An end-cloud coordination municipal road old material regeneration hot-mix cold-paving integrated construction method

Abstract

This invention relates to the field of municipal road maintenance and road material recycling construction technology, and particularly to an end-to-cloud collaborative integrated hot-mix and cold-laying construction method for municipal road material recycling. The end-side control unit generates a batch identifier for recycled materials and uploads it to the cloud-side control platform. The cloud-side control platform constructs a prediction model based on the batch characterization data. The cloud-side control platform sends the parameter boundary set, constraint proof data, parameter version identifier, and validity conditions to the end-side control unit. The end-side control unit controls the construction equipment to complete milling, heating and dispersing, hot mixing, cold paving, and compaction, collecting construction measurements to form a residual vector. When the residual vector meets the deviation judgment condition, the end-side control unit sends the residual vector to the cloud-side control platform. The end-side control unit updates the parameter boundary set according to the differential parameter package and rolls back to the previous parameter version identifier when the rollback condition is met. This invention achieves dynamic generation and online adjustment of construction parameters to meet workability and compaction quality constraints.

Description

Technical Field

[0001] This invention relates to the field of municipal road maintenance and road material recycling construction technology, and in particular to an integrated hot-mix and cold-laying construction method for municipal road material recycling using a cloud-based collaborative approach. Background Technology

[0002] During the maintenance and renovation of municipal roads, a large amount of old asphalt pavement material (reclaimed material) is generated. In order to improve the resource utilization rate of reclaimed material, existing recycling construction operations usually include old pavement milling, old material treatment (such as heating, dispersing and mixing recycling), and paving and compaction. Gradually, solutions that integrate and automate multiple processes are emerging to improve the continuity of on-site construction and the degree of equipment coordination.

[0003] For example, the technical solution with publication number CN102121224A discloses a technical approach for an integrated cold and hot composite recycling process. This approach integrates the process chain and equipment related to recycled materials, and incorporates modules at the equipment level for automatic control of operating conditions, process temperature display and control, automatic batching and metering, and automatic display and control, to achieve automated control and integrated operation organization of the recycling process. This type of solution can achieve a certain degree of integration and automatic control of the recycling process and equipment, and is suitable for the organizational needs of continuous on-site construction operations.

[0004] However, in actual municipal road recycling construction, the source and condition of recycled materials often vary from batch to batch, and on-site environmental conditions also change. Simultaneously, the connection between multiple processes is uncertain, easily causing changes in material condition and workability windows, thus coupling with the quality constraints of the paving and compaction stages. Under these uncertain coupling conditions, existing integrated recycling solutions, primarily based on local automatic control of equipment, typically focus on parameter setting and process control for single machines or local processes. They lack an end-to-cloud collaborative mechanism that combines on-site data acquisition, process event information, and cloud-based model calculation capabilities. This makes it difficult to generate and adjust construction parameters continuously and iteratively for different batches of recycled materials and different construction sections, resulting in a strong reliance on manual experience corrections during construction and difficulty in consistently meeting preset construction quality constraints.

[0005] Therefore, the main technical problem to be solved is: under the construction control system of end-to-end cloud collaboration, how to generate and dynamically adjust the recycling construction parameters in the face of uncertain conditions such as batch differences of old materials and fluctuations in process connection (including time delay fluctuations) so that the recycling paving process can continuously meet the preset workability constraints and compaction quality constraints. Summary of the Invention

[0006] To overcome the aforementioned technical deficiencies, the present invention aims to provide an integrated hot-mix cold-paving construction method for municipal roads using recycled materials in a cloud-edge collaborative manner. The cloud side generates a parameter boundary set and proof data based on batch characterization data through constraint solving, and distributes it in a versioned manner. The terminal side selects execution values ​​within the boundaries to complete construction and triggers directional incremental recalculation, differential updates, and rollbacks on the cloud side based on residuals. This achieves dynamic generation and online adjustment of construction parameters to meet workability and compaction quality constraints.

[0007] This invention discloses an integrated hot-mix and cold-pave construction method for recycled municipal road materials using a cloud-based collaborative approach. The method is executed collaboratively by a cloud-based control platform and a terminal control unit. The terminal control unit connects to and controls the construction equipment, which includes at least milling equipment, heating and dispersing equipment, mixing equipment, paving equipment, and compaction equipment. The method includes:

[0008] S1, the end-side control unit generates batch identifiers for municipal road waste materials entering the recycling process, collects batch characterization data corresponding to the batch identifiers, and uploads the batch characterization data and construction section identifiers to the cloud-side control platform.

[0009] S2, the cloud-based control platform constructs a prediction model based on batch characterization data, and performs constraint solving under the condition that at least cold-laying workability constraints and compaction quality constraints are met, to obtain parameter boundary sets and constraint proof data, and generates parameter version identifiers;

[0010] S3, the cloud-side control platform sends the parameter boundary set, constraint proof data, parameter version identifier, and validity conditions to the end-side control unit. After the end-side control unit verifies and passes the verification, it binds and stores the old material batch identifier with the parameter version identifier.

[0011] S4, the end-side control unit selects execution values ​​within the parameter boundary set to control the construction equipment to complete milling, heating and dispersing, hot mixing, cold paving and compaction, and collects construction measurement values ​​to form a residual vector;

[0012] S5, when the residual vector meets the deviation judgment condition, the end-side control unit sends the residual vector to the cloud-side control platform; the cloud-side control platform determines the parameter subset that needs to be updated based on the residual vector and the sensitivity matrix and performs directional incremental recalculation to generate a new parameter version identifier, and distributes it in the form of differential parameter package; the end-side control unit updates the parameter boundary set according to the differential parameter package and rolls back to the previous parameter version identifier when the rollback condition is met.

[0013] Preferably, the parameter boundary set includes at least the discharge temperature boundary, the paving speed boundary, and the number of compaction passes boundary; the constraint proof data includes at least the predicted paving start temperature and the upper limit of the allowable delay; the validity condition includes at least the batch characterization data falling within the preset boundary range; and the prediction model is used to calculate the predicted paving start temperature and the upper limit of the allowable delay.

[0014] Preferably, the prediction model includes at least the components for obtaining the predicted paving start temperature. Temperature decay prediction model, where the predicted paving start temperature is... It satisfies the exponential temperature decay form.

[0015] Preferably, the predicted temperature at the start of paving satisfy:

[0016]

[0017] in, For discharge temperature, For ambient temperature, The time delay between the completion of material discharge and the start of paving. The temperature decay coefficient is It is an exponential function.

[0018] Preferably, the upper limit of allowable latency Lower limit temperature of cold-laid workability The reverse calculation yields the result, which satisfies the following:

[0019]

[0020] in, It is the natural logarithmic function, and This is the preset lower limit temperature for cold-laying workability.

[0021] Preferably, the end-side control unit calculates the time delay from the completion of material discharge to the start of paving based on the process event timestamps. And satisfy:

[0022]

[0023] in, This is the timestamp for the completion of material output. The timestamp for the start of the stall event.

[0024] Preferably, the paving speed boundary is Furthermore, the end-side control unit is based on the time delay between the completion of material discharge and the start of paving. With the maximum allowed latency Select the paving speed within the paving speed boundary. ,satisfy:

[0025]

[0026] in, To control the length of the paving, To allow for a safety margin of duration, This is the lower limit constant for duration.

[0027] Preferably, when communication between the end-side control unit and the cloud-side control platform is interrupted, the end-side control unit selects execution values ​​within the parameter boundary set to continuously control the construction equipment, and after communication is restored, it transmits the process event timestamps, including at least the material discharge completion event timestamp and the paving start event timestamp, as well as the residual vector, to the cloud-side control platform; the end-side control unit generates a construction record corresponding to the old material batch identifier, and the construction record includes at least the parameter version identifier, the process event timestamp sequence, the execution value sequence, and the residual vector sequence, and stores it in association with the construction section identifier.

[0028] Preferably, the residual vector is ,in:

[0029]

[0030] in, This is a temperature measurement value. This is the measured compaction density. These are measured values ​​of workability indicators; predicted values ​​of paving start temperature. Predicted compaction density and predicted values ​​of work performance indicators All were calculated by the prediction model of claim 2; This indicates transpose.

[0031] Preferably, the deviation determination criteria include a residual threshold and a continuous determination, and satisfy the following:

[0032]

[0033] in, It is a norm 2. The residual threshold, It is an integer greater than or equal to 2.

[0034] Preferably, the sensitivity matrix is ​​as follows:

[0035]

[0036] in , The parameter vector to be updated is used; the cloud-based control platform determines the parameter subset based on the sensitivity matrix and the residual vector.

[0037] Preferably, the directional incremental recalculation includes calculating the parameter increment. parameter increment By solving the linear equation, we obtain:

[0038]

[0039] in, The regularization coefficient is . It is an identity matrix.

[0040] Preferably, the parameter subset is a pair The cloud-based control platform only performs targeted incremental recalculation on the parameter items corresponding to the subset of parameters that contribute the most.

[0041] Preferably, the differential parameter package contains only the set of parameter items that have changed relative to the previous parameter version identifier, the applicable batch range, and the applicable construction section range, and includes a verification summary for end-side consistency verification; the end-side control unit performs a local replacement update on the parameter boundary set based on the differential parameter package.

[0042] Preferably, the rollback conditions include at least the following: after the end-side control unit applies the differential parameter package, it satisfies the following within a preset observation window:

[0043]

[0044] in, For the updated residual vector, The residual vector before the update. This is the residual increment threshold; the end-side control unit rolls back to the previous parameter version identifier when the rollback conditions are met.

[0045] Compared with existing technologies, the above technical solution has the following advantages:

[0046] 1. Achieve closed-loop construction control through end-to-cloud collaboration, enhancing the stability and controllability of the integrated hot-mix and cold-laying construction process using recycled materials. The cloud side generates parameter boundary sets based on batch characterization data and distributes them to the end-to-end. The end-to-end selects execution values ​​within these boundaries to control each process. Residual feedback triggers directional incremental recalculation and differential updates on the cloud side, enabling online adjustment of construction parameters according to changes in working conditions, thereby reducing the impact of working condition fluctuations on construction quality.

[0047] 2. Under the premise of meeting the cold-laying workability constraints, quantitative constraints and scheduling of process connection delays are implemented to improve organizational fault tolerance. Verifiable constraint proof data is generated by predicting temperature decay and inversely calculating the upper limit of allowable delay. This data is then converted into executable paving speed constraints at the end side, enabling the end side to dynamically select the paving speed based on the delay margin, thereby reducing the risk of insufficient paving start temperature leading to substandard workability.

[0048] 3. Control is implemented through a set of parameter boundaries rather than single-point setpoints, balancing adaptability and stability. The cloud-side outputs the discharge temperature boundary, paving speed boundary, and compaction pass number boundary. The end-side selects the execution value within the boundary range by combining real-time measurement and equipment status, thereby maintaining a feasible construction domain under uncertain working conditions and improving adaptability to material differences, environmental changes, and equipment fluctuations.

[0049] 4. Targeted incremental recalculation is achieved through residual vectors and sensitivity matrices, reducing unnecessary full-scale perturbations and improving update efficiency. When a deviation judgment is triggered, the cloud side can determine the subset of parameters that contribute significantly to the deviation based on the sensitivity matrix and perform targeted updates, generating differential parameter packages for local replacement, thereby reducing additional fluctuations introduced by parameter updates and improving update speed.

[0050] 5. Provides differential update and rollback mechanisms to improve the security and recoverability of online parameter updates. After applying the differential parameter package, the endpoint can determine whether the update has aggravated the deviation based on rollback conditions. If the conditions are met, it rolls back to the previous parameter version identifier, and ensures consistency between the endpoint and cloud versions through version chain freezing and recalculation mechanisms, thereby reducing the adverse effects of erroneous updates on the construction process.

[0051] 6. Achieve batch-level data organization and traceability to enhance the traceability and quality verification capabilities of the construction process. By using old material batch identifiers, batch characterization data, parameter version identifiers, execution values, process event timestamps, and residual vectors are associated and stored to form traceable construction records, facilitating the location analysis, quality verification, and cause tracing of abnormal batches.

[0052] 7. Maintain continuous control within the boundary under adverse network conditions such as communication interruptions, and retransmit critical data after recovery to improve on-site operational stability. On the end side, during a link failure, continue control using the most recently validated parameter boundary set, and employ local persistent caching and a tiered retransmission mechanism to supplement deviation information and construction records after network recovery, thereby reducing the impact of network instability on construction continuity.

[0053] 8. Integrate workability indicators and compaction quality into a unified deviation evaluation system to improve the ability to identify comprehensive quality risks. By constructing a residual vector that includes temperature, density, and workability components and using its L2 norm for deviation determination, the overall degree of deviation in the construction process can be more comprehensively reflected, thereby triggering parameter updates more promptly and improving quality stability. Attached Figure Description

[0054] Figure 1 This diagram illustrates the overall architecture and data interaction relationships of the integrated hot-mix and cold-pave construction method for municipal roads using end-to-end cloud collaboration and recycled materials.

[0055] Figure 2 This is a schematic diagram showing the temperature decay prediction curve and temperature measurement comparison of recycled mixture from the completion of discharge to the start of paving.

[0056] Figure 3 This is a schematic diagram showing the variation of the L2 norm of the residual vectors for multiple batches of recycled materials with each batch.

[0057] Figure 4 A schematic diagram comparing the temperature residuals before and after directional incremental recalculation.

[0058] Figure 5 This is a schematic diagram of the lower limit curve of paving speed under different selection rules or different parameter configurations.

[0059] Figure 6 A schematic diagram showing the curve of the allowable time delay limit as a function of the discharge temperature.

[0060] Figure 7 This is a schematic diagram of the tracking curve of the workability index on the batch dimension and the constraint band of the workability index.

[0061] Figure 8 This is a schematic diagram showing the relationship between the number of compaction passes and the predicted compaction density.

[0062] Figure 9 This is a schematic diagram illustrating the steps of an integrated hot-mix and cold-laying construction method for recycled municipal road materials using a cloud-based collaborative approach. Detailed Implementation

[0063] The advantages of the present invention will be further illustrated below with reference to the accompanying drawings and specific embodiments.

[0064] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0065] The terminology used in this disclosure is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The singular forms “a,” “the,” and “the” as used in this disclosure and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0066] It should be understood that although the terms first, second, third, etc., may be used in this disclosure to describe various information, such information should not be limited to these terms. These terms are used only to distinguish information of the same type from one another. For example, without departing from the scope of this disclosure, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0067] In the description of this invention, it should be understood that the terms "longitudinal", "lateral", "up", "down", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0068] In the description of this invention, unless otherwise specified and limited, it should be noted that the terms "installation", "connection" and "linking" should be interpreted broadly. For example, they can refer to mechanical or electrical connections, or internal connections between two components. They can be direct connections or indirect connections through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms according to the specific circumstances.

[0069] In the following description, suffixes such as "module," "part," or "unit" used to denote elements are used only for the convenience of the description of the invention and have no specific meaning in themselves. Therefore, "module" and "part" can be used interchangeably.

[0070] See Figure 9As shown, this embodiment provides an integrated hot-mix and cold-pave construction method for municipal road recycled materials using a cloud-based collaborative approach. The method is executed collaboratively by a cloud-based control platform and a terminal control unit. The terminal control unit connects to and controls the construction equipment, which includes at least milling equipment, heating and dispersing equipment, mixing equipment, paving equipment, and compaction equipment. The method includes: S1, the terminal control unit generates batch identifiers for municipal road recycled materials entering the recycling process, collects batch characterization data corresponding to the batch identifiers, and uploads the batch characterization data and construction section identifiers to the cloud-based control platform; S2, the cloud-based control platform constructs a prediction model based on the batch characterization data and performs constraint solving under conditions that satisfy at least cold-pavement workability constraints and compaction quality constraints, obtaining parameter boundary sets and constraint proof data, and generating parameter version identifiers; S3, the cloud-based control platform... The control platform sends the parameter boundary set, constraint proof data, parameter version identifier, and validity conditions to the end-side control unit. After the end-side control unit verifies the data, it binds and stores the old material batch identifier with the parameter version identifier. In S4, the end-side control unit selects the execution value within the parameter boundary set limit to control the construction equipment to complete milling, heating and dispersing, hot mixing, cold paving and compaction, and collects construction measurement values ​​to form a residual vector. In S5, when the residual vector meets the deviation judgment condition, the end-side control unit sends the residual vector to the cloud-side control platform. The cloud-side control platform determines the parameter subset that needs to be updated based on the residual vector and the sensitivity matrix and performs directional incremental recalculation to generate a new parameter version identifier, which is then distributed in the form of a differential parameter package. The end-side control unit updates the parameter boundary set according to the differential parameter package and rolls back to the previous parameter version identifier when the rollback condition is met.

[0071] In this embodiment, firstly... Figures 1 to 8 Describe it.

[0072] Figure 1 The diagram illustrates the overall architecture and data interaction of the end-cloud collaborative integrated hot-mix and cold-pavement construction method for municipal road waste material recycling. An end-side control unit is installed at the construction site. This control unit is electrically connected to milling equipment, heating and dispersing equipment, mixing equipment, paving equipment, and compaction equipment. It is also connected to a sensor acquisition module and a storage module. This module is used to batch-divide the municipal road waste materials entering the recycling process, generate batch identifiers, collect batch characterization data, and generate upload messages. Figure 1 The diagram also illustrates the uplink and downlink information flows between the cloud-side control platform and the edge-side control unit. The uplink information flow includes at least batch characterization data and construction section identifiers, process event timestamps, residual vectors, rollback notifications, and construction record summaries. The downlink information flow includes at least parameter boundary sets, constraint proof data, parameter version identifiers and validity conditions, and differential parameter packages. Figure 1The structure shown embodies the closed-loop relationship of this method. The end-side control unit selects the execution value within the boundary to control the construction equipment based on the parameter boundary set issued by the cloud-side control platform. The end-side control unit generates a residual vector based on the construction measurement value to trigger the cloud-side control platform to perform directional incremental recalculation. The cloud-side control platform generates a differential parameter package and issues it to realize parameter version update. If the rollback condition is met after the update, the end-side control unit rolls back to the previous parameter version identifier, thus forming a closed-loop control link for end-cloud collaboration.

[0073] Figure 2 The diagram shows the temperature decay prediction curve and temperature measurement value comparison of the recycled mixture from the completion of discharge to the start of paving. The temperature decay prediction curve is calculated by the cloud-based control platform based on the temperature decay prediction model. The model takes the ambient temperature, discharge temperature, temperature decay coefficient and the time delay from the completion of discharge to the start of paving as inputs, and outputs the predicted temperature value at the start of paving. Figure 2 The corresponding temperature measurement value is obtained by the end-side control unit through an infrared thermometer array near the start of paving and then aggregated by a window. By comparing the predicted curve with the measured value, the degree of fit of the prediction model to the actual temperature decay process can be characterized, and a basis can be provided for the calculation of the temperature residual in the subsequent residual vector. Figure 2 The lower limit temperature line for cold paving workability is also shown. This lower limit temperature line characterizes the minimum temperature requirement for workability in the cold paving process. When the predicted paving start temperature or the measured temperature is lower than this lower limit temperature, it indicates that workability may be insufficient, thus requiring adjustments such as increasing the discharge temperature, reducing the organization delay, or increasing the paving speed. Therefore, Figure 2 This is used to intuitively illustrate the significance of the temperature decay prediction model and the role of cold-laid workability constraints in constraint solving.

[0074] Figure 3 The diagram shows the curves of the residual vector L2 norm as a function of multiple batches of recycled materials. The residual vector is constructed by the end-side control unit based on the difference between the construction measurement value and the cloud-side prediction value. The residual vector L2 norm is used to characterize the comprehensive degree of deviation in three aspects: temperature, compaction density and workability. Figure 3 The horizontal axis represents the batch index, and the vertical axis represents the L2 norm of the residual vector. The larger the value of the L2 norm of the residual vector, the greater the overall deviation and the worse the process stability; conversely, the smaller the L2 norm of the residual vector, the smaller the overall deviation and the better the process stability. Figure 3 The sharp peaks in the curve usually correspond to sudden changes in operating conditions, such as fluctuations in the moisture content of old materials, changes in ambient temperature, fluctuations in the time delay of the structure, or changes in the equipment status, which increases the difference between the predicted value and the measured value. Figure 3It is used to intuitively demonstrate the impact of the edge-cloud collaborative closed loop on process stability, and to provide a visual reference for setting the deviation judgment threshold and the number of consecutive judgments. It is also used to illustrate the trend of the overall level of the residual vector L2 norm decreasing after directional incremental recalculation and differential update.

[0075] Figure 4 A schematic diagram showing the comparison curves of temperature residuals before and after directional incremental recalculation is presented. The temperature residuals are obtained by the difference between the measured temperature value and the predicted temperature at the start of paving. Figure 4 The data includes at least two types: pre-update and post-update data. The pre-update temperature residual corresponds to the prediction model output of the end-side control unit using the old parameter version identifier; the post-update temperature residual corresponds to the prediction model output of the end-side control unit using the new parameter version identifier updated by the differential parameter package issued by the cloud-side control platform. By comparing the amplitude and fluctuation of the temperature residual before and after the update, the effect of directional incremental recalculation on correcting temperature prediction errors can be intuitively characterized. That is, when the amplitude of the temperature residual decreases and the fluctuation decreases after the update, it indicates that the cloud-side control platform can determine the parameter subset based on the residual vector and sensitivity matrix and perform directional updates, making the prediction results closer to the measurement results. Therefore, Figure 4 This is used to illustrate the convergence and effectiveness of directional incremental recalculation and differential update in the temperature dimension, and to support the necessity of the rollback mechanism, that is, when the residual increases after the update, a rollback can be triggered to ensure process stability.

[0076] Figure 5 The diagram shows the lower limit curve of paving speed under different selection rules or different parameter configurations, which is used to illustrate that the selection of paving speed within the boundary is not arbitrary, but a constraint result derived from factors such as the upper limit of allowable time delay and safety margin. Figure 5 The horizontal axis can be understood as a variable related to organization delay or delay margin, while the vertical axis represents the lower limit of paving speed. The lower limit of paving speed is jointly determined by the paving control length, the upper limit of allowable delay, the safety margin duration, and the lower limit constant of duration. When the result of subtracting the safety margin duration from the upper limit of allowable delay becomes smaller, the lower limit of paving speed will increase to ensure that the corresponding paving control length is completed within the specified time, thereby reducing the risk that the paving start temperature will be lower than the workability lower limit temperature. Figure 5 This demonstrates that after receiving the allowable delay limit issued by the cloud-side control platform, the end-side control unit can convert the verification data into executable paving speed constraints and select an execution value for the paving speed that satisfies the lower bound constraint within the paving speed boundary range; simultaneously... Figure 5 It is also used to explain that when the calculated lower limit of speed exceeds the upper limit of paving speed boundary, an over-limit response should be triggered or the parameter boundary set should be resolved.

[0077] Figure 6The diagram shows the curve of the allowable time delay upper limit as a function of the discharge temperature. The allowable time delay upper limit is obtained by the cloud-side control platform based on the temperature decay coefficient, ambient temperature, lower limit temperature for cold laying workability, and discharge temperature. Figure 6 The horizontal axis represents the discharge temperature, and the vertical axis represents the upper limit of the allowable time delay. The curve shape reflects that increasing the discharge temperature can increase the upper limit of the allowable time delay. That is, under the condition that the ambient temperature and the lower limit of the working temperature remain unchanged, increasing the discharge temperature can expand the range of time delay that the process organization can tolerate, thereby improving the fault tolerance of on-site organization fluctuations. Figure 6 This is used to illustrate the quantitative relationship between the discharge temperature boundary and the upper limit of the allowable time delay, and to support the cloud-based control platform in determining the discharge temperature boundary / paving speed boundary in constraint solving, etc.

[0078] Figure 7 The diagram shows the tracking curve of the workability index in the batch dimension and the schematic diagram of the workability index constraint band. The workability index is a normalized constructability characterization quantity, which can be constructed from factors such as the mixing torque measurement value and the temperature difference normalization term. The workability index measurement value is calculated by the end-side control unit based on the data obtained by the sensor acquisition module, and the workability index prediction value is calculated by the cloud-side control platform based on the prediction model or the empirical model. Figure 7 Typically, a lower and upper limit threshold for workability indicators are set to form a constraint band. When the measured or predicted value of the workability indicator falls within this constraint band, it indicates that the material has acceptable workability under the current process conditions. When it falls outside the constraint band, it indicates that there may be issues such as excessive hardness, excessive softness, or abnormal fluidity, which need to be corrected by adjusting the discharge temperature, mixing cycle, paving speed, or organization delay. Figure 7 This is used to illustrate the observability and necessity of the workability index as a constraint solution input and residual vector component, and to demonstrate the improvement effect of the edge-cloud collaborative closed loop on workability stability.

[0079] Figure 8 A schematic diagram of the relationship between the number of compaction passes and the predicted compaction density is shown, along with a target density threshold line. The predicted compaction density can be composed of a density reference term, a pass gain term, and a temperature window gain term. The number of compaction passes is the execution value selected by the end-side control unit within the compaction pass boundary. Figure 8 The horizontal axis represents the number of compaction passes, and the vertical axis represents the predicted compaction density. The curve reflects that increasing the number of compaction passes usually increases the compaction density; meanwhile, Figure 8 Multiple curves are used to represent the density increase trend under different paving start temperatures. When the paving start temperature is low, the density increase slope may decrease, and the minimum number of passes required to reach the target density threshold will increase accordingly. Figure 8To intuitively explain the formation logic of the compaction pass number boundary, the cloud-based control platform can deduce the lower bound of the compaction pass number through the constraint that "the predicted compaction density value is not lower than the target density threshold," and generate the compaction pass number boundary in the constraint solution; simultaneously... Figure 8 It is also used to illustrate the role of the linkage control between the number of compaction passes and the temperature window in ensuring compaction quality.

[0080] This embodiment uses a municipal road renovation project as an example. Milling equipment, heating and dispersing equipment, mixing equipment, paving equipment, and compaction equipment are deployed at the construction site. An edge-cloud collaborative construction control system is configured to generate, distribute, execute, and adjust parameters online. The edge-cloud collaborative construction control system includes a cloud-side control platform, edge-side control units, sensor acquisition modules, and storage modules. The cloud-side control platform is deployed in the cloud or at a central location, while the edge-side control units are deployed at the construction site and electrically connected to the construction equipment and sensor acquisition modules.

[0081] The cloud-based control platform receives batch characterization data and construction section identifiers uploaded by the end-side control unit, constructs a prediction model and performs constraint solving, outputs parameter boundary sets and constraint proof data, and generates parameter version identifiers. When the end-side control unit triggers a deviation report, the cloud-based control platform performs directional incremental recalculation based on the residual vector and sensitivity matrix, generates a new parameter version identifier, and distributes it in the form of differential parameter packages. The end-side control unit is used to divide the continuous old material flow into old material batches and generate old material batch identifiers, collect batch characterization data corresponding to the old material batch identifiers and complete the upload; the end-side control unit receives and verifies the parameter boundary set, constraint proof data, parameter version identifier and validity conditions sent by the cloud side, and binds and stores the old material batch identifier with the parameter version identifier after passing the verification; the end-side control unit selects the execution value within the limit of the parameter boundary set to control the construction equipment to complete milling, heating and dispersing, hot mixing, cold paving and compaction, and collects construction measurement values ​​to form a residual vector; the end-side control unit is also used to maintain continuous control within the boundary when communication is interrupted and to retransmit key data after communication is restored, while generating construction records corresponding to the old material batch identifiers to support traceability and verification.

[0082] The recycled material batch identifier is a unique identifier for each batch of recycled material entering the recycling process, based on its particle size. This identifier is used to achieve closed-loop consistency across the entire process: batch characterization data – parameter version – executed value – residual – update. Batch characterization data is a structured dataset representing the material state, environmental state, and microstructure of a specific batch of recycled material. It includes at least fields such as recycled material moisture content, initial recycled material temperature, and ambient temperature, and optionally fields such as gradation index, wind speed, flow statistics, and equipment status summary. The parameter boundary set is a description of the feasible range of construction parameters provided by the cloud-side control platform after constraint solving. It includes at least the discharge temperature boundary, paving speed boundary, and compaction pass number boundary. Constraint proof data consists of verifiable evidence fields provided by the cloud-side control platform for the feasibility of the parameter boundary set. It includes at least the predicted paving start temperature and the upper limit of allowable delay. The parameter version identifier is an identifier used by the cloud-side control platform for version management of a specific solution output, used for differential update and rollback consistency control. The residual vector is the error vector obtained by the end-side control unit from the difference between the construction measurement value and the predicted value, used for deviation judgment and cloud-side directional incremental recalculation. The sensitivity matrix is ​​the partial derivative matrix of the predicted output for the parameters to be updated, used by the cloud-side control platform to determine the subset of parameters that need updating and to achieve directional updates. The differential parameter package is a differential data package sent by the cloud-side control platform to the end-side control unit after encapsulating the differences between the old and new versions, used for partial replacement updates by the end-side control unit. The rollback condition is the trigger condition used by the end-side control unit to determine if the deviation worsens after the update and to revert to the previous parameter version identifier.

[0083] The sensor acquisition module is used to acquire measurements and events related to batch characterization data, constraint proof data verification, and residual vector construction, including but not limited to ambient temperature measurement, old material moisture content measurement, old material initial temperature measurement, discharge temperature measurement, paving start temperature measurement, compaction density measurement, workability index measurement, conveying flow rate measurement, and process event timestamps. The storage module is used to cache the upload queue, store construction records, store historical parameter versions and data required for rollback, and supports re-upload in case of link failure.

[0084] For ease of understanding, steps S1 to S5 will be described in detail in this example.

[0085] Step S1: "Generate batch identifiers for recycled materials and collect batch characterization data according to batches":

[0086] The end-side control unit generates batch identifiers for municipal road waste entering the recycling process. This embodiment employs a composite segmentation method combining "quality threshold triggering + time window fallback + anomaly triggering." The end-side control unit reads the cumulative mass output by the conveyor belt scale and sets the target cumulative mass for each batch as the batch target mass. For example, it can be set to 3.0 tons; when the cumulative mass reaches 3.0 tons, the end-side control unit generates a batch boundary event, ending the current batch and starting the next batch. If equipment start-up / shutdown or conveying fluctuations cause the cumulative mass to fail to reach the threshold for an extended period, the end-side control unit will use the maximum batch duration. As a fallback condition, for example, a 6-minute timeframe can be set. If the current batch has not reached 3.0 tons within 6 minutes, a batch boundary event will also be generated. Anomaly triggering is used to handle situations such as equipment shutdown, conveying interruption, and sensor failure. When the end-side control unit detects an anomaly trigger signal, it immediately terminates the current batch and starts the next batch, while simultaneously writing the anomaly cause into the batch context object.

[0087] To reduce cross-batch mixing caused by material retention at batch boundaries, the end-side control unit sets a transition window duration. For example, it can be set to 5 seconds; the end-side control unit marks the samples within the transition window as boundary transition samples and reduces their weights during subsequent aggregation. The specific way to reduce the weights is to assign lower aggregation weights to boundary transition samples than normal samples during batch aggregation, for example, setting the weight of samples within the transition window to 0.2 and the weight of samples outside the transition window to 1.0, in order to reduce the impact of boundary aliasing.

[0088] The old material batch identifier is generated using a "field concatenation + checksum" method. The old material batch identifier includes at least the following fields: construction section identifier, end-side control unit identifier, date and time, batch number, and checksum. For example, the construction section identifier field is "SEG03", the end-side control unit identifier field is "EDGE01", the date and time field is the year, month, day, hour, minute, and second of the batch start event timestamp, for example, "20260202-100500", the batch number field is "0007", and the checksum field is either CRC16 or the first few bits of a hash digest, for example, "A3F9". After the end-side control unit generates the old material batch identifier, it writes the old material batch identifier into the batch context object and uses it as the primary key field for subsequent data packets. Simultaneously, the old material batch identifier is written into the construction record index field to ensure consistency in the subsequent "data-parameter-execution-residual-update" association.

[0089] The end-side control unit collects batch characterization data corresponding to the old material batch identifier and forms a batch characterization data package. The batch characterization data includes at least the old material moisture content characterization, the old material initial temperature characterization, and the ambient temperature characterization, and optionally includes gradation index characterization, wind speed characterization, old material flow statistics characterization, and equipment operating status summary characterization. The collection and aggregation process is executed according to the following default implementation:

[0090] Reclaimed material moisture content characterization: A near-infrared moisture content sensor is installed above the reclaimed material conveyor belt. The near-infrared moisture content sensor outputs the moisture content measurement value at a fixed sampling frequency; the default sampling frequency is 2Hz. The end-side control unit performs amplitude limiting filtering on the moisture content measurement value (the upper and lower limits of the amplitude limiting can be set according to the sensor range, for example, 0% to 10%), and then performs sliding median filtering (the default window length is 5 samples) to remove instantaneous jump points; after filtering, the batch average moisture content and batch variance are calculated within the batch window as the moisture content characterization fields of that batch of reclaimed material.

[0091] Initial temperature characterization of recycled materials: A temperature acquisition device is installed before the recycled materials enter the heating and dispersing equipment. The temperature acquisition device outputs the initial temperature measurement value of the recycled materials at a fixed sampling frequency; the default sampling frequency is 2Hz. The end-side control unit takes the median of the temperature sequence within the batch window to obtain the initial temperature characterization of the recycled materials, and calculates the upper quartile and lower quartile to form a temperature fluctuation characterization field, so that the cloud-side control platform can judge whether the initial temperature of the recycled materials is abnormal or fluctuates too much.

[0092] Ambient temperature characterization: Ambient temperature data acquisition devices are deployed at locations away from direct heat sources on the construction site. These devices output ambient temperature measurements at a low frequency; the default sampling period is 60 seconds. The end-side control unit aligns the ambient temperature measurements by nearest neighbor or linear interpolation based on the batch start timestamp, forming a batch ambient temperature characterization field.

[0093] Grading index characterization: An image acquisition device is arranged above the old material conveying channel to acquire images of the old material surface and perform particle size estimation in the end control unit or edge vision module to obtain several quantiles of particle size distribution, such as 10% quantile, 50% quantile and 90% quantile, as gradation index characterization fields.

[0094] Wind speed characterization: Anemometers are deployed in the paving area, with a default sampling frequency of 1Hz. The end-side control unit calculates the average and peak wind speeds within a batch window as wind speed characterization fields, which are used by the cloud-side control platform to measure the temperature attenuation coefficient. Verification and correction.

[0095] The flow rate of recycled materials is characterized by the instantaneous flow rate output by the conveyor belt scale. The default sampling frequency is 1Hz. The end control unit calculates the average flow rate and peak flow rate within the batch window as the flow rate statistical representation fields.

[0096] The edge control unit adds a sampling timestamp to the above-mentioned sampled data and corrects clock drift through the time synchronization module. The default time synchronization method is as follows: the edge control unit obtains the calibration time from the time source every 60 seconds and estimates the local clock drift rate, and performs linear drift compensation on the sampling timestamp to ensure that all measured quantities and event quantities are aligned under the same time reference.

[0097] The edge control unit generates a batch characterization data packet, which includes at least the old material batch identifier, construction section identifier, batch start event timestamp, batch end event timestamp, and a set of batch characterization data fields. A checksum is calculated for the message body and written to the packet header. The checksum is generated by default using a hash digest or CRC checksum. The cloud-side control platform recalculates and compares the checksum upon receipt to verify its integrity.

[0098] For ease of understanding, Example 1 will be provided in this embodiment:

[0099] Set batch target quality The maximum duration of the batch is 3.0 tons. The transition window lasts for 6 minutes. The time interval is 5 seconds. The construction section is identified as "SEG03", and the end-side control unit is identified as "EDGE01". When the cumulative mass of the conveyor belt scale reaches 3.0 tons, the end-side control unit triggers a batch boundary event and ends the current batch. The timestamp for the start of the next batch is 2025-02-02 10:05:00, the batch number is 7, and the end-side control unit generates a used material batch identifier "SEG03-EDGE01-20250202-100500-0007-A3F9". After amplitude limiting and median filtering, the average moisture content of this batch is 3.2%, the median value of the initial temperature sensor for the used material is 18°C, the ambient temperature is 10°C, the wind speed is 2.8 m / s, and the average conveying flow rate is 0.55 t / min. The end-side control unit forms a batch characterization data package containing the above fields and a verification summary.

[0100] Step S2: "Upload batch characterization data and construction section identifiers to the cloud-based control platform":

[0101] The edge control unit uploads batch characterization data and construction section identifiers to the cloud-based control platform. This embodiment employs an upload mechanism of "layered upload queues + priority + end-to-end confirmation and retransmission + local persistent caching". The edge control unit divides the uploaded data into batch characterization data packets, deviation information packets, and construction record summary packets, and sets upload queues and priorities for each, with deviation information packets having the highest priority. The edge control unit serializes the batch characterization data packets into structured messages and writes them to the upload cache queue. The message header includes at least the old material batch identifier, construction section identifier, message type, message sequence number, verification summary, and retransmission count. The message body contains a set of batch characterization data fields and timestamp information.

[0102] The edge control unit sends messages to the data access module of the cloud control platform via a communication link, with the default transmission method being message queue protocol or HTTPS interface. After sending, the edge control unit waits for an acknowledgment, which includes at least the message sequence number, verification result, and storage status. If no acknowledgment is received within a preset timeout period, retransmission is exponentially backed up, and the retransmission count is incremented and written to the message header. If the retransmission limit is reached and the transmission still fails, the message is retained in a local persistent queue and marked for network recovery. Upon receiving the message, the cloud control platform performs verification digest verification and field integrity checks. If the verification passes, the message is stored and an acknowledgment is returned; if the verification fails, a rejection acknowledgment with the reason for failure is returned. The cloud control platform performs idempotent processing on message sequence numbers, meaning that if messages with the same batch identifier and message sequence number arrive repeatedly, they are not stored repeatedly; instead, a processed acknowledgment is returned directly, thus avoiding duplicate writing caused by network outage retransmissions.

[0103] When the network is interrupted, the edge control unit records the timestamp of the network interruption start and continuously writes messages to the local persistent queue. After the network is restored, the data is retransmitted in priority order, with the default retransmission order being: deviation information packet → batch characterization data packet → construction record summary packet, thereby prioritizing the recovery of key data required for closed-loop update.

[0104] For ease of understanding, Example 2 will be provided in this embodiment:

[0105] The edge control unit encapsulates the batch identification data packet with the old material batch identifier "SEG03-EDGE01-20250202-100500-0007-A3F9" into a message and assigns it the message sequence number "MSG-000125", and sends it to the cloud control platform. If no acknowledgment is received within 3 seconds, it retransmits the message at backoff intervals of 1 second, 2 seconds, and 4 seconds, and writes the retransmission count into the message header. If the cloud control platform receives "MSG-000125" repeatedly, it returns a processed confirmation. After receiving the acknowledgment, the edge control unit removes the message from the queue and records the delivered status.

[0106] Step S3: "Build a prediction model on the cloud side and solve it under constraints, outputting the parameter boundary set, constraint proof data, parameter version identifier, and validity conditions":

[0107] After receiving the batch characterization data packets, the cloud-based control platform enters the prediction model construction and constraint solving stage. Based on the batch characterization data, the cloud-based control platform constructs a prediction model. This model is used at least to calculate the predicted paving start temperature and the upper limit of the allowable time delay, and can optionally be used to calculate the predicted compaction density and workability index. Under the condition that at least cold-paving workability constraints and compaction quality constraints are met, the cloud-based control platform performs constraint solving, outputs the parameter boundary set and constraint proof data, and generates a parameter version identifier.

[0108] The predicted paving start temperature value is used. This means that the following conditions are met:

[0109]

[0110] Among them, ambient temperature The ambient temperature at the construction site and the discharge temperature are mentioned. The discharge temperature of the recycled mixture, and the time delay from the completion of discharge to the start of paving. Temperature decay coefficient is the time between the completion of material discharge and the start of paving. These are the parameters for the temperature decay model.

[0111] For ease of understanding, an example is provided in this embodiment: Let the ambient temperature be... discharge temperature Temperature decay coefficient Delay .

[0112] Step 1: .

[0113] Step 2: .

[0114] Step 3: .

[0115] Step 4: .

[0116] Step 5: .

[0117] The allowed latency limit is adopted. This indicates that the lower limit temperature for cold-laid workability is adopted. This means that the result is obtained by inverse calculation:

[0118]

[0119] For ease of understanding, an example is provided in this embodiment: Let the ambient temperature be... discharge temperature Lower limit temperature for cold-laid workability Temperature decay coefficient .

[0120] Step 1: .

[0121] Step 2: .

[0122] Step 3: Ratio .

[0123] Step 4: .

[0124] Step 5: .

[0125] Step 6: .

[0126] Figure 2 and Figure 6 This is used to illustrate the changing trends of temperature prediction and the upper limit of allowable delay.

[0127] To ensure the predictive model is usable in different construction sections and under different environmental conditions, the cloud-based control platform adjusts the temperature attenuation coefficient. The implementation method employs "segment grouping calibration + online residual correction". In the segment grouping calibration stage, the cloud-based control platform groups historical construction records according to the construction segment identifier. From each group, it extracts a sample set of "discharge temperature measurement, ambient temperature measurement, time delay from discharge to paving, and paving start temperature measurement". Using an exponential decay model as the structure, it fits the temperature decay coefficient to obtain the initial temperature decay coefficient for that construction segment. In the online residual correction stage, when the end-side control unit reports the temperature residual... Furthermore, when a deviation is triggered, the cloud-based control platform incorporates the temperature decay coefficient as part of the parameter vector to be updated during directional incremental recalculation. It uses the residual vector to drive a small update of the temperature decay coefficient, setting an upper limit on the update magnitude to prevent drastic changes in the temperature decay coefficient due to noise. The cloud-based control platform manages the updated temperature decay coefficient along with the parameter version identifier and records version iteration logs for easy traceability.

[0128] The compaction density prediction value is adopted. This indicates that the number of compaction passes is used. This indicates the predicted compaction density value. Regression can be used:

[0129]

[0130] Density reference coefficient The baseline density term, density regression coefficient For example, the gain coefficient is the pass-through coefficient, and the temperature regression coefficient is the temperature regression coefficient. This refers to the temperature window gain coefficient. The cloud-based control platform can maintain the data according to the construction section identification. , , And it will be updated with each version iteration; Figure 8 The variation trends of the number of compaction passes and the predicted compaction density under different paving start temperatures are shown to explain the influence of temperature window and number of passes on density prediction.

[0131] The predicted values ​​of the work performance indicators are used. Indicated. Work performance index measurements are taken using work performance index measurements. This indicates that, to provide a default implementation, this embodiment uses the "mixing torque measurement value". +Temperature difference normalization" is used to construct the workability indicators. The mixing torque measurement value is adopted. This indicates that the torque normalization benchmark adopts the torque benchmark. This indicates that the temperature difference normalization benchmark adopts the temperature difference benchmark. This indicates that the predicted value of the work performance index is... Linear normalization and truncation are applied:

[0132]

[0133] Among them, the working bias coefficient Working torque weighting coefficient Weighting coefficient for working temperature difference These are the model parameters. Figure 7 The tracking trend of the performance index within the constraint band is shown to illustrate the observability of the performance index.

[0134] For ease of understanding, an example is provided in this embodiment: Let... , , ;set up , ;set up , , .

[0135] Step 1: .

[0136] Step 2: .

[0137] Step 3: Linear combination = 0.20 + 0.50 × 0.85 + 0.25 × 1 = 0.875.

[0138] Step 4: The value remains 0.875 after truncation.

[0139] therefore =0.875.

[0140] When the cloud-based control platform performs constraint solving, it must input at least the following information: construction section identifier, batch characterization data, equipment capability boundaries (e.g., controllable range of discharge temperature, controllable range of paving speed, controllable range of compaction passes), prediction model parameters (e.g., temperature decay coefficient, density regression coefficient, workability model coefficient), and preset thresholds (e.g., lower limit temperature for cold paving workability, target density threshold, workability index constraint zone threshold, etc.). The constraint set of the cloud-based control platform must include at least: the predicted paving start temperature is not lower than the lower limit temperature for cold paving workability; the upper limit of allowable delay meets the requirements for process organization and implementation; the predicted compaction density is not lower than the target density threshold; and the predicted workability index is within the preset constraint zone. Under the premise of satisfying the constraint set, the cloud-based control platform generates a feasible solution set and constructs a parameter boundary set and proof data from the feasible solution set.

[0141] The cloud-based control platform employs a discretization-based screening method: discharge temperature, paving speed, and number of compaction passes are discretized within the equipment capacity boundary at step sizes to form candidate combinations. For each candidate combination, the platform uses a temperature decay model to calculate the predicted paving start temperature and determine if it meets the cold-laying workability constraint. Simultaneously, it uses a density prediction model to calculate the predicted compaction density and determine if it meets the target density threshold. Furthermore, it uses a workability model or constraint bands to determine if the workability constraint is met. Candidate combinations that satisfy all constraints are included in the feasible solution set. To ensure that the parameter boundary set is not widened by extreme outliers, the cloud-based control platform can sort the feasible solution set in the order of "prioritizing minimum adjustment range, prioritizing minimum energy consumption, and prioritizing minimum rework risk," and select a subset of the top-ranked feasible solutions to construct the boundary. When constructing the boundaries, the cloud-based control platform calculates the minimum and maximum values ​​of the discharge temperature for the selected feasible solution set to form the discharge temperature boundary, calculates the minimum and maximum values ​​of the paving speed to form the paving speed boundary, and calculates the minimum and maximum values ​​of the number of compaction passes to form the compaction pass boundary, thereby obtaining the parameter boundary set.

[0142] The constraint proof data must include at least the predicted paving start temperature. With the maximum allowed latency The cloud-based control platform will also write key input summaries used to verify the validity of the data into the verification data, such as fields like the ambient temperature of this batch, temperature decay coefficient, and discharge temperature, so that the end-side control unit can perform consistency verification during calibration.

[0143] The validity conditions must include at least the following: the batch characterization data fields fall within the model's applicable boundary range. The model's applicable boundary range can be obtained by the cloud-side control platform based on historical training samples, such as the moisture content range, the initial temperature range of the recycled material, and the ambient temperature range. When the batch characterization data exceeds the applicable boundary range, the cloud-side control platform sets the validity condition to "not met" and includes an "inapplicability reason code" in the parameter package. When the edge control unit verifies that the validity condition is not met, it marks the batch as "requiring manual confirmation" and prompts the user to adopt a conservative strategy or stop the system for confirmation, thereby preventing model misuse.

[0144] The cloud-based control platform generates a parameter version identifier for this output. The parameter version identifier is generated according to the format of "construction section identifier + batch range + version serial number + check code". The parameter version identifier is then mapped to the old material batch identifier for traceability.

[0145] Step S4: "Parameter distribution and end-side verification, version binding, selection of execution values ​​within the boundary, and control of construction equipment to complete the operation":

[0146] The cloud-based control platform sends the parameter boundary set, constraint proof data, parameter version identifier, and validity conditions to the edge control unit. The cloud-based control platform encapsulates this information into a parameter package. The parameter package header must include at least the parameter version identifier, applicable construction section range, applicable batch range, validity condition summary, verification summary, and issuance sequence number. The parameter package body includes discharge temperature boundary, paving speed boundary, compaction pass number boundary, and proof data fields. After sending the parameter package, the cloud-based control platform waits for a receipt confirmation from the edge control unit. The confirmation receipt must include at least the parameter version identifier, verification result, and storage result.

[0147] After receiving the parameter packet, the edge control unit performs a consistency check, which includes at least: verification summary verification, parameter version identifier format verification, applicable construction section range check, applicable batch range check, validity condition check, and verification of the consistency of supporting data. If the check passes, the edge control unit binds and stores the old material batch identifier with the parameter version identifier and establishes a mapping relationship in the construction record. If the check fails, the edge control unit returns a rejection receipt and records the reason for the failure locally. Based on this, the cloud-based control platform can reissue the parameter packet or prompt for manual intervention.

[0148] The end-side control unit selects execution values ​​within the parameter boundary set. The default execution value selection strategy is to use the boundary median as the initial execution value and make fine adjustments based on real-time working conditions without exceeding the boundary. The fine-tuning criteria include at least: the trend of process connection delay, the deviation of the discharge temperature measurement value, the deviation of the paving start temperature measurement value, and the equipment load status. The execution values ​​selected by the end-side control unit include at least: the discharge temperature setpoint, the paving speed setpoint, and the compaction pass setpoint. At the same time, the end-side control unit can select heating power setpoint, conveying speed setpoint, mixing time setpoint, vibration mode setpoint, etc. as equipment-level auxiliary execution values. The auxiliary execution values ​​should also meet the equipment capacity limitations and be consistent with the main execution values.

[0149] The end-side control unit issues control commands through the equipment control interface module to control the construction equipment to complete milling, heating and dispersion, hot mixing, cold paving, and compaction. To ensure feasibility, this embodiment provides a default command interaction mechanism: the end-side control unit sets a command sequence number and confirmation mechanism for each equipment control command, namely, "send command → wait for equipment confirmation → resend after timeout → maximum number of resends → trigger interlock or shutdown protection if it fails." Equipment confirmation can come from equipment PLC feedback or equipment communication module acknowledgments, and the confirmation content includes at least the command sequence number, execution status, and key feedback quantities (such as real-time discharge temperature, real-time paving speed, etc.). The end-side control unit writes "command sequence number - confirmation status - timestamp - execution value - feedback value" into the construction record for subsequent traceability.

[0150] To ensure process safety and data consistency, the end-side control unit is equipped with interlocking and exception handling logic. The default interlocks include: when the measured discharge temperature is lower than the lower limit of the discharge temperature boundary, paving start is prohibited and a prompt to increase the discharge temperature or adjust the organization is given; when the paving equipment detects insufficient material supply (e.g., the auger load is too low or the temperature array continuous coverage signal is missing), the compaction equipment is prohibited from starting; when the compaction density detector has no data or exceeds the range, the batch is marked as a measurement abnormal batch and a conservative compaction strategy is initiated, while the fault code and timestamp are recorded in the construction record; when a communication abnormality causes the inability to obtain the latest parameter package, the chain break continuous control strategy is initiated (see step S5).

[0151] Step S5 "Calculate the process event delay at the end and select the paving speed within the paving speed boundary, and perform chain break continuity control and retransmission":

[0152] The end-side control unit calculates the time delay from material discharge completion to paving start based on the process event timestamp. The material discharge completion event timestamp is the same as the material discharge completion event timestamp. This indicates that the paving start event timestamp is used. The timestamp for the material discharge completion event is triggered by two conditions by default: the "material gate opening signal" and the "discharge temperature stabilization signal" at the discharge port of the mixing equipment. Specifically, when the material gate opens and the measured discharge temperature fluctuates less than a preset temperature fluctuation threshold within a certain number of consecutive sampling periods, the timestamp for the material discharge completion event is recorded. The timestamp for the paving start event is triggered by two conditions by default: the "movement start signal" and the "temperature array continuous coverage signal" at the paving equipment. Specifically, when the paving starts and the temperature array detects that the continuous material strip coverage has reached a preset length, the timestamp for the paving start event is recorded. The end-side control unit ensures that the event timestamps are aligned under the same clock reference through a time synchronization module, using PTP time synchronization by default and calibrating the time every 60 seconds.

[0153] The time delay from the completion of material discharge to the start of paving satisfies:

[0154]

[0155] For ease of understanding, an example is provided in this embodiment: timestamp of material discharge completion event ( =10:05:30, timestamp of the stall start event =10:23:10.

[0156] Step 1: Time difference = 17 minutes and 40 seconds.

[0157] Step 2: Difference in seconds = 17 × 60 + 40 = 1060 seconds.

[0158] Step 3: Minute difference = 1060 / 60 ≈ 17.6667 min.

[0159] therefore min.

[0160] The end-side control unit is based on the time delay from the completion of material discharge to the start of paving. With the maximum allowed latency At the paving speed boundary Select paving speed And satisfy:

[0161]

[0162] Among them, the paving control length Safety margin duration for paving operations corresponding to the batch of recycled materials To compensate for additional fluctuations, the lower limit constant of duration. This is to prevent unreasonable speeds caused by an excessively small denominator. The end-side control unit ensures the paving speed while satisfying this inequality constraint. The batch falls within the paving speed boundary; when the calculated lower limit of the speed exceeds the upper limit of the paving speed boundary, the end-side control unit marks the batch as "out-of-limit batch", triggers an interlock prompt to increase the discharge temperature boundary or adjust the paving cycle, and reports the reason for the over-limit to the cloud-side control platform for re-solution.

[0163] For ease of understanding, an example is provided in this embodiment: Let the paving control length be... m, maximum allowable latency min, safety margin duration min, lower limit constant for duration =3min.

[0164] Step 1: .

[0165] Step 2: .

[0166] Step 3: .

[0167] therefore .

[0168] When communication between the edge control unit and the cloud control platform is interrupted, the edge control unit activates the continuity control module. The default behavior of the continuity control module includes: continuing to use the most recently validated parameter boundary set and proof data to select execution values ​​within the boundaries for continuous control of the construction equipment; writing all data to be uploaded to a local persistent queue; recording the start timestamp of the continuity interruption; sending heartbeat probes at fixed intervals; and recording the end timestamp of the continuity interruption and starting retransmission after communication is restored. The retransmission field list includes at least: old material batch identifier, parameter version identifier, and material discharge completion event timestamp. timestamp of the stall start event The data includes the following: discharge temperature measurement, paving start temperature measurement, compaction density measurement, workability index measurement, residual vector, and deviation judgment result. The default upload order is: deviation information packet → batch characterization data packet → construction record summary packet, to prioritize the restoration of closed-loop update capability.

[0169] The end-side control unit generates construction records corresponding to the old material batch identifier. The construction records include at least parameter version identifiers, process event timestamp sequences, execution value sequences, and residual vector sequences, and are stored in association with construction section identifiers.

[0170] The end-side control unit collects construction measurement values ​​to form a residual vector. To enable implementation by those skilled in the art, this embodiment provides a default measurement implementation:

[0171] Temperature measurement value An infrared thermometer array is deployed in the paving area, with a default sampling frequency of 2Hz. The end-side control unit sets a measurement window near the paving start event timestamp, for example, a ±10s window centered on the paving start event timestamp. The median of the output from the thermometer array within this window is taken as the paving start temperature measurement value, and this value is used as the temperature measurement value. If the temperature measurement array is obstructed or the measurement is abnormal, the end-side control unit will remove the abnormal sample and replace it with the measurement value of the backup point infrared thermometer; if the backup thermometer is also abnormal, the batch will be marked as a batch with abnormal temperature measurement and the conservative strategy will be triggered, and the fault code will be recorded.

[0172] Compacted density measurement value After compaction, density measurements are collected using a density detector, which can be a nuclear density meter or a non-nuclear density sensor. The default sampling method is: the paving control length corresponding to each batch of recycled material. A fixed number of density sampling points are set within the range, for example, one sampling point every 30m. The end-side control unit takes the average density of the sampling points in this batch as the compaction density measurement value. The maximum and minimum values ​​are recorded as fluctuation indicators. If the measured value of a sampling point exceeds the equipment range or is missing, the sampling point is removed and an anomaly is recorded in the construction record; if the number of valid sampling points is lower than the preset lower limit, the batch is marked as an abnormal density measurement batch and a conservative strategy is triggered.

[0173] Performance index measurement values The default implementation uses the torque sensor output of the mixing equipment. The mixing torque measurement value is the measured value of the mixing equipment. The default sampling frequency is 10Hz. The end-side control unit sets a sampling window during the mixing stabilization stage, such as a 30-second window before the discharge completion event timestamp. It performs a moving average filter on the torque measurement sequence within the window (window length, for example, 50 samples) and takes the average as the mixing torque measurement value for that batch. The end-side control unit normalizes the mixing torque measurement values ​​to construct workability index measurement values: the torque reference adopts the torque reference. This indicates that the average torque value of recent qualified batches or the equipment calibration value is used by default; the end-side control unit first calculates the normalized torque value. This, along with the temperature difference normalization term, forms the fields required for predicting performance indicators and calculating residuals. To ensure that the measured and predicted performance indicators are on the same scale, this embodiment uses the measured performance indicator values... Defined as "by and The value obtained after consistent normalization structure calculation means that the edge control unit uses the same normalization parameters as the cloud side. Calculated and will and The difference is taken as the working residual. If the torque sensor fails, the end-side control unit uses the paving auger load signal as a substitute measurement source, and the substitute source is marked in the construction record to ensure data traceability.

[0174] The residual vector is the residual vector ,satisfy:

[0175]

[0176] Among them temperature residual Density residual Working residuals satisfy:

[0177]

[0178] Among the temperature measurements Compacted density measurement value Compared with work performance index measurement value Measurements from the end side; predicted paving start temperature Predicted compaction density Compared with the predicted value of work performance indicators It comes from a cloud-based prediction model and is used consistently across versions.

[0179] The end-side control unit performs deviation detection, and the deviation detection conditions are:

[0180]

[0181] The residual threshold is the residual threshold. The number of consecutive establishments is the number of consecutive establishments. Note: Residual threshold With regularization coefficient Residual threshold for different technical quantities Used for end-side deviation determination, regularization coefficient Regularized stabilizing terms used for solving linear equations on the cloud side.

[0182] When the deviation judgment condition is met, the end-side control unit sends a deviation information packet to the cloud-side control platform. The deviation information packet includes at least: old material batch identifier, parameter version identifier, and residual vector. Time stamp of material discharge completion event timestamp of the stall start event and measurement summary fields related to deviation (e.g.) , , ).

[0183] For ease of understanding, an example is provided in this embodiment: Let the temperature measurement value be... Predicted temperature at the start of paving Compacted density measurement value Predicted compaction density Work performance index measurement values Predicted values ​​of work performance indicators .

[0184] Step 1: .

[0185] Step 2: .

[0186] Step 3: .

[0187] Step 4: Sum of Squares .

[0188] Step 5: .

[0189] For the end-side control unit to perform deviation determination, a residual threshold is set. Number of consecutive establishments If the L2 norm of the residual vectors from two consecutive samplings is 1.2453 and 1.1030 respectively, then both samplings satisfy the condition... If the condition is true twice consecutively, a deviation judgment is triggered; if the second value is 0.95, the consecutive condition is broken and the deviation judgment is not triggered.

[0190] After receiving the residual vector, the cloud-based control platform calculates the sensitivity matrix and determines the parameter subset. The sensitivity matrix is... ,satisfy:

[0191]

[0192] Where the output vector satisfy:

[0193]

[0194] The parameter vector to be updated is the parameter vector to be updated. The cloud-based control platform can, according to The parameter subset is selected based on the largest contribution or component size, and targeted incremental recalculation is performed only on this subset.

[0195] For ease of understanding, an example will be provided in this embodiment: Let the residual vector be... Sensitivity matrix for:

[0196]

[0197] but:

[0198]

[0199] calculate:

[0200]

[0201]

[0202]

[0203] therefore The first component is the largest, and the corresponding parameter item contributes the most, so it can be given priority to be included in the parameter subset.

[0204] Targeted incremental recalculation of parameter increment The parameter increments are obtained by solving the linear equations:

[0205]

[0206] Where the regularization coefficient is the regularization coefficient. The identity matrix is ​​the identity matrix. .

[0207] For ease of understanding, an example is provided in this embodiment: following example G. and And take the regularization coefficient. =0.05.

[0208] Step 1: From example G, we get .

[0209] Step 2: Calculation:

[0210]

[0211] Step 3: Construction:

[0212]

[0213] make Solve the system of equations:

[0214]

[0215] The result obtained by Gaussian elimination is:

[0216]

[0217] therefore:

[0218]

[0219] The cloud-based control platform generates a new parameter version identifier and distributes it in the form of a differential parameter package. The differential parameter package includes at least the new version identifier, the previous version identifier, the set of changed parameter items, the applicable batch range, the applicable construction section range, and a verification summary. After the end-side control unit verifies the package, it performs a partial replacement update on the parameter boundary set. To ensure the correctness of the differential package merging, the end-side control unit also performs a version chain check, confirming that the previous version identifier carried in the differential parameter package matches the current local version identifier; otherwise, it rejects the differential package and requests the cloud to reissue the full parameter package.

[0220] The end-side control unit sets rollback conditions, and the rollback conditions are met:

[0221]

[0222] The updated residual vector is the updated residual vector. The residual vector before the update is the same as the residual vector before the update. The residual increment threshold is the residual increment threshold. When the edge control unit meets the rollback conditions, it rolls back to the previous parameter version identifier and uploads a rollback notification. After receiving the rollback notification, the cloud-side control platform freezes the version chain before rollback and regenerates subsequent differential parameter packages based on the version after rollback to ensure version chain consistency.

[0223] For ease of understanding, an example is provided in this embodiment: Let... all over, Set the number of compaction passes. , , .

[0224] Step 1: .

[0225] Step 2: .

[0226] Step 3: .

[0227] Step 4: .

[0228] It should be noted that this embodiment sets up an implementation group and a control group. The control group adopts a regeneration integrated control method mainly based on local automatic control of equipment, lacking the constraint solving, residual directional update, and differential rollback mechanism of end-to-cloud collaboration; the implementation group adopts the end-to-cloud collaborative closed-loop method of this embodiment. Both groups are implemented in adjacent construction sections, with consistent sources of recycled materials, consistent equipment models and personnel configurations, and similar weather conditions during the construction period. The statistical caliber is: based on recycled material batches, with 30 batches; the statistical caliber for the paving start temperature compliance rate is " Not lower than the lower limit temperature for cold laying workability The percentage of batches meeting the compaction density standard; the statistical caliber for the compaction density compliance rate is " The percentage of batches that meet or exceed the target density threshold; the statistical caliber for the number of delay exceedances is " Exceed The number of occurrences of ""; the number of manual interventions refers to the number of times equipment settings are manually modified or execution values ​​are forcibly rewritten on-site; the number of rework / reinforcement times refers to the number of times rework or reinforcement is carried out due to substandard temperature or density.

[0229] Comparison of Key Quality and Organizational Indicators (Statistics of 30 Batches) Table 1

[0230] index Comparison group (equipment locally controlled) Implementation Team (End-to-Cloud Collaborative Closed-Loop) The paving start temperature meets the standard (not lower than the lower limit temperature for cold paving workability). 86.2% 97.8% Compaction density compliance rate (not lower than the target threshold) 88.5% 96.9% Number of times latency exceeded the limit (exceeding the allowed latency limit) 14 times 3 times Number of times manual intervention is required for construction parameters 11 times 2 times Number of rework / reinforcement attempts (due to temperature or density deviations) 6 times 1 time

[0231] Table 1 presents the comparative results of key quality and organizational indicators between the implementation group and the control group under the statistical scope of 30 batches of recycled materials. The implementation group adopted an end-to-cloud collaborative closed-loop approach, while the control group primarily used local equipment control. In Table 1, "Paving Start Temperature Compliance Rate" represents the percentage of batches whose measured paving start temperature is not lower than the lower limit temperature for cold paving workability, reflecting the cold paving workability assurance capability; "Compacted Density Compliance Rate" represents the percentage of batches whose measured compacted density is not lower than the target density threshold, reflecting the compaction quality assurance capability; "Number of Time Delay Exceeding Limits" represents the number of times the time delay from material discharge to paving start exceeds the allowable time delay limit, reflecting organizational fault tolerance and scheduling capabilities; "Number of Manual Interventions in Construction Parameters" represents the number of times equipment settings were manually modified or execution values ​​were forcibly rewritten, reflecting the degree of automation and closed-loop control; and "Number of Rework / Reinforcement" represents the number of rework or reinforcement measures required due to temperature or density deviations, reflecting the impact on quality stability and construction costs. Table 1 is used to visually demonstrate the improvements in quality compliance and organizational efficiency of the edge-cloud collaborative closed loop from the perspective of engineering results.

[0232] Table 2 Comparison of Stability and Closed-Loop Capability (Statistics of 30 Batches)

[0233] Experimental Data Comparison Table

[0234] index control group Implementation Group Mean of residual vector with L2 norm (normalized) 1.00 0.62 Number of deviation detection triggers (consecutive occurrences) 9 times 4 times Differential parameter packet distribution times 0 times 5 times Rollback trigger count 0 times 1 time Cumulative duration of communication interruption (measured on-site) 17min 18min Number of construction interruptions during chain break 2 times 0 times

[0235] Table 2 presents the comparison results of the implementation group and the control group in terms of stability and closed-loop capability under the same statistical scope of 30 batches of used materials. In Table 2, the "mean value of the L2 norm of the residual vector" represents the average level of the comprehensive deviation of temperature, density, and workability; the smaller the value, the more stable the process. The "number of deviation judgment triggers" represents the number of triggers that meet the residual threshold and are continuously valid; the smaller the value, the less system fluctuation or the better the model fit. The "number of differential parameter package distributions" represents the number of times the cloud-side control platform triggers directional incremental recalculation and differential updates, reflecting the actual operating frequency of the closed-loop update mechanism. The "number of rollback triggers" represents the number of times rollback is triggered when the deviation worsens after the update, reflecting the role of the update safety mechanism. The "cumulative duration of communication interruption" represents the background of network fluctuations on site. The "number of construction interruptions during link loss" represents whether construction is forced to stop under communication interruption conditions; the smaller the value, the stronger the end-side continuous control capability during link loss. Table 2 is used to demonstrate the stability and traceability of the end-cloud collaborative closed loop from the perspective of control process and system capability.

[0236] From Table 1 and Table 2 and Figure 3 , Figure 4 , Figures 6-8 It is evident that the implementation team improved the compliance rates of paving start temperature and compaction density, reduced the number of times time delay exceeded limits, the number of times manual intervention was required, and the number of times rework was required. Furthermore, the team was able to maintain continuous control within the boundary even when communication was interrupted and retransmit data after the communication was restored, thereby improving construction stability and traceability.

[0237] It should be noted that the embodiments of the present invention have better implementability and are not intended to limit the present invention in any way. Any person skilled in the art may use the above-disclosed technical content to change or modify it into equivalent effective embodiments. However, any modifications or equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of the technical solution of the present invention.

Claims

1. A cloud-based collaborative construction method for integrated hot-mix and cold-laying of recycled materials for municipal roads, characterized in that: The method is executed collaboratively by a cloud-based control platform and an end-side control unit. The end-side control unit connects to and controls the construction equipment, which includes at least milling equipment, heating and dispersing equipment, mixing equipment, paving equipment, and compaction equipment. The method includes: S1, the end-side control unit generates batch identifiers for municipal road waste materials entering the recycling process, collects batch characterization data corresponding to the batch identifiers, and uploads the batch characterization data and construction section identifiers to the cloud-side control platform. S2, the cloud-based control platform constructs a prediction model based on the batch characterization data, and performs constraint solving under the condition that at least cold-laying workability constraints and compaction quality constraints are met, to obtain parameter boundary set and constraint proof data, and generates parameter version identifier; S3, the cloud-side control platform sends the parameter boundary set, the constraint proof data, the parameter version identifier, and the validity conditions to the end-side control unit. After the end-side control unit verifies the data, it binds and stores the old material batch identifier with the parameter version identifier. S4, the end-side control unit selects execution values ​​within the parameter boundary set to control the construction equipment to complete milling, heating and dispersing, hot mixing, cold paving and compaction, and collects construction measurement values ​​to form a residual vector; S5, when the residual vector meets the deviation judgment condition, the end-side control unit sends the residual vector to the cloud-side control platform; the cloud-side control platform determines the parameter subset that needs to be updated based on the residual vector and the sensitivity matrix and performs directional incremental recalculation to generate a new parameter version identifier, and sends it out in the form of a differential parameter package; the end-side control unit updates the parameter boundary set according to the differential parameter package and rolls back to the previous parameter version identifier when the rollback condition is met.

2. The method according to claim 1, characterized in that, The parameter boundary set includes at least the discharge temperature boundary, the paving speed boundary, and the number of compaction passes boundary; the constraint proof data includes at least the predicted paving start temperature and the upper limit of the allowable delay. The validity conditions include at least the batch characterization data falling within a preset boundary range; the prediction model is used to calculate the predicted paving start temperature and the upper limit of the allowable delay.

3. The method according to claim 2, characterized in that, The prediction model includes at least the components for obtaining the predicted paving start temperature. Temperature decay prediction model, wherein the predicted paving start temperature value It satisfies the exponential temperature decay form.

4. The method according to claim 3, characterized in that, The predicted temperature at the start of paving satisfy: in, For discharge temperature, For ambient temperature, The time delay between the completion of material discharge and the start of paving. The temperature decay coefficient is It is an exponential function.

5. The method according to claim 4, characterized in that, The allowed delay limit Lower limit temperature of cold-laid workability The reverse calculation yields the result, which satisfies the following: in, It is the natural logarithmic function, and This is the preset lower limit temperature for cold-laying workability.

6. The method according to claim 5, characterized in that, The end-side control unit calculates the time delay from the completion of material discharge to the start of paving based on the process event timestamp. And satisfy: in, This is the timestamp for the completion of material output. The timestamp for the start of the stall event.

7. The method according to claim 6, characterized in that, The paving speed boundary is The end-side control unit is based on the time delay from the completion of material discharge to the start of paving. With the stated allowable delay limit Select paving speed within the paving speed boundary ,satisfy: in, To control the length of the paving, To allow for a safety margin of duration, This is the lower limit constant for duration.

8. The method according to claim 6, characterized in that, When communication between the end-side control unit and the cloud-side control platform is interrupted, the end-side control unit selects execution values ​​within the parameter boundary set to continuously control the construction equipment. After communication is restored, it transmits at least the process event timestamps, including the material discharge completion event timestamp and the paving start event timestamp, as well as the residual vector, to the cloud-side control platform. The end-side control unit generates a construction record corresponding to the old material batch identifier. The construction record includes at least a parameter version identifier, a process event timestamp sequence, an execution value sequence, and a residual vector sequence, and is stored in association with the construction section identifier.

9. The method according to claim 1, characterized in that, The residual vector is ,in: in, This is a temperature measurement value. This is the measured compaction density. These are workability index measurements; the predicted paving start temperature. The predicted compaction density value and the predicted values ​​of the workability indicators All were calculated by the prediction model described in claim 2; This indicates transpose.

10. The method according to claim 9, characterized in that, The deviation determination criteria include a residual threshold and a continuity determination, and satisfy the following: in, It is a norm 2. The residual threshold, It is an integer greater than or equal to 2.

Images