-healing asphalt cold patch material according to the present invention;

[0026] Figure 2 This is a schematic diagram showing a comparison of the diagnostic effects of the present invention and existing technologies on mixing conditions. Detailed Implementation

[0027] 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, not all, of the embodiments of the present invention. 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.

[0028] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0029] This invention discloses an intelligent control method for mixing pressure-sensitive self-healing asphalt cold patch material, referring to... Figure 1 This includes steps S1-S4:

[0030] S1. Collect the torque timing data of the mixing motor.

[0031] In an optional embodiment, a high-frequency torque sensor can be installed on the drive motor of the asphalt cold patch mixing plant, or the data feedback function of the motor frequency converter itself can be used with a fixed sampling period, for example, such as... Real-time acquisition of motor torque time series data during the mixing process.

[0032] Specifically, since gelcoat breakage is an instantaneous event, the decrease in torque may occur within hundreds of milliseconds. If the sampling frequency is too low, for example, once per second, this abrupt change cannot be captured. Therefore, high-frequency sampling is required to ensure the accuracy of the raw data.

[0033] In this way, by acquiring torque data in real time at high frequency, a high-resolution raw data foundation is provided for subsequent refined dynamic feature analysis, ensuring that instantaneous events such as gelcoat rupture can be captured.

[0034] S2. Construct dynamic feature vectors for torque time series data at each time point based on torque time series data. The dynamic feature vectors are used to characterize the asymmetry and transient kurtosis of torque time series data fluctuations.

[0035] In an optional embodiment, a two-dimensional dynamic feature vector of the torque time series data at each time point can be constructed based on the obtained torque time series data. This vector consists of an asymmetric fluctuation index and a transient spike index, which are used to characterize the asymmetry and transient spike degree of the torque time series data fluctuation.

[0036] Specifically, the asymmetric fluctuation index is used to assess the directionality of mixing torque fluctuations to distinguish fluctuations from those with different physical causes. Specifically, when component agglomeration leads to solid agglomeration and increased resistance, the torque manifests as a continuously upward positive pulse, at which point the upward fluctuation energy... Much greater than the downward energy This causes the index to continuously approach +1; gel coat cracking is due to a decrease in viscosity, and the torque manifests as a momentary downward negative pulse, at which point the downward energy... Much greater than the upward energy This causes the index to momentarily approach -1; normal mixing fluctuations are symmetrical around the mean. This causes the index to continue approaching 0.

[0037] In this optional embodiment, the local average torque is calculated within a sliding window centered at time point t and with a width of 2W+1. Where W is the half width of the sliding window, Indicates the first The torque value at each sampling time, for example, is 30.

[0038] Furthermore, the uplink energy within each window is calculated separately. and downward energy Finally, the asymmetric volatility index is constructed, which satisfies the following relationship:

[0039]

[0040] in, Let be the asymmetric fluctuation index at time t. For a point in time, This is the sum of squares of all fluctuations above the local average torque within the sliding window. This is the sum of squares of all fluctuations below the local average torque within the sliding window. To prevent extremely small positive numbers with a denominator of zero, for example, the value is taken as... .

[0041] In this optional embodiment, the transient spike index is used to assess the persistence of fluctuations. Specifically, normal mixing or component agglomeration are persistent fluctuations, and the current fluctuation amplitude is similar to its historical background fluctuation amplitude, causing the index to approach 1; while gelcoat breakage is a sudden event, and the current fluctuation amplitude is significantly higher than its stable historical background, causing the index to be much greater than 1 instantaneously.

[0042] Specifically, the absolute value of the second-order difference of the torque signal can be used to assess the degree of abrupt change, thereby calculating transient fluctuations. Transient fluctuations satisfy the following relationship:

[0043]

[0044] in, for The torque value at that moment. and These are the torque values ​​at times t+1 and t-1, respectively. Next, the transient peak exponent can be calculated based on the obtained transient fluctuations. The transient peak exponent satisfies the following relationship:

[0045]

[0046] in, The transient peak exponent at time t, This represents the length of the background window; for example, a value of 50. Using time indexing, the transient spike index represents the current transient fluctuation and its immediate neighbors. The ratio of the historical background fluctuation mean within the window.

[0047] Thus, by constructing a two-dimensional dynamic feature vector composed of an asymmetric fluctuation index and a transient spike index, the implicit operating condition information in the one-dimensional time series signal is successfully made explicit, providing a separable feature space for subsequent differentiation between two anomalies with completely different causes: aggregation and rupture.

[0048] S3. Use the local anomaly factor algorithm to perform unsupervised analysis on the feature space formed by dynamic feature vectors and calculate the local anomaly factor corresponding to each time point.

[0049] In an optional embodiment, since normal mixing and component agglomeration are continuous processes, they will form two high-density clusters in the feature space, while gelcoat breakage is an occasional event and is a sparse outlier in the feature space. The Local Anomaly Factor (LOF) algorithm is used to evaluate the degree of sparsity anomaly of a data point relative to its neighborhood. Therefore, the LOF algorithm can be used to perform unsupervised analysis on the feature space composed of dynamic feature vectors to calculate the local anomaly factor corresponding to each time point.

[0050] Specifically, all dynamic feature vectors are used as the input dataset for the local anomaly factor algorithm. After standardizing the input dataset, the local anomaly factor of the vector at each time point is calculated. The local anomaly factor satisfies the following relationship:

[0051]

[0052] Where k is the number of neighborhood points, for example, a value of 50. For the k-neighborhood, is the locally reachable density, and o is the index of the data point in the neighborhood. When t is in a normal or clustered state, the data point is in a dense region, and its density is similar to the average density of its neighborhood, resulting in its local anomaly factor being close to 1; when t is in a gel coat breakage state, the data point is an isolated point, and its density is much lower than the average density of its neighborhood points, resulting in its local anomaly factor being much greater than 1.

[0053] Thus, by introducing the local anomaly factor algorithm, the prior knowledge that gelcoat breakage must be a sparse outlier in the feature space is utilized, enabling the diagnostic model to automatically ignore the continuous fluctuations of high-density clusters such as component aggregation and to generate an extremely sensitive response to the instantaneous burst signal of gelcoat breakage.

[0054] S4. Closed-loop control is performed based on local anomaly factors and dynamic eigenvectors to achieve intelligent control of the mixing process.

[0055] In an optional embodiment, a gelcoat breakage alarm threshold can be preset, for example, a value of 3, to determine the degree of mutation of local abnormal factors; a component aggregation diagnosis threshold, for example, a value of 0.6, to determine the degree of positive deviation of asymmetric fluctuations; a mixing uniformity judgment threshold, for example, a value of 0.2, to determine whether asymmetric fluctuations return to symmetry; and a standard mixing time, for example, a value of 300 seconds.

[0056] Furthermore, the system monitors local anomalies in real time. If the value of such an anomaly exceeds the preset gelcoat breakage alarm threshold, the system diagnoses it as a gelcoat breakage anomaly. At this point, the system immediately executes the highest priority action, stopping the mixing process, classifying the batch as a high-risk batch, and sending the highest-level alarm to the central control room, prompting manual inspection to check if the mixing rate or temperature exceeds the limits.

[0057] In this optional embodiment, if the local abnormal factor does not exceed the gel coat breakage alarm threshold, the system then checks the asymmetric fluctuation index. If the index is greater than the preset component agglomeration diagnosis threshold, the system diagnoses that component agglomeration has occurred in the mixture. At this time, the system executes a medium-priority action, appropriately extending the mixing time, for example, by 30 seconds, and issues a mixing unevenness alarm.

[0058] In this optional embodiment, if the mixing time has exceeded the standard mixing time, and the local abnormal factors have not exceeded the gel coat breakage alarm threshold, and the asymmetric fluctuation index is not greater than the mixing uniformity judgment threshold, the system diagnoses that mixing is complete. At this time, the system performs routine operations, stops mixing, determines it as a qualified batch, and proceeds to the discharge process.

[0059] like Figure 2The figures shown are schematic comparisons of the diagnostic effects of the present invention and existing technologies in mixing conditions. Figure (a) is a schematic diagram of anomaly diagnosis using existing technology based on moving range, while Figure (b) is a schematic diagram of anomaly diagnosis based on dynamic feature vectors and the LOF algorithm in an embodiment of the present invention. As can be seen, in Figure (a), the original torque fluctuates drastically during both the component agglomeration and gelcoat rupture stages, causing the anomaly score of the existing technology to continuously or instantaneously exceed the alarm threshold in both stages. The control system cannot distinguish between these two conditions requiring opposite control commands. In Figure (b), under the same original torque, the local anomaly factor score of the present invention is successfully suppressed to a low level during the component agglomeration stage, without triggering an alarm. However, during the gelcoat rupture stage, the local anomaly factor score generates a spike exceeding the alarm threshold. This indicates that the present invention successfully ignores the continuous fluctuation interference of component agglomeration and accurately captures the sudden event of gelcoat rupture, achieving effective differentiation between the two conditions.

[0060] Thus, by designing a priority-based multi-branch control strategy, the highest priority response to gel coat breakage is ensured. At the same time, the asymmetric fluctuation index is used to distinguish between agglomeration and completion states, achieving safe, efficient, and intelligent closed-loop control of the mixing process.

[0061] This invention also discloses a pressure-sensitive self-healing asphalt cold patch mixing intelligent control system, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, a pressure-sensitive self-healing asphalt cold patch mixing intelligent control method according to the present invention is implemented.

[0062] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0063] In the description of this specification, "multiple" or "several" means at least two, such as two, three or more, unless otherwise expressly and specifically defined.

[0064] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.

Claims

1. A pressure-sensitive self-healing asphalt cold patch mixing intelligent control method, characterized in that, include: Collect torque timing data of the mixing motor; Based on the torque time series data, a dynamic feature vector of the torque time series data is constructed at each time point. The dynamic feature vector is used to characterize the asymmetry and transient kurtosis of the torque time series data fluctuation. The local anomaly factor algorithm is used to perform unsupervised analysis on the feature space formed by the dynamic feature vectors, including: identifying two high-density clusters in the feature space that represent normal mixing and component agglomeration conditions respectively; and identifying sparse anomaly points in the feature space that represent gel coat cracking conditions, wherein the asymmetric features of the sparse anomaly points are close to -1 and the transient kurtosis features are much greater than 1. Calculate the local anomaly factor corresponding to each time point; Closed-loop control is performed based on the local anomaly factor and the dynamic feature vector to achieve intelligent control of the mixing process. The closed-loop control includes: when the local anomaly factor at time t is greater than the gelcoat breakage alarm threshold, gelcoat breakage is diagnosed and mixing is stopped immediately; when the local anomaly factor is not greater than the gelcoat breakage alarm threshold and the asymmetric feature is greater than the component agglomeration diagnosis threshold, component agglomeration is diagnosed and mixing time is extended; when the mixing time is greater than the standard mixing time and the local anomaly factor is not greater than the gelcoat breakage alarm threshold, and the corresponding asymmetric feature is not greater than the mixing uniformity judgment threshold, mixing is diagnosed as complete and mixing is stopped.

2. The intelligent control method for mixing pressure-sensitive self-healing asphalt cold patch material according to claim 1, characterized in that, The dynamic feature vector is a two-dimensional feature vector, which includes: an asymmetric fluctuation index, used to evaluate the directionality of the torque time series data fluctuation; and a transient spike index, used to evaluate the persistence of the torque time series data fluctuation.

3. The intelligent control method for mixing pressure-sensitive self-healing asphalt cold patch material according to claim 2, characterized in that, The asymmetric volatility index satisfies the following relationship: in, Let be the asymmetric fluctuation index at time t. For a point in time, This is the sum of squares of all fluctuations above the local average torque within the sliding window. This is the sum of squares of all fluctuations below the local average torque within the sliding window. To prevent extremely small positive numbers with a denominator of zero.

4. The intelligent control method for mixing pressure-sensitive self-healing asphalt cold patch material according to claim 2, characterized in that, The transient spike exponent satisfies the following relationship: in, The transient peak exponent at time t, For a point in time, For the current transient fluctuation, The length of the background window. For time indexing, To prevent extremely small positive numbers with a denominator of zero.

5. The intelligent control method for mixing pressure-sensitive self-healing asphalt cold patch material according to claim 4, characterized in that, The transient fluctuations satisfy the following relationship: in, for The torque value at that moment. and These are the torque values ​​at time t+1 and t-1, respectively.

6. The intelligent control method for mixing pressure-sensitive self-healing asphalt cold patch material according to claim 1, characterized in that, The unsupervised analysis of the feature space formed by the dynamic feature vectors using the local anomaly factor algorithm includes: Use all dynamic feature vectors as the input dataset; The input dataset is standardized. The local anomaly factor for each time point is calculated based on the standardized dataset.

7. The intelligent control method for mixing pressure-sensitive self-healing asphalt cold patch material according to claim 1, characterized in that, The specific process for collecting the torque time series data of the mixing motor is as follows: torque time series data is collected in real time at a fixed sampling period by means of a torque sensor installed on the drive motor of the mixing machine, or by means of feedback data from the frequency converter of the drive motor.

8. A pressure-sensitive, self-healing asphalt cold patch mixing intelligent control system, characterized in that, include: The processor and memory, wherein the memory stores computer program instructions, which, when executed by the processor, implement a pressure-sensitive self-healing asphalt cold patch mixing intelligent control method according to any one of claims 1-7.

Images