Mixing station mixture transportation carbon emission real-time correction method

Abstract

The invention relates to the technical field of carbon emission analysis, in particular to a mixing station mixture transportation carbon emission real-time correction method. The method comprises the following steps: carrying out transport vehicle multi-source data acquisition and synchronous analysis processing on a mixture transport vehicle by using a multi-source sensor built in the transport vehicle to generate synchronous transport vehicle multi-source data; obtaining the parameters of the mixture of the transportation mixing station; establishing a transportation stirring barrel carbon emission prediction model based on the transportation mixing station mixture parameters and the synchronous transportation vehicle multi-source data, and analyzing the real-time carbon emission data of the transportation stirring barrel through the transportation stirring barrel carbon emission prediction model; carrying out real-time carbon emission analysis processing on the traveling of the transport vehicle through the synchronous transport vehicle multi-source data, and generating real-time carbon emission data of the transport vehicle; and the real-time correction operation of the mixture transportation carbon emission of the mixing station is executed through the real-time carbon emission data of the transportation stirring barrel and the real-time carbon emission data of the transportation vehicle. According to the invention, accurate carbon emission correction in mixture transportation is realized.

Description

Real-time correction method for carbon emissions from mixing plant material transportation Technical Field

[0001] This invention relates to the field of carbon emission analysis technology, and in particular to a method for real-time correction of carbon emissions from the transportation of mixed materials at a mixing plant. Background Technology

[0002] With the acceleration of urbanization and the continuous expansion of infrastructure construction, mixing plants are widely used in road engineering, bridge engineering, and various construction projects. The mixtures produced by these plants need to be transported long or short distances by specialized vehicles. During this process, factors such as fuel consumption of the transport vehicles, insulation of the mixing drums, auxiliary mixing operations, and road traffic conditions all directly or indirectly generate significant carbon emissions. Currently, carbon emission accounting and control have become a crucial focus in the engineering construction field, requiring accurate calculation and real-time correction of carbon emissions during transportation while ensuring construction quality and progress. However, existing methods for real-time correction of carbon emissions from mixing plant transport typically rely on statistical models of fuel consumption or empirical formulas based on vehicle mileage and operating conditions. These methods neglect the impact of environmental factors, road conditions, and changes in the mixture's own state during transportation on energy consumption and carbon emissions, leading to significant discrepancies between calculated and actual carbon emissions. Furthermore, the lack of detailed modeling of the operating characteristics of the mixing drums in transport vehicles fails to reflect the dynamic impact of the mixing drums under different operating modes on overall energy consumption and carbon emissions. Summary of the Invention

[0003] Based on this, the present invention provides a method for real-time correction of carbon emissions during the transportation of mixed materials at a mixing plant, in order to solve at least one of the above-mentioned technical problems.

[0004] To achieve the above objectives, a method for real-time correction of carbon emissions from the transportation of mixed materials at a mixing plant includes the following steps: Step S1: Utilizing multi-source sensors built into the transport vehicle to collect and synchronously analyze multi-source data of the mixed material transport vehicle, generating synchronous multi-source data of the transport vehicle, wherein the synchronous multi-source data of the transport vehicle includes environmental perception data, transport road data, transport vehicle operation data, and auxiliary operation data of the transport vehicle's mixing drum; Step S2: Obtaining parameters of the mixed materials at the mixing plant; Based on the environmental perception data, the auxiliary operation data of the transport vehicle's mixing drum, and the parameters of the mixed materials at the mixing plant, establishing a mapping relationship between the operating characteristics and carbon emissions of the transport mixing drum, generating a carbon emission prediction model for the transport mixing drum; Step S3: Real-time correction of carbon emissions from the environmental perception data... Real-time data analysis generates real-time environmental perception data; based on the parameters of the mixture at the transportation mixing plant and the auxiliary operation data of the mixing drum of the transport vehicle, real-time mixing drum control operation characteristic analysis is performed to generate real-time mixing drum control operation characteristic data; the real-time mixing drum control operation characteristic data and real-time environmental perception data are transmitted to the carbon emission prediction model of the transportation mixing drum for real-time carbon emission prediction processing to generate real-time carbon emission data of the transportation mixing drum; Step S4: based on the transportation road data and the transportation vehicle operation data, real-time carbon emission analysis processing of the transportation vehicle is performed to generate real-time carbon emission data of the transportation vehicle; Step S5: real-time correction operation of carbon emission of the mixture transportation at the mixing plant is performed using the real-time carbon emission data of the transportation mixing drum and the real-time carbon emission data of the transportation vehicle.

[0005] Furthermore, step S1 includes the following steps: using the multi-source sensors built into the transport vehicle to collect multi-source data of the mixed material transport vehicle, generating multi-source data of the transport vehicle; performing multi-source data time synchronization and preprocessing on the multi-source data of the transport vehicle to generate synchronized multi-source data of the transport vehicle.

[0006] Further, step S2 includes the following steps: Step S21: Obtain the parameters of the mixture at the transportation mixing plant; Step S22: Perform attribute characteristic analysis of the mixture at the transportation mixing plant based on the parameters of the mixture at the transportation mixing plant, and generate attribute characteristic data of the transportation mixture; Step S23: Perform structural characteristic analysis of the transportation mixture based on the attribute characteristic data of the transportation mixture, and generate structural characteristic data of the transportation mixture; Step S24: Perform global carbon emission characteristic analysis of the conventional state and structural changes of the transportation mixture based on the structural characteristic data of the transportation mixture, and generate global carbon emission characteristic data of the mixture structure; Step S25: Perform auxiliary operation data of the mixing drum of the transportation vehicle... Step S26: Based on environmental perception data and the specific data of the mixing tank's operating mode, perform energy consumption characteristic analysis of the mixing tank's operating mode to generate energy consumption characteristic data of the mixing tank's operating mode; Step S27: Based on the energy consumption characteristic data of the mixing tank's operating mode, perform carbon emission characteristic analysis of the mixing tank equipment to generate carbon emission characteristic data of the mixing tank equipment; Step S28: Based on the global carbon emission characteristic data of the mixture structure and the carbon emission characteristic data of the mixing tank equipment, establish the mapping relationship between the operating characteristics and carbon emission of the transport mixing tank, and generate a carbon emission prediction model for the transport mixing tank.

[0007] Further, step S24 includes the following steps: Step S241: Analyze the carbon emission characteristics of each mixture structure based on the transport mixture structure characteristic data to generate mixture structure carbon emission characteristic data; Step S242: Analyze the mixture structure change based on the transport mixture structure characteristic data to generate mixture structure change data; Step S243: Analyze the mixture structure change sensitivity characteristics based on the mixture structure change data to generate mixture structure change sensitivity characteristic data; Step S244: Analyze the mixture structure change time series characteristics based on the mixture structure change data to generate mixture structure change time series characteristic data; Step S245: Analyze the carbon emission trend characteristics of mixture structure change based on the mixture structure change sensitivity characteristic data and the mixture structure change time series characteristic data to generate mixture structure change carbon emission trend characteristic data; Step S246: Analyze the global carbon emission characteristics of the mixture structure based on the mixture structure carbon emission characteristic data and the mixture structure change carbon emission trend characteristic data to generate global mixture structure carbon emission characteristic data.

[0008] Further, step S26 includes the following steps: Step S261: Extract the corresponding auxiliary operating energy consumption data of the mixing tank through the specific data of the mixing tank operation mode, and perform preliminary energy consumption characteristic analysis of the mixing tank operation mode based on the specific data of the mixing tank operation mode and the corresponding auxiliary operating energy consumption data of the mixing tank to generate preliminary energy consumption characteristic data of the mixing tank operation mode; Step S262: Perform unsteady-state heat flux density analysis on the structural characteristic data of the transported mixture to generate unsteady-state heat flux density data of the mixture; Step S263: Based on the environmental perception data and the unsteady-state heat flux density data of the mixture, perform environmental impact parameter adjustment processing on the preliminary energy consumption characteristic data of the mixing tank operation mode to generate energy consumption characteristic data of the mixing tank operation mode.

[0009] Furthermore, step S263 includes the following steps: designing environmental perception impact quantification parameters based on environmental perception data; performing environmental perception impact benefit quantification analysis on the unsteady heat flux density data of the mixture based on the environmental perception impact quantification parameters, generating environmental perception impact benefit quantification data on the mixture's thermal effect; designing environmental impact factors for the energy consumption of the mixing drum operation mode based on the environmental perception impact benefit quantification data on the mixture's thermal effect, generating environmental impact factors for the energy consumption of the mixing drum operation mode; and adjusting the environmental impact parameters of the initial energy consumption characteristic data of the mixing drum operation mode using the environmental impact factors for the energy consumption of the mixing drum operation mode, generating energy consumption characteristic data of the mixing drum operation mode.

[0010] Further, step S28 includes the following steps: Step S281: Using the specific data of the mixing tank operation mode corresponding to the carbon emission characteristic data of the mixing tank equipment, perform cross-feature analysis of the transportation mixing tank operation on the global carbon emission characteristic data of the mixture structure corresponding to the sensitivity characteristic data of the mixture structure change and the temporal characteristic data of the mixture structure change, to generate cross-feature data of the transportation mixing tank operation; Step S282: Using the cross-feature data of the transportation mixing tank operation, perform full-connected fusion feature processing on the global carbon emission characteristic data of the mixture structure and the carbon emission characteristic data of the mixing tank equipment, to generate full-connected fusion feature data of the transportation mixing tank carbon emission; Step S283: Using a preset multi-scale convolution kernel, perform carbon emission fusion feature analysis on the full-connected fusion feature data of the transportation mixing tank carbon emission based on the differences in the mixing tank operation scale, to generate multi-scale fusion feature data of the transportation mixing tank operation carbon emission; Step S284: Based on the multi-scale fusion feature data of the transportation mixing tank operation carbon emission, establish the mapping relationship between the operation characteristics and carbon emission of the transportation mixing tank, and generate a carbon emission prediction model for the transportation mixing tank.

[0011] Furthermore, step S3 includes the following steps: Step S31: Analyze the real-time transportation demand of the mixture based on the parameters of the mixing plant to generate real-time transportation demand data; perform real-time data analysis on the environmental perception data to generate real-time environmental perception data; Step S32: Analyze the real-time mixing drum control operation characteristics based on the real-time transportation demand data of the mixture and the auxiliary operation data of the mixing drum of the transport vehicle to generate real-time mixing drum control operation characteristic data; Step S33: Transmit the real-time mixing drum control operation characteristic data and the real-time environmental perception data to the carbon emission prediction model of the transport mixing drum for real-time carbon emission prediction processing to generate real-time carbon emission data of the transport mixing drum.

[0012] Furthermore, step S4 includes the following steps: Step S41: Perform vehicle operation status sequence analysis on the transport vehicle operation data to generate vehicle operation status sequence data; Step S42: Perform cumulative energy consumption feature analysis on the operation status sequence based on the vehicle operation status sequence data to generate cumulative energy consumption feature data of the operation status sequence; Step S43: Perform energy consumption cumulative feature correction processing on the cumulative energy consumption feature data of the operation status sequence based on the parameters of the mixture at the transport mixing plant and the transport road data to generate cumulative energy consumption feature data of the operation condition correction; Step S44: Perform real-time carbon emission analysis processing on the transport vehicle's operation based on the cumulative energy consumption feature data of the operation condition correction to generate real-time carbon emission data of the transport vehicle.

[0013] Furthermore, step S43 includes the following steps: analyzing the impact characteristics of vehicle transportation based on the parameters of the mixed material at the transportation mixing plant and the transportation road data, and generating vehicle transportation impact condition characteristic data; analyzing the driving behavior characteristics of each vehicle transportation condition based on the vehicle transportation impact condition characteristic data and the vehicle operation state sequence data, and generating vehicle transportation condition driving behavior characteristic data; and performing condition difference correction processing on the cumulative energy consumption characteristic data of the operation state sequence based on the vehicle transportation condition driving behavior characteristic data and the corresponding vehicle transportation impact condition characteristic data, and generating operation condition corrected cumulative energy consumption characteristic data.

[0014] The beneficial effects of this application are that the present invention uses multi-source sensors built into the vehicle to synchronously collect and preprocess data on environmental perception, road conditions, vehicle operating status, and auxiliary operation of the mixing drum, solving the problems of information loss and time misalignment caused by traditional single data sources or asynchronous data. The synchronous transport vehicle multi-source data, after time synchronization and preprocessing, has higher data integrity, spatiotemporal consistency, and availability, supporting accurate identification of instantaneous operating conditions (such as sudden road congestion, slope changes, or sudden temperature changes), and providing high-quality input for subsequent energy consumption assessment and carbon emission calculation. This facilitates real-time data quality verification, anomaly detection and compensation (improving robustness) at the edge or cloud, enhances the comparability and traceability of samples during model training, and significantly improves the timeliness, accuracy, and engineering applicability of subsequent carbon emission prediction and correction. Furthermore, it facilitates functions such as operation and maintenance alarms, historical backtracking, and compliance auditing, promoting the implementation of refined energy consumption management and emission reduction decisions in the transportation sector. By coupling the physical properties of the mixture with environmental perception and agitator-assisted operation data, this model systematically reveals the energy consumption-carbon emission mapping relationship between the mixture's structural characteristics, unsteady-state thermal behavior, and agitator operation modes. Through layer-by-layer analysis of mixture properties, structural change sensitivity, heat flux density characteristics, and agitator operation mode energy consumption characteristics, and employing cross-scale fusion (e.g., multi-scale convolution) to establish feature fusion and mapping relationships, a carbon emission prediction model can be constructed that reflects both the intrinsic differences of materials and adapts to fluctuations in the operating environment. This model can make more refined predictions of energy consumption and carbon emissions for mixtures with different proportions and structural states under different environments and operation modes, significantly reducing estimation bias caused by neglecting material-equipment interactions. It also supports online model updates and adaptive calibration, making real-time carbon emission predictions more reliable and interpretable, and providing decision-making basis for agitator control, insulation strategy adjustment, and transportation route / scheduling optimization based on the prediction results, thereby achieving refined carbon reduction and energy consumption optimization throughout the transportation process. By coupling environmental perception, mixture parameters, and real-time control characteristics of the mixing drum in the time dimension, the timeliness and accuracy of carbon emission prediction for the mixing drum are significantly improved. Real-time analysis of environmental perception data captures the direct impact of short-term abrupt changes in temperature, humidity, wind speed, and precipitation on the heat dissipation / insulation behavior of the mixture. Real-time control characteristics extracted from real-time transport needs of the mixture and auxiliary operating data of the mixing drum reflect the immediate contributions of mixing frequency, rotation speed, and heating / insulation actions to energy consumption. These two data points are fed into an established carbon emission prediction model for the mixing drum, enabling rapid and fine-grained predictions of energy consumption and carbon emissions under unsteady conditions. This can be used for online correction of carbon emission estimates and to drive adaptive control of the mixing drum (e.g., adjusting rotation speed and insulation strategies to reduce energy consumption). This achieves real-time compensation for material heat loss and equipment energy consumption, reduces estimation errors, and provides a reliable basis for abnormal operating condition alarms and immediate maintenance decisions.The ability to accurately calculate and calibrate the carbon emissions of transport vehicles during driving, with a refined and condition-aware approach. By analyzing the operating state sequence of vehicle operation data and accumulating energy consumption characteristics, driving behavior information such as instantaneous speed, acceleration / deceleration events, idling duration, slope, and gear usage is converted into quantifiable energy consumption accumulation characteristics. Then, by combining the correction of the influence of transport road data (such as road surface type, slope, congestion, signal distribution, etc.) and mixture parameters on the working conditions, the systematic errors in the estimation can be eliminated, and more realistic vehicle real-time carbon emission data can be obtained. The real-time carbon emission data at the mixing drum end and the vehicle end are fused for online correction of the overall carbon emissions in the mixture transportation link of the mixing plant. By simultaneously considering the heat loss caused by the change in the material structure and the energy consumption of the mixing drum equipment, as well as the energy consumption during vehicle driving due to road conditions and driving behavior, the two types of real-time data are mutually verified and compensated, which can significantly reduce the deviation and uncertainty of single-sided estimation. Based on this correction result, instant carbon emission notification, refined scheduling, accurate carbon footprint attribution, and compliance declaration are achieved. At the same time, the corrected real-time data can also be used as high-quality labels for model adaptive update and historical retrospective analysis, forming a closed-loop optimization to improve the stability and interpretability of long-term prediction.

[0015] Therefore, the real-time correction method for the carbon emissions of the mixture transportation in the mixing plant of the present invention addresses the problem of carbon emission calculation deviation in the prior art, which only relies on fuel consumption statistical models or empirical formulas and ignores environmental factors, road conditions, and changes in the state of the mixture itself. A real-time carbon emission correction method that can integrate environmental perception data, road condition information, parameters of the mixture transported in the mixing plant, and operating characteristics of the mixing drum is proposed. By establishing a multi-dimensional data-driven prediction and correction mechanism, it can not only refine the modeling of the dynamic impact of energy consumption and carbon emissions of the mixing drum under different operating modes, but also achieve real-time and accurate correction of the overall carbon emissions of the vehicle under complex working conditions, thus significantly improving the accuracy and reliability of carbon emission calculation and providing more practical technical support for energy conservation and emission reduction in the transportation link of the mixing plant. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] FIG. 1 is a schematic diagram of the step flow of a real-time correction method for the carbon emissions of the mixture transportation in the mixing plant of the present invention; FIG. 2 is a detailed implementation step flow diagram of step S2 in FIG. 1; The realization, functional features, and advantages of the purpose of the present invention will be further described in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] The technical method of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those skilled in the art within the scope of the present invention without creative efforts belong to the scope of protection of the present invention.

[0018] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. Functional entities may be implemented in software, or in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods. The term "and / or" as used herein includes any and all combinations of one or more of the associated items listed.

[0019] To achieve the above objectives, please refer to Figures 1 and 2. This invention provides a method for real-time correction of carbon emissions during the transportation of mixed materials at a mixing plant. In an embodiment of this invention, please refer to Figure 1, which is a flowchart illustrating the steps of the method. The method includes the following steps: Step S1: Using multi-source sensors built into the transport vehicle, multi-source data of the mixed material transport vehicle is collected and synchronously analyzed to generate synchronous multi-source data of the transport vehicle. This synchronous multi-source data includes environmental perception data, transport road data, transport vehicle operation data, and auxiliary operation data of the mixing drum. In this embodiment, a multi-frequency GNSS receiver (sampling frequency 10Hz, RTK positioning accuracy) is installed at the front of the mixed material transport vehicle. Using a unified time reference of ≤0.02m, a low-noise triaxial inertial measurement unit (IMU, sampling frequency 200Hz, acceleration resolution ≤0.01m / s², angular velocity resolution ≤0.01° / s) is fixed on the vehicle chassis. The engine and transmission system collect engine speed, engine torque, oil pump pulse, and wheel speed signals via CAN bus (sampling frequency 50Hz). An angle encoder and torque sensor (sampling frequency 100Hz, range calibrated according to vehicle model) are installed on the shaft of the mixing drum. At least three thermocouples are arranged on the inner surface of the mixing drum for sampling the temperature of the mixture (sampling frequency 1Hz). An ambient temperature and relative humidity sensor and a three-cup anemometer (sampling frequency 1Hz) are installed on the top of the vehicle. An on-board imaging unit (resolution ≥2MP, frame rate ≥10fps) is installed on the side of the vehicle for road surface feature recognition. All sensor signals are aligned with GNSS pulses (PPS) and UTC timestamps. High-frequency signals are first denoised using a zero-phase second-order Butterworth low-pass filter (5Hz cutoff frequency), then isolated outliers are removed using a Hampshire filter (1s window width). Low-frequency environmental quantities are resampled to a uniform 1Hz time grid using linear interpolation. An extended Kalman filter is used to fuse GNSS and IMU data to solve for vehicle speed, attitude, and road slope, with time synchronization error controlled within ±10ms. After the above processing, environmental perception data (including instantaneous temperature, humidity, wind speed, and road image features), transportation road data (including road type identification, slope, and section speed limits), transportation vehicle operation data (including instantaneous vehicle speed, acceleration, engine torque, and fuel pulse), and mixer drum auxiliary operation data (including speed, torque curve, heating current, and internal temperature curve) are generated moment-by-moment along a uniform time axis. The data integrity ratio and quality score are calculated at each moment, and moments with quality below a threshold are marked for subsequent processing and anomaly alarms.

[0020] Step S2: Obtain the parameters of the transport mixing plant mixture; establish the mapping relationship between the operating characteristics and carbon emissions of the transport mixing drum based on environmental perception data, auxiliary operation data of the transport vehicle mixing drum, and the parameters of the transport mixing plant mixture, and generate a carbon emission prediction model for the transport mixing drum; in this embodiment of the invention, the parameters of the transport mixing plant mixture are obtained, including aggregate particle size distribution, asphalt or binder content (mass percentage), mineral powder content, moisture content, discharge temperature, mixture density, specific heat capacity and thermal conductivity of each component, etc., and are calibrated according to the formula sheet and discharge metering record to form a mixture parameter table. Based on the mixture parameters, attribute characteristic analysis is performed, and the equivalent specific heat capacity and equivalent thermal conductivity are calculated by weighting the component volume fraction, and the transport mixture attribute characteristic data are output; further, the structural characteristics of the mixture are derived based on the particle size distribution and porosity model, and the structural characteristic data of the transport mixture are obtained by using the particle packing model and the porosity-thermal conduction coupling relationship. A global carbon emission characteristic analysis was conducted on the mixture's structural properties under both normal and structural variations. Under normal conditions, the first-order energy conservation equation and the time integral of the product of stirring friction torque and angular velocity were used to measure the heat loss per unit mass. For structural variations, an unsteady-state heat flux density analysis was employed, and the radial or layered temperature field within the mixture was discretized using the finite difference method, with the time step satisfying the explicit scheme stability condition. This generated the mixture's sensitivity and temporal characteristics to structural variations. At the mixing drum end, mode labels were extracted through operation mode-specific analysis: the operation mode was identified based on the torque-speed curve, heater current waveform, and insulation duty cycle, and the average mechanical power and heating power under each mode were calculated to obtain preliminary energy consumption characteristic data for the mixing drum's operation mode. Subsequently, environmental perception data and the mixture's unsteady-state heat flux density were used for environmental impact factor design. The environmental impact factor, as a function of wind speed, ambient temperature, and humidity, was used to adjust the parameters of the preliminary energy consumption characteristics, outputting the final energy consumption characteristic data for the mixing drum's operation mode. Based on the global carbon emission characteristics of the mixture structure and the carbon emission characteristics of the mixing tank equipment, operational cross-feature analysis is performed to generate a fully connected fusion feature matrix. Multi-scale convolutional kernels are used to extract carbon emission features corresponding to mixture changes at short-term, medium-term, and long-term scales. Then, a carbon emission prediction function is constructed through several layers of nonlinear mapping and regression. During the model training phase, least squares fitting is performed using historical calibration samples with L2 regularization. After training, the model coefficients, feature standardization parameters, and multi-scale convolutional kernel weights are saved for real-time inference.

[0021] Step S3: Perform real-time data analysis on the environmental perception data to generate real-time environmental perception data; perform real-time mixing drum control operation characteristic analysis based on the mixing parameters of the transportation mixing plant and the auxiliary operation data of the mixing drum of the transportation vehicle to generate real-time mixing drum control operation characteristic data; transmit the real-time mixing drum control operation characteristic data and the real-time environmental perception data to the transportation mixing drum carbon emission prediction model for real-time carbon emission prediction processing of the transportation mixing drum to generate real-time carbon emission data of the transportation mixing drum; in this embodiment of the invention, the environmental perception signal is first subjected to two-stage denoising: the low-frequency trend is extracted by a 600-second sliding median filter, the short-term disturbance is smoothed by a 60-second sliding median filter, and the high-frequency noise is filtered out by a 0.5Hz low-pass filter; calculate the statistics (mean, standard deviation, linear trend) based on the 60-second and 600-second windows to form environmental window features, and extract the environmental perception data at the current moment accordingly to obtain real-time environmental perception data. The real-time transportation demand analysis of the mixed material uses the target outbound temperature, allowable temperature difference, and expected transportation time as inputs. Combined with the current temperature curve of the mixed material, a first-order predictor is used to predict the temperature decay curve for the next 30 minutes and calculate the heat loss rate per unit time, outputting real-time transportation demand data for the mixed material. The auxiliary operation signal of the mixing drum calculates the instantaneous mechanical power at a 2-second resolution and performs pattern matching according to a preset operation mode dictionary. The mode dictionary includes at least three categories: constant speed mixing, intermittent mixing, and continuous heating. The average rotational speed, average torque, heating duty cycle, and mode duration for each type are extracted as real-time mixing drum control operation characteristic data. The environmental window features, real-time transportation demand of the mixed material, and real-time mixing drum control operation characteristics are assembled into a standardized feature vector. Batch forward inference is performed using a sliding small window (window length 5 seconds, step size 1 second). The forward inference outputs real-time carbon emission data of the mixing drum per second based on the trained transportation mixing drum carbon emission prediction model. The inference process implements online monitoring of residuals: if the observed residuals exceed 30% or the absolute increment exceeds the preset carbon dioxide content threshold, the exponential weighted residual average method is used to correct the model bias term online. The correction learning rate is set to 0.01, and continuous outliers are labeled and recorded as offline backtracking samples for subsequent recalibration.

[0022] Step S4: Based on transportation road data and transportation vehicle operation data, perform real-time carbon emission analysis and processing of transportation vehicle operation to generate real-time carbon emission data of transportation vehicles. In this embodiment of the invention, the time-series signal is divided into states, and the driving process is divided into acceleration, deceleration, idling, and cruising segments according to the instantaneous longitudinal acceleration threshold. Idle speed is determined when the vehicle speed is below 1 km / h and the engine speed is in the idle range. For each state segment, the instantaneous traction power is calculated according to the physical power model. The traction power consists of three parts: rolling resistance, slope gravity component, and air resistance. Rolling resistance is obtained by multiplying the rolling friction coefficient by the total mass of the vehicle and the gravitational acceleration, and then by the vehicle speed. The slope component is obtained by multiplying the vehicle mass, gravitational acceleration, and slope angle by the vehicle speed. Air resistance is obtained by multiplying the air density, drag coefficient, frontal area, and the cube of the vehicle speed, and then by a constant factor. Engine output power is obtained by multiplying engine torque by crankshaft angular velocity. Engine operating condition mapping uses a fuel consumption map (BSFC table) and bilinear interpolation to look up the instantaneous fuel consumption rate on the engine speed and torque coordinates. Instantaneous fuel mass flow rate is obtained by multiplying engine output power by the interpolated fuel consumption rate. The cumulative energy consumption is obtained by integrating the instantaneous fuel mass flow rate for each state segment, with the integration step size consistent with the synchronization time base to ensure time alignment. Operating condition corrections are performed to address differences in transportation conditions. A multiple linear regression model is constructed to output correction coefficients. The regression independent variables include the average slope of the road segment, load mass, average wind speed, and initial temperature of the mixture. The regression model uses ridge regression (L2 regularization) with the following parameters: regularization strength α = 0.1, feature standardization uses zero-mean unit variance transformation, and five-fold cross-validation is used to determine the stability of the regression coefficients. Regression residuals are monitored in real time using sliding window statistics. If the residuals exceed a set threshold, an offline recalibration process is triggered. The fuel mass flow rate, after correction, is converted into a carbon dioxide emission rate based on the fuel carbon emission factor. The conversion uses the physical relationship between fuel mass and fuel carbon content, multiplied by the molar mass ratio of carbon to carbon dioxide. The output is a real-time carbon emission time series of transport vehicles with a resolution of per second or per minute. To ensure data reliability, the time series signal undergoes anomaly detection before processing. Anomalies include signal loss, numerical abrupt changes, or physical inconsistencies. Anomaly samples are marked and recorded for subsequent quality audits and model retraining.

[0023] Step S5: Perform real-time correction of carbon emissions from the transportation of the mixing plant's aggregates using real-time carbon emission data from the transport mixing drum and the transport vehicles.

[0024] In this embodiment of the invention, when performing real-time correction of carbon emissions from the transportation of mixed materials at the mixing plant, the real-time carbon emission data of the mixing drum and the transportation vehicle are first aligned according to the same time slice to ensure a one-to-one correspondence between the two types of data in the time dimension. For the real-time carbon emission data of the mixing drum within each time slice, based on the difference between its data acquisition cycle and the vehicle's real-time carbon emission data acquisition cycle, time interpolation is used to match the data length, ensuring that the two types of data maintain the same number of data points within the same time slice. Real-time correction feedback is then performed based on the aligned real-time carbon emission data of the mixing drum and the transportation vehicle as the overall emissions from the transportation of mixed materials.

[0025] Furthermore, step S1 includes the following steps: using the multi-source sensors built into the transport vehicle to collect multi-source data of the mixed material transport vehicle, generating multi-source data of the transport vehicle; performing multi-source data time synchronization and preprocessing on the multi-source data of the transport vehicle to generate synchronized multi-source data of the transport vehicle.

[0026] In this embodiment of the invention, several sensing units are arranged on the transport vehicle according to functional domains. Their positions, installation methods, and electrical interfaces are determined according to vehicle structure and engineering electrical specifications. A high-precision GNSS receiving module is used for positioning and time reference. The module outputs a unified UTC timestamp and pulse synchronization signal, serving as a unified time reference for all vehicle data. Vehicle dynamics signals are acquired by a three-axis inertial measurement unit (IMU) mounted on the chassis. The IMU and GNSS share rack grounding and shielding measures to reduce electromagnetic interference. Vehicle power and engine operating condition signals are read through the onboard controller network bus. Reading items include engine speed, engine torque, fuel pulse or flow indication, wheel speed, etc. The electrical interface is consistent with the vehicle's CAN bus protocol. Signals related to the mixing tank consist of an axial encoder, torque sensor, internal temperature sensing array, and heating circuit current / voltage measurement device. Temperature points are distributed axially and circumferentially along the tank body to reflect the non-uniformity of the temperature field. Environmental quantities are collected by a roof-mounted meteorological sensor, including ambient temperature, relative humidity, and wind speed. Road surface characterization is obtained jointly by a vehicle-side imaging unit and a short-range laser ranging unit to extract road texture and pothole distribution. The raw values ​​output by each sensor are recorded as standardized sensor metadata items along with their inherent identifier, timestamp, unit of measurement, and calibration coefficient. To ensure measurement reliability, multiple calibration procedures are implemented: static zero-bias calibration, cross-temperature drift calibration, and range linearization calibration. Calibration configuration files and calibration certificates are saved along with the sensor metadata. At the data acquisition end, basic hardware filtering and anti-aliasing circuitry are implemented for the signal at the sampling link, and a circular buffer is maintained at the acquisition node to cope with transient communication interruptions. The acquisition process provides event logging and alarm output for hardware failures, communication frame drops, and sensor self-test anomalies. Anomaly logs include time windows, sensor IDs, and raw observation segments for subsequent traceability and maintenance. The power supply and grounding of all acquisition devices are deployed according to vehicle engineering specifications, and redundant power management and data caching mechanisms are used to avoid short-term data loss. GNSS pulse signals and UTC timestamps are used as the vehicle-wide synchronization reference, and all acquisition units are time-stamp aligned using this time reference. Anti-aliasing hardware pre-pass low-pass processing is implemented for high-frequency sensor signals. Subsequently, in the synchronization processing stage, signals with different sampling rates are resampled and aligned: high-frequency signals are processed by time-domain downsampling and window function low-pass filtering to maintain band-limited characteristics, while low-frequency signals are upsampled by spline interpolation or linear interpolation to match the global time grid. The timing synchronization logic includes a synchronization error estimation module, which estimates the time offset of each signal through cross-correlation analysis and pulse time difference measurement and applies correction. After correction, the verification step uses cross-correlation peak values ​​and time consistency checks.Multi-layer preprocessing is performed on the raw signal: noise suppression employs a robust statistical outlier removal method, first using median filtering to identify isolated abrupt changes, and then using window-based anomaly detection to remove short-pulse noise; continuity verification uses physical constraints to determine anomalies, such as instantaneous acceleration exceeding the vehicle structure's allowable range or temperature change rate exceeding the estimated material heat capacity range, which are then marked as anomalous segments. For CAN bus signals, frame integrity checks and physical consistency mapping are performed. Intra-frame fields are converted into physical quantities according to a defined signal decoding table, and dimensional consistency checks are performed on the decoded quantities. Attitude and position fusion uses an extended Kalman filter as the sensor fusion framework. The state vector includes position, velocity, attitude angle, and inertial sensor bias. Filter configuration parameters include the process noise covariance matrix structure and measurement noise covariance matrix structure, initial covariance matrix description, and observation update frequency setting. Observation updates use alternating correction of GNSS position and inertial measurements to eliminate short-term drift. Temperature sensing implements cold junction compensation and nonlinear correction. After alignment, the time-series data is used to calculate several basic derived quantities, such as instantaneous vehicle speed, longitudinal acceleration, road slope estimation, integral power of mixing torque, and instantaneous heating power. A data quality score is also calculated for each time point, composed of coverage indicators, sensor health indicators, and statistical stability indicators. The output synchronous multi-source transport vehicle data is indexed using a unified time grid and includes fields for environmental perception data, transport road data, transport vehicle operation data, and mixer drum auxiliary operation data. Data quality labels and anomaly annotations are also added to each time point for subsequent carbon emission modeling and online inference.

[0027] Further, as an embodiment of the present invention, referring to Figure 2, which is a detailed flowchart of step S2 in Figure 1, step S2 in this embodiment includes the following steps: Step S21: Obtaining the parameters of the mixture at the transportation mixing plant; In this embodiment of the present invention, the parameters of the mixture at the transportation mixing plant are obtained through factory metering records and laboratory property measurements. The factory metering records include the formula composition, the mass percentage of each component, the factory discharge temperature, and the batch number; the laboratory property measurements include particle size distribution determination, moisture content determination, mixture density determination, specific heat capacity determination of each component, and thermal conductivity determination. Particle size distribution is determined by sieving or laser particle size analyzer and a particle size distribution curve is generated; moisture content is quantified by oven drying at room temperature or Karl Fischer titration; specific heat capacity is determined by differential scanning calorimetry or adiabatic calorimetry and calibrated by a calibration curve; thermal conductivity is measured by steady-state or transient heat flow meter method and a temperature dependence curve is fitted. Each measurement result is annotated with the measurement uncertainty and calibration certificate number, and recorded in a standard format as a mixture parameter set. The parameter set fields include: formulation code, percentage of formulation components, factory temperature, moisture content, density, specific heat capacity curve, and thermal conductivity curve. To ensure parameter reliability, repeated measurements and intra-batch variation analysis are implemented. Abnormal results trigger a retest process and generate a quality control report. The final output mixture parameters are provided in a structured format as input for subsequent property analysis and model building. All parameters are accompanied by measurement method descriptions and measurement conditions to ensure the traceability of the property interpretation chain.

[0028] Step S22: Analyze the property characteristics of the mixture at the transportation mixing plant based on the parameters of the mixture, generating property characteristic data of the transportation mixture. In this embodiment of the invention, the component mass percentage is converted to volume fraction and the equivalent specific heat capacity is calculated using a mass or volume weighted method. The equivalent thermal conductivity is constructed based on a thermal resistance series-parallel model and the equivalent thermal conductivity coefficient is solved using component thermal conductivity and porosity parameters. The influence of moisture content is treated by separating adsorbed and free water. Free water is estimated using a latent heat loss model for additional heat loss during temperature decrease, while adsorbed water is included in the equivalent specific heat capacity correction term using a specific heat correction factor. Particle size distribution is quantified and mapped to estimated porosity values ​​using statistical indicators (median diameter, fine particle ratio, coarse particle ratio). The relationship between porosity and contact thermal resistance is obtained by mapping from experimental calibration curves, which are recorded as parameters in the attribute feature vector. The attribute feature vector consists of equivalent specific heat capacity, equivalent thermal conductivity, estimated porosity, moisture content decomposition term, and particle size statistical indicators. The attribute feature analysis process includes consistency verification and uncertainty propagation calculation. Uncertainty propagation uses a linear approximation method to transfer the measurement uncertainty of each physical property to the equivalent parameter and form a confidence interval. All output items are accompanied by uncertainty descriptions for subsequent structural characteristic analysis.

[0029] Step S23: Analyze the structural characteristics of the transport mixture based on its property data to generate structural characteristic data. In this embodiment, the structural characteristic analysis establishes a pore distribution and connectivity characterization based on a particle packing model. The contact probability between particles is derived using sieving results, and the thermal contact area ratio is calculated using a contact network model. The thermal contact area ratio and contact thermal resistance together define the effectiveness of the solid-solid heat conduction path. For multi-scale structural heterogeneity, a layered or radially discrete model is adopted. The equivalent thermal conductivity and specific heat capacity of each discrete unit describe the local heat capacity and heat conduction rate, and the interface thermal resistance term reflects the influence of interface thermal resistance. The structural characteristic analysis also includes a description of the particle-slurry phase interface behavior. The thermal resistance at the interface is determined by the interface wettability and interface contact pressure parameters. The interface parameters are obtained through laboratory compression and heat conduction tests and formed into a parameter table. The structural characteristic data output fields include: porosity distribution curve, connectivity coefficient, contact thermal resistance distribution, layered equivalent thermal conductivity distribution, local heat capacity distribution, etc., and are accompanied by mesh description and boundary condition suggestions for spatial discretization solution, so as to perform time-varying thermal field solution based on the structural characteristics or for subsequent unsteady-state sensitivity analysis.

[0030] Step S24: Based on the structural characteristic data of the transport mixture, perform global carbon emission characteristic analysis of the transport mixture under normal conditions and structural changes, generating global carbon emission characteristic data of the mixture structure. In this embodiment of the invention, the normal state analysis starts from the energy conservation equation, expressing the temperature evolution per unit mass of mixture in the form of heat capacity, convective heat transfer, and conduction terms; the convective heat transfer flux is given by the product of the surface heat transfer coefficient and the temperature difference, and the surface heat transfer coefficient is used as a parameterized expression of the relationship between the environmental function and the windward speed. The structural change analysis obtains the local heat flux abrupt change caused by changes in particle position or contact area through the solution of unsteady-state heat flux density, obtains the temperature time series response using the spatiotemporal discretization method, and calculates the unit time heat loss difference under several representative structural change scenarios. Sensitivity analysis is performed based on the multi-scenario solution results, and the contribution share of each structural parameter to heat loss is quantified by the variance decomposition method, outputting the sensitivity feature vector and the structural change time series feature sequence. The heat loss time series is then converted into additional energy consumption demand. The stirring friction power and heating energy consumption are combined and converted into the corresponding carbon emission time series. Finally, global carbon emission characteristic data of the mixture structure is generated. The fields include the normal state heat loss curve, the incremental carbon emission curve triggered by structural changes, the list of sensitive parameters and the corresponding contribution rate matrix.

[0031] Step S25: Based on the auxiliary operation data of the transport vehicle's mixing drum, perform a specific analysis of the mixing drum's operation mode to generate specific data for the mixing drum's operation mode. In this embodiment of the invention, an operation mode set is defined and identified using physical quantity signatures. The signature consists of shaft speed, output torque, heater current duty cycle, mixing cycle, and heat preservation duty cycle. Historical operation segments are clustered to identify common mode prototypes and representative statistical characteristics are calculated for each prototype, including average mechanical power spectrum, heating power spectrum, and mode duration distribution. The mode-specific data includes mode identifier, representative power spectrum, spectral energy distribution, typical duration distribution, and mode transition probability matrix. To reflect dynamic control behavior, the control input time series and response time series are further combined, and control response delay and overshoot characteristics are calculated. The analysis process implements anomaly labels for sudden non-standard operating behaviors, with the anomaly labels indicating the event start and end time, anomaly type, and possible fault indicators. The output mode-specific data serves as the mode index and power mapping basis for subsequent energy consumption characteristic analysis.

[0032] Step S26: Based on environmental perception data and specific data of the mixing tank operation mode, perform energy consumption characteristic analysis and processing of the mixing tank operation mode to generate energy consumption characteristic data of the mixing tank operation mode. In this embodiment of the invention, mechanical energy consumption and heating energy consumption are quantified separately: mechanical energy consumption is obtained by integrating torque and angular velocity to obtain the instantaneous mechanical power spectrum, and then integrated over the mode duration to obtain the mode mechanical energy consumption distribution; heating energy consumption is obtained by integrating current and voltage to obtain the heating power spectrum and allocated to the mode time period according to the heat preservation duty cycle. Environmental impact is applied to the surface heat transfer coefficient and convective heat dissipation term through parameterized functions, thereby changing the heat loss prediction and affecting the heating energy consumption compensation requirement. The environmental impact factor is jointly determined by wind speed, ambient temperature and humidity and expressed by polynomials or empirical functions. The energy consumption characteristic data output items include: mode mechanical energy consumption time series, mode heating energy consumption time series, environmental correction factor sequence, frequency domain representation of mode energy consumption spectrum and cumulative energy consumption distribution. To support online applications, the energy consumption characteristic analysis provides a fast approximate solution process to calculate model energy consumption estimates and uncertainty assessments within a short time window, and outputs energy consumption baseline values ​​and confidence intervals for subsequent carbon emission conversion.

[0033] Step S27: Analyze the carbon emission characteristics of the mixing tank equipment based on the energy consumption characteristic data of the mixing tank's operating mode, generating carbon emission characteristic data for the mixing tank equipment. In this embodiment of the invention, energy consumption is converted into carbon emissions according to the energy carrier: the mechanical energy consumption driven by fuel is converted based on fuel quality and carbon content coefficients, and engine / transmission system efficiency losses are considered during the conversion process; the heating energy consumption supplied by electricity is converted based on the power generation carbon emission factor of the power supply network, and heater efficiency and system heat loss are considered. The equipment carbon emission characteristic data includes two types of indicators decomposed by mode: direct emissions and indirect emissions, a unit energy consumption carbon emission coefficient array, equipment energy efficiency factor, and cumulative carbon emission time series. To reflect the impact of equipment performance degradation and maintenance status on carbon emissions, the analysis process includes an equipment energy efficiency depreciation model, which takes equipment operating years, maintenance intervals, and failure frequency as inputs and outputs efficiency correction terms. The output equipment carbon emission characteristic data is used to map with the global characteristics of the mixture structure in the mapping step and serves as the basis for direct energy-carbon conversion for carbon emission prediction.

[0034] Step S28: Based on the global carbon emission characteristic data of the mixture structure and the carbon emission characteristic data of the mixing tank equipment, establish the mapping relationship between the operation characteristics and carbon emission of the transport mixing tank, and generate a carbon emission prediction model for the transport mixing tank.

[0035] In this embodiment of the invention, two types of features—global carbon emission feature data of the mixture structure and carbon emission feature data of the mixing tank equipment—are fully connected and fused along the time axis to generate a fused feature matrix. The fusion process includes feature alignment, normalization, and anomaly label mask merging. The model adopts a multi-branch convolutional structure. The short-term branch is responsible for extracting high-frequency instantaneous features, the medium-term branch is responsible for extracting periodic and mode-switching features, and the long-term branch is responsible for extracting trend and structural change sensitivity features. Each branch consists of several convolutional layers, normalization layers, and nonlinear activation layers. The branch outputs are concatenated along the channel dimension and mapped to the regression output node through several fully connected layers. The model configuration parameters include: input feature dimension and time window length definition, number of convolutional branch layers and number of channels per layer configuration, activation function type, normalization layer type, number of regression head layers and unit type, and loss function type is mean squared error with L2 regularization added to control model complexity. Training adopts supervised learning and cross-validation strategies. The training samples include historical working condition calibration samples and multi-scenario synthetic samples generated in the laboratory. In the validation stage, a time series segmentation strategy is used to evaluate the robustness of the model. The model is deployed as a real-time inference path and includes online residual monitoring and bias adaptive correction mechanisms to ensure that it continuously provides corrected carbon emission prediction outputs under actual operating conditions.

[0036] Further, step S24 includes the following steps: Step S241: Analyze the carbon emission characteristics of each mixture's structural characteristics based on the structural characteristic data of the transport mixture, generating structural carbon emission characteristic data for the mixture. In this embodiment of the invention, the structural indicators of the mixture, such as particle composition, particle size distribution, porosity, density, and contact area, are evaluated item by item. By establishing a microscopic energy transfer model of the particles, the energy consumption of each structural unit during transport and mixing due to mechanical stirring, vibration, and temperature changes is quantified. In the analysis, energy consumption is decomposed into mechanical work consumption and heat loss. Mechanical work consumption is related to the rotational power of the mixing drum and the distribution of stirring resistance, while heat loss is calculated through local heat flux density and material thermal conductivity. Subsequently, the energy consumption is converted into carbon emission contribution according to the standard carbon emission coefficient, and the carbon emission of each structural unit is statistically summarized to generate structural carbon emission characteristic data of the mixture, including local carbon emission distribution, carbon emission ratio, energy consumption time series, and uncertainty range, providing an accurate quantitative basis for subsequent structural change analysis.

[0037] Step S242: Analyze the structural changes of the transported mixture to generate structural change data. In this embodiment, a multi-dimensional dataset of mixture structural characteristics is established based on the temperature distribution data, moisture content data, and particle size distribution data obtained from real-time monitoring of the mixture during transportation. The structural characteristics of the mixture mainly involve key parameters such as internal temperature gradient, particle density, and bonding performance. An infrared temperature sensor array is used to spatially record the temperature at different locations inside the transport mixing drum to obtain the longitudinal and radial temperature distribution curves of the mixture. Subsequently, a dielectric constant monitoring device is used to continuously monitor the moisture content in the mixture, thereby obtaining dynamic changes in moisture over time and location. For particle size distribution characteristics, an acoustic scattering monitoring device installed on the mixing drum wall is used to extract the distribution ratio and packing state of particles of different sizes in the mixture through real-time analysis of the acoustic reflection spectrum. All original monitoring results are cleaned and time-series resampling processed, then fused according to a unified timestamp and spatial location identifier to form analytically usable mixture structural characteristic data. Next, based on the multidimensional characteristic curve fitting method, the evolution of different parameters in the time series and spatial distribution is modeled, thereby identifying characteristic changes such as temperature non-uniformity, moisture content fluctuation and particle rearrangement, generating mixture structure change data for subsequent sensitivity characteristic analysis.

[0038] Step S243: Perform sensitivity characteristic analysis of mixture structure change based on the mixture structure change data to generate mixture structure change sensitivity characteristic data. In this embodiment, the mixture structure change data is classified according to different change dimensions, including temperature change dimension, moisture content change dimension, and particle size distribution change dimension. For the temperature change dimension, an energy consumption sensitivity calculation model based on thermodynamic equations is established. By comparing the stirring resistance change curves under different temperature gradients, the sensitivity coefficient of temperature change on transportation energy consumption is extracted. For the moisture content change dimension, empirical parameters from the mechanical experimental database are used to correlate moisture content fluctuations with changes in mixture viscosity and flowability. The sensitivity of moisture content to vehicle fuel consumption and carbon emissions is calculated by the trend of viscous resistance changes. For the particle size distribution change dimension, based on the discrete element method simulation results, the friction loss caused by particle rearrangement is converted into an additional energy consumption value, and the difference is compared with the baseline energy consumption under standard operating conditions to obtain the sensitivity characteristic parameters of particle structure change. Subsequently, the sensitivity parameters of each dimension are normalized and weighted using the analytic hierarchy process to form unified mixture structure change sensitivity characteristic data. The final output sensitivity feature data not only quantitatively describes the impact of the mixture on carbon emissions under different structural changes, but also provides key input parameters that can be used for real-time correction calculations, enabling the carbon emission correction model for transport vehicles to dynamically reflect the actual impact of the mixture state on energy consumption and carbon emissions.

[0039] Step S244: Perform temporal feature analysis of mixture structure change based on the mixture structure change data to generate mixture structure change temporal feature data. In this embodiment of the invention, the mixture structure change data is used as input, and unified calibration processing of the time axis and spatial axis is performed to ensure that each structure change record contains a clear timestamp and in-bucket position identifier (axial or circumferential index). Layered preprocessing is performed on the time series signal, including baseline drift removal, short-term pulse anomaly removal, and multi-scale smoothing. Then, time windows are divided according to three scales: short-term, medium-term, and long-term, and feature extraction is performed on each time window. Short-term features focus on instantaneous changes and peak detection, employing instantaneous slope estimation, peak location identification, and duration statistics to characterize sudden structural rearrangement events. Medium-term features emphasize periodicity and spectral components, using frequency domain analysis (Fourier transform or wavelet transform) to extract the dominant frequency energy ratio, energy spectral density, and power distribution within the frequency band to characterize vibration-induced repetitive structural changes. Long-term features focus on trends and cumulative effects, using piecewise regression and sliding cumulative calculations to determine cumulative porosity changes, cumulative contact area changes, and cumulative thermal resistance changes. To reflect spatial distribution differences, the above features are calculated for axial and circumferential sub-blocks within each time window, forming a time-space two-dimensional feature matrix. Further, change point detection is implemented to locate structural turning points. A robust statistical-based piecewise detection method identifies significant structural transition moments and labels the corresponding environmental and operational triggering conditions (e.g., amplitude surge segments, temperature jump segments, or continuous high-frequency vibration segments). Uncertainty assessment is performed on the extracted time-series features. Confidence intervals are generated using bootstrap resampling or variance propagation of time-series residuals. The output is a time-series feature dataset of mixture structure changes with quality labels. Fields include time index, spatial index, short, medium and long-term feature vectors, change point labels and uncertainty indexes for subsequent trend and sensitivity analysis.

[0040] Step S245: Based on the sensitivity characteristic data of mixture structure change and the time series characteristic data of mixture structure change, perform carbon emission trend characteristic analysis of mixture structure change to generate carbon emission trend characteristic data of mixture structure change. In this embodiment of the invention, the sensitivity characteristic data of mixture structure change and the time series characteristic data of mixture structure change are used as inputs to perform the formation and quantification of carbon emission trend characteristics. First, for each structural parameter (such as porosity, contact area, local thermal conductivity), a weighted superposition logic is used to multiply its sensitivity coefficient with the corresponding time series change amplitude at each time step to obtain the instantaneous carbon emission increment contribution sequence. If the sensitivity coefficient is time-varying during transportation, the dynamic sensitivity matrix within the time window is convolved with the time series change at each time step to obtain a more time-varying responsive increment sequence. The instantaneous increment sequence is integrated over time to obtain the cumulative carbon emission increment curve, and the trend characteristic items are extracted by the piecewise fitting method, including the start and end times of the main rising / falling phases, the phase slope (representing the carbon emission increase / decrease rate), the peak time, and the peak amplitude. To identify key stages, a dual screening process of thresholds and change points is implemented: thresholds are defined based on historical benchmarks and confidence intervals, while change points are located using robust piecewise regression to pinpoint inflection points. Frequency domain and periodic effects are assessed by performing spectral analysis on the instantaneous increment sequence to obtain the periodic energy proportion, which is then used as an indicator of trend volatility. To quantify the parameter contribution structure, a contribution matrix is ​​constructed, whose elements are defined as the cumulative contribution of a structural parameter to the carbon emission increment within a given time window. This matrix can be used to identify the parameters and time periods that dominate the carbon emission increment. The output is carbon emission trend characteristic data of mixture structure changes, with fields including the instantaneous increment sequence, cumulative increment curve, stage trend parameters (start and end, slope, peak value), spectral energy index, contribution matrix, and corresponding uncertainty estimate, which are used for dynamic input and early warning rules in the carbon emission prediction model.

[0041] Step S246: Perform global carbon emission characteristic analysis of the mixture structure using the mixture structure carbon emission characteristic data and the mixture structure change carbon emission trend characteristic data to generate global carbon emission characteristic data of the mixture structure.

[0042] In this embodiment of the invention, the time series of carbon emission characteristic data of mixture structure and carbon emission trend characteristic data of mixture structure change are processed for time alignment and dimension consistency. Inverse variance weighting is used to fuse the static structural carbon emission baseline and dynamic trend increment, thereby forming a time-by-time global carbon emission estimation sequence. To reveal spatial distribution differences, parallel fusion is performed on discrete units within the bucket to obtain a global carbon emission field based on a spatial grid and spatial statistics (such as local peak positions, spatial averages, and spatial variances) are calculated. Further, statistical features are extracted through multi-scale aggregation, including sliding window mean, sliding window peak, cumulative total, and percentiles at different time scales, to support subsequent visualization and alarm threshold setting. To assess and express uncertainty, a Monte Carlo resampling strategy is used to randomly perturb the sensitivity coefficient and time series increment and generate confidence intervals, which are output along with time. In addition, a parameter contribution matrix is ​​constructed to display the cumulative contribution ranking of each structural parameter to the global carbon emission throughout the transportation process, and summary statements by vehicle and batch are provided for measurement attribution. The output global carbon emission characteristic data of the mixture structure includes time-by-time global carbon emission curves, spatial distribution fields, peak and mean statistics, cumulative emissions, parameter contribution matrices and confidence intervals, etc., providing comprehensive and calibrated multi-scale feature inputs for the carbon emission prediction model of transport mixing tanks, and providing quantitative basis for real-time correction, regulation strategy formulation and offline model retraining.

[0043] Further, step S26 includes the following steps: Step S261: Extract the corresponding auxiliary operating energy consumption data of the mixing tank through the specific data of the mixing tank operation mode, and perform preliminary energy consumption characteristic analysis of the mixing tank operation mode based on the specific data of the mixing tank operation mode and the corresponding auxiliary operating energy consumption data of the mixing tank, generating preliminary energy consumption characteristic data of the mixing tank operation mode; In this embodiment of the invention, the specific data of the mixing tank operation mode is parsed. The specific data of the operation mode includes the rotation speed of the mixing tank, the frequency of rotation direction switching, the idling time of the mixing tank, the load mixing time, and the coupling relationship with the vehicle driving speed and road conditions. By setting data grouping rules, the operation mode is divided into four categories: continuous rotation, intermittent rotation, low-speed mixing, and high-speed mixing. For each type of mode, the real-time energy consumption record of the auxiliary drive system of the mixing tank is obtained using the vehicle current sensor and the hydraulic pressure sensor, and the energy consumption curve corresponding to each mode is extracted. The energy consumption curve is correlated with the specific data of the operation mode to form a time-segmented mode-energy consumption comparison table. Subsequently, through energy density calculation and time-period energy consumption accumulation processing, preliminary energy consumption characteristic data for the operating mode is generated. Specific fields include average power, peak power, energy consumption per unit time, and load / idle ratio. This embodiment ensures that the energy consumption differences under different operating modes can be quantified, and the final output preliminary energy consumption characteristic data of the stirred tank operating mode serves as the baseline data input for subsequent heat flux density and environmental correction analysis.

[0044] Step S262: Perform unsteady-state heat flux density analysis on the structural characteristic data of the transport mixture to generate unsteady-state heat flux density data. In this embodiment of the invention, unsteady-state heat flux density analysis is performed on the structural characteristic data of the transport mixture. The structural characteristic data of the mixture includes the mixture density distribution, particle contact area, porosity, specific heat capacity, and initial temperature field distribution. First, a heat transfer path topology is established, treating the mixture as a porous medium structure, and heat transfer channels are constructed through the inter-particle contact surfaces and pore air layers. The distribution of heat flux density over time is analyzed using a transient heat conduction and convection coupling method. During the analysis, input variables include the temperature boundary conditions of the inner wall of the mixing tank, the internal temperature gradient of the mixture, and the effect of the mixing state inside the tank on the local heat transfer coefficient. The transient heat flux density of each structural unit is calculated using a multi-time-slice method, thereby forming a time-seriesd heat flux density matrix. The generated unsteady-state heat flux density data of the mixture includes local heat flux density curves, the overall average heat flux density variation trend, and the location of the heat flux density peak interval. This data provides direct input for assessing the heat loss and energy consumption correction of the mixing tank, and can be cross-compared with the energy consumption characteristic data of the mixing tank mode.

[0045] Step S263: Based on environmental perception data and unsteady heat flux density data of the mixture, adjust the environmental impact parameters of the initial mixing drum operation mode energy consumption characteristic data to generate mixing drum operation mode energy consumption characteristic data.

[0046] In this embodiment of the invention, environmental sensing data is combined with unsteady-state heat flux density data of the mixture to adjust environmental impact parameters of the preliminary mixing drum operation mode energy consumption characteristic data. Environmental sensing data includes external temperature, humidity, wind speed, solar radiation intensity, and road slope. First, a normalized feature vector is established for the environmental parameters to ensure that various environmental indicators have uniform dimensions. Then, the unsteady-state heat flux density data of the mixture is mapped to the environmental parameter vector in a multidimensional manner to form a heat flux-environment coupling factor, which is used to quantify the corrective effect of the external environment on energy consumption performance. For example, under high external temperature conditions, the heat flux density gradient decreases, thus reducing the insulation load of the mixing drum, and the corresponding energy consumption needs to be reduced proportionally; while in strong wind and low temperature environments, the heat flux density gradient increases, enhancing heat dissipation from the outer wall of the mixing drum, requiring an upward adjustment of the operation mode energy consumption. By establishing a multidimensional weighting coefficient matrix, these correction factors are superimposed on the preliminary energy consumption characteristic data to generate the final mixing drum operation mode energy consumption characteristic data. The output includes energy consumption curves under multiple operating modes, environmental correction coefficients, and corrected energy consumption statistics, providing more accurate basic data on energy consumption for real-time correction of carbon emissions from subsequent mixture transportation.

[0047] Furthermore, step S263 includes the following steps: designing environmental perception impact quantification parameters based on environmental perception data; performing environmental perception impact benefit quantification analysis on the unsteady heat flux density data of the mixture based on the environmental perception impact quantification parameters, generating environmental perception impact benefit quantification data on the mixture's thermal effect; designing environmental impact factors for the energy consumption of the mixing drum operation mode based on the environmental perception impact benefit quantification data on the mixture's thermal effect, generating environmental impact factors for the energy consumption of the mixing drum operation mode; and adjusting the environmental impact parameters of the initial energy consumption characteristic data of the mixing drum operation mode using the environmental impact factors for the energy consumption of the mixing drum operation mode, generating energy consumption characteristic data of the mixing drum operation mode.

[0048] In this embodiment of the invention, the environmental sensing data includes measured values ​​of ambient temperature, relative humidity, wind speed, wind direction, solar radiation intensity, precipitation status, road surface roughness index, instantaneous vehicle speed, and relative wind speed. Statistics are calculated for these physical quantities in both short-term and long-term time windows. Short-term statistics characterize pulsating effects, such as short-term wind gusts and sudden temperature changes, while long-term statistics reflect slow trends. Based on heat transfer theory and engineering experience, a set of physical quantitative parameters is constructed: local convective heat transfer coefficient, equivalent radiative heat transfer coefficient, surface evaporation mass transfer coefficient, boundary layer stability index, and net thermal driving force index. The convective heat transfer coefficient uses an empirical correlation formula of wind speed-windward relationship and combines it with wind tunnel calibration curves to obtain a mapping table, reflecting the amplification or attenuation effect of vehicle shape and local wind field on heat transfer; the radiative heat transfer coefficient is obtained by mapping the material surface emissivity to the equivalent sky temperature; the evaporation mass transfer coefficient is quantitatively calculated from the empirical relationship between relative humidity and convective mass transfer; and the boundary layer stability index is given by a non-dimensional combination of ambient temperature difference and wind speed to characterize the strength of thermal convection. Statistical characteristics include instantaneous values, moving averages, standard deviations, pulsation intensity, and peak frequency, which are normalized using physical quantities to form a parameter vector with consistent dimension. Parameter calibration is achieved by combining laboratory heat flux measurements, wind tunnel tests, and field comparative tests to obtain correction coefficients. The calibration process employs batch-repeated experiments to estimate parameter uncertainties and generate confidence intervals. The final output is a set of time-resolution environmental perception impact quantification parameters, with each element carrying an uncertainty estimate and quality label, serving as the input basis for subsequent heat flux density correction and energy consumption factor design. The unsteady-state heat flux density data of the mixture represents the instantaneous heat flux distribution facing outwards at different locations inside the mixing tank over time. The heat flux field is first decomposed into convection components, radiation components, and latent heat of vaporization components according to physical mechanisms. The convection component is expressed as the product of the local convective heat transfer coefficient and the surface-ambient temperature difference, with the heat transfer coefficient obtained using the aforementioned mapping table of environmentally perceived parameters. The radiation component is estimated as the product of surface emissivity and the equivalent radiative temperature difference, which is derived from solar radiation intensity and sky radiation temperature. The evaporation component is estimated by mapping the latent heat loss to the vapor pressure difference, convective mass transfer coefficient, and surface moisture activity of the mixture. The time-by-time difference between the predicted values ​​of the above physical model and the measured unsteady-state heat flux density is calculated, and a data-driven residual corrector is used to model the difference to compensate for complex coupling effects not covered by the physical model. The residual corrector adopts an ensemble regression model structure, which is an ensemble composed of several base learners. During training, the mean squared error is used as the loss and regularization constraints are added to prevent overfitting. The model hyperparameters are determined and recorded through a cross-validation procedure. The output is quantitative data on the environmentally perceived impact of the mixture's thermal effect, given by spatial cell and time scale. The data fields include the increase or decrease in local heat flux caused by the environment, the corresponding energy consumption difference estimate, and the corresponding uncertainty interval.The quantitative results can be used to evaluate the effect of external environment on the gain or suppression of instantaneous heat loss, and can also provide physical order of magnitude for the subsequent numerical design of energy consumption factor.

[0049] The environmental impact data of the mixture's thermal effect are quantified by spatial and temporal integration to convert it into an overall additional heat energy demand, which represents the additional heating power required to maintain the target temperature of the mixture. For the mechanical energy consumption of the mixing tank operation mode, changes in moisture content, particle rearrangement, and viscosity are first mapped to increments in viscosity and friction loss, and then the additional mechanical power demand is calculated using the torque-angular velocity relationship. To represent both long-term trends and respond to short-term fluctuations, an environmental impact factor is designed as a combined adjustment for each operation mode, comprising a multiplicative factor and an additive compensation term. The multiplicative factor characterizes the relative amplification or attenuation of baseline energy consumption under changes in overall heat load, while the additive compensation term compensates for short-term peak demand and start-up pulse energy consumption. The factor design process uses a model piecewise regressor to fit the environmental benefit quantification data with the observed model energy consumption increments. The regressor structure is linear regression combined with L2 regularization to maintain model stability. The input vector includes the spatial average heat flux increment, environmental baseline quantity, and model feature vector; the output is the value of the multiplicative factor and the additive compensation term. The fitting process employs multiple batches of calibration samples accompanied by residual analysis to estimate factor uncertainties. Factor results are recorded in time series format and possess time-varying capabilities to reflect the dynamic impact of environmental conditions during operation. The final output consists of energy consumption environmental impact factors and corresponding confidence indices for the mixing tank's operating mode, differentiated by operating mode and time, used for parametric adjustment of baseline energy consumption characteristics. At each time step, preliminary operating mode energy consumption characteristics are retrieved according to the current operating mode identifier. These characteristics include the baseline mechanical power time series and the baseline heating power time series. Based on the multiplicative factor of the operating mode energy consumption environmental impact factors, the baseline power series is proportionally amplified or reduced point by point, and additive compensation terms are superimposed onto the baseline peak segment according to time windows to correct instantaneous demand. To prevent non-physical jumps caused by transient anomalies, smoothing filtering and upper and lower limit constraints are applied to the corrected power series. The upper and lower limits are jointly defined by the rated power of the mixing tank equipment, historical reliable peak statistics, and safety margin. Uncertainty propagation is accomplished through analytical methods or Monte Carlo sampling: confidence intervals are generated based on the uncertainty estimates of the factors and baseline power and output over time. If environmental observations are unavailable or anomalies occur, time interpolation is used and the confidence level is lowered to alert subsequent audits. All correction steps record the residual time series and serve as a sample set for online bias correction and offline recalibration. The correction results are output as energy consumption characteristic data for the stirred tank operation mode, including the corrected mechanical power curve, the corrected heating power curve, the sequence of environmental impact factors used, the uncertainty interval for each step, and residual records, for subsequent carbon emission conversion, real-time correction, and model iteration.

[0050] Further, step S28 includes the following steps: Step S281: Using the mixing tank operation mode-specific data corresponding to the carbon emission characteristic data of the mixing tank equipment, a cross-feature analysis of the transportation mixing tank operation is performed on the mixture structure change sensitivity characteristic data and the mixture structure change time series characteristic data corresponding to the global carbon emission characteristic data of the mixture structure, generating transportation mixing tank operation cross-feature data; In this embodiment of the invention, based on the mixing tank equipment carbon emission characteristic data and the mixing tank operation mode-specific data, a cross-feature analysis of the transportation mixing tank operation is performed on the structure change sensitivity characteristics and structure change time series characteristics recorded in the global carbon emission characteristic data of the mixture structure. The two types of time series features are aligned on the time axis, a unified time grid is used, and interpolation and labeling are performed on the missing time series segments; subsequently, convolutional correlation analysis is performed on the mixing tank mode power spectrum and the mixture structure sensitivity time series according to the sliding time window to identify the coupling relationship between the mode power change and the structure sensitivity response under time delay. The cross-feature construction includes: time-progressive product terms to reflect instantaneous coupling strength, lag product sequences to reveal delay effects, cross-correlation peaks and peak lags to quantify response rates, and frequency-domain-based coherence coefficients to characterize periodic coupling. To capture nonlinear interactions, canonical correlation analysis and mutual information metrics are introduced to perform high-order coupling quantification on multivariate time-series segments; a pattern-structure cross matrix is ​​constructed for each operating mode, with matrix elements including instantaneous coupling energy, cumulative coupling contribution, maximum response lag, and in-band coherence energy. Cross-features also include statistical summary terms: mean, standard deviation, skewness, and kurtosis within the window, to reveal the distribution characteristics of coupling strength. The final output of the transport mixing tank operating cross-feature data is a time-series-based matrix set, with fields including time index, pattern identifier, coupling energy, response lag, coherence index and statistics, as well as data quality labels and uncertainty estimates corresponding to each item. This cross-feature data is used in subsequent fully connected fusion and multi-scale feature extraction steps.

[0051] Step S282: The global carbon emission feature data of the mixture structure and the carbon emission feature data of the mixing tank are processed using the cross-feature data of the transport mixing tank operation to generate fully connected fusion feature data of the transport mixing tank carbon emission. In this embodiment of the invention, the purpose of performing fully connected fusion feature processing on the global carbon emission feature data of the mixture structure and the carbon emission feature data of the mixing tank based on the cross-feature data of the transport mixing tank operation is to construct a high-dimensional feature representation with extensive interaction terms for subsequent multi-scale analysis. The fusion processing first performs normalization and standardization on all input features, adopts zero-mean unit variance transformation, and retains the original uncertainty information. Then, a fully connected fusion matrix is ​​constructed: an outer product operation is performed on the main feature set to form second-order cross features, and a time-delayed embedding is generated in the time dimension using a sliding window to construct a time-series interaction layer. To control dimensionality growth, a layer-by-layer dimensionality reduction strategy was applied. Principal component analysis was first used to extract explanatory variables, and then sparse linear mapping was employed to compress the linear redundancy of cross features, preserving the interaction subspaces with strong explanatory power. Quadratic and cubic combination features were added to important interaction terms to retain nonlinear interaction information. Next, a set of descriptive statistics were calculated on the fusion matrix, including temporal cumulative energy, peak contribution parameter, parameter correlation matrix, and conditional mutual information score. These statistics supplement the fully connected fusion features. The fusion features were also expanded according to a spatial grid to form a local fusion vector for each spatial unit, facilitating subsequent multi-scale convolutional operations to handle spatial-temporal coupling. The final output is the fully connected fusion feature data of carbon emissions from transport mixing tanks. Its structure includes a time axis, spatial axis, fusion vector, second-order and higher-order interaction indices, statistical summaries, and uncertainty assessments for each item, ensuring that subsequent models can fully learn the impact of mixing tank operation and mixture structure interaction on carbon emissions under multi-dimensional input.

[0052] Step S283: Analyze the carbon emission fusion feature data of the transport mixing tank at different operating scales using a preset multi-scale convolutional kernel to generate multi-scale carbon emission fusion feature data. In this embodiment, multi-scale convolutional kernels are used to perform carbon emission fusion feature analysis on the fully connected fusion feature data of the transport mixing tank's carbon emission to reveal significant patterns and temporal relationships at different operating scales. The implementation method involves constructing a multi-branch one-dimensional temporal convolutional network with three branches, corresponding to short-term, periodic, and long-term scales. The short-term branch uses continuous convolutional layers with small receptive fields and kernel lengths of 3 and 5, aiming to capture instantaneous pulses and high-frequency coupling features. The medium-term branch uses kernel lengths of 7 and 15 and introduces dilated convolution to expand the receptive field and cover the medium-term process of mode switching. The long-term branch uses a kernel length of 31 and a larger dilation rate to extract slow evolution and structural change trends. Each convolutional layer is followed by batch normalization and linear rectified activation units, and residual connections are used within the branches to stabilize training and accelerate convergence. The number of convolutional kernels is configured with 64, 128, and 128 channels in the short-term, medium-term, and long-term branches, respectively. The stacking depth of the convolutional layers is three layers in the short-term, four layers in the medium-term, and four layers in the long-term. Global average pooling is performed at the end of each convolutional sequence to obtain a scale-level summary vector. Multi-scale features are concatenated by channel dimension and then subjected to dimensionality reduction mapping. The dimensionality reduction module uses a single fully connected layer for mapping and normalization to generate the final multi-scale mixing tank carbon emission fusion feature vector. In addition, to capture frequency domain characteristics, a time-frequency transform module is incorporated before the convolutional branches to perform short-time Fourier transform on the input features and use the frequency domain energy spectrum as the parallel channel input to the convolutional branches, enhancing the sensitivity to periodic perturbations. The output multi-scale feature data includes short-term, medium-term, and long-term feature components, frequency domain energy indices, and residual diagnostic terms, providing a rich multi-scale representation for the regression model.

[0053] Step S284: Based on the multi-scale mixing tank operation carbon emission fusion feature data, establish the mapping relationship between the operation characteristics and carbon emission of the transport mixing tank, and generate a carbon emission prediction model for the transport mixing tank.

[0054] In this embodiment of the invention, a carbon emission prediction model for transport mixing tanks is established. The model structure is a multi-branch convolutional-regression fusion network. The input dimensions are the time window length and the fusion feature dimension. The time window length is set to a sliding window of several seconds to several minutes to cover short, medium, and long-term features. The main body of the network extracts features in parallel using a three-branch convolutional module (corresponding to the configuration in step S283). The branch output channels are concatenated and mapped to the scalar regression output through two fully connected regression heads. The regression head configuration parameters are as follows: the first fully connected layer has 256 units and the activation function is a linear rectified unit; the second fully connected layer has 64 units and the activation function is a linear rectified unit; the output layer is a single neuron without an activation function to directly output the predicted carbon emission value. The model training uses mean squared error loss and adds an L2 regularization term to constrain the model complexity; the optimizer uses adaptive moment estimation and is configured with an initial learning rate on the order of a tiny magnitude. The batch size and the number of training rounds are set according to the calibrated sample size and the real-time update window. To achieve online adaptive capabilities, the model deployment includes residual monitoring and bias fine-tuning modules: during the inference phase, the predicted residuals are calculated and the bias term of the regression head is updated using an exponentially weighted average method to reflect the latest operating condition shift, while historical residual records are retained for periodic offline retraining. The model output includes uncertainty estimation, and distributed predictions are generated and confidence intervals are calculated during inference using the Monte Carlo dropout method. The final output carbon emission prediction model for transport mixing tanks receives fused features in a sliding window manner during the real-time inference path and generates time-by-time carbon emission predictions. It also provides confidence intervals, residual diagnostics, and energy consumption decomposition information for control feedback to support subsequent real-time carbon emission correction and closed-loop operation and maintenance decisions.

[0055] Further, step S3 includes the following steps: Step S31: Analyze the real-time transportation demand of the mixture based on the parameters of the mixing plant, and generate real-time transportation demand data; perform real-time data analysis on the environmental perception data, and generate real-time environmental perception data; In this embodiment of the invention, the transportation demand is analyzed based on the real-time parameter data of the mixture. The real-time parameters of the mixture include the mix proportion, physical density, initial temperature, target construction temperature, and allowable transportation time window. These parameters are collected in a structured manner by the mixing plant's metering and discharge system at the moment the material is loaded, and the transportation demand is determined through a regularized parameter mapping relationship. For example, when the difference between the initial temperature of the mixture and the target construction temperature is large, the system will calculate the upper limit of the required transportation time and generate a time-sensitive transportation demand label; when the viscosity and aggregate content of the mixture exceed a certain threshold, the system will generate a high-speed maintenance requirement for the mixing drum to prevent material stratification and settling. Based on this, combined with fleet scheduling parameters, such as transportation distance, road slope, and traffic density, real-time transportation demand data of the mixture is formed. This data includes the target speed range, allowable temperature drop range, maximum delay time, and corresponding priority transportation level. Simultaneously, real-time environmental perception data is processed. This data is acquired from multiple external sensors, including air temperature, humidity, wind speed, road surface temperature, and traffic flow density. The processing logic involves first performing data denoising, and then calculating the dynamic trend of the environment using distributed moving average filtering. The output real-time environmental perception data includes a meteorological condition distribution matrix and traffic flow temporal characteristics.

[0056] Step S32: Based on the real-time transportation demand data of the mixture, perform real-time mixing drum control operation characteristic analysis on the auxiliary operation data of the mixing drum of the transport vehicle to generate real-time mixing drum control operation characteristic data; In this embodiment of the invention, real-time control characteristic analysis is performed on the auxiliary operation data of the mixing drum of the transport vehicle. The auxiliary operation data of the mixing drum is acquired by torque sensor, angular velocity sensor and internal temperature sensor installed on the drum body, with a sampling frequency of not less than 50Hz to ensure time resolution. The mixing drum speed curve is compared with the transportation demand speed range. When the actual speed deviates from the target range, the system records the degree of deviation and the duration of deviation. Then, the nonlinear relationship between the mixing drum motor torque and energy consumption is calculated to determine the additional energy cost when maintaining transportation demand. Secondly, the temperature curve of the mixture inside the drum is compared with the allowable temperature drop range in real time. When the temperature drop rate exceeds the threshold, an indicator for accelerated mixing or increased heat preservation energy consumption command is triggered. In the feature extraction stage, real-time mixing drum control operation characteristic data is generated through multivariate time series analysis, including: mean square error of speed deviation, temperature drop rate gradient, energy overload coefficient, drum body vibration amplitude spectrum and control command trigger frequency. These high-dimensional features are used to quantify the degree of dynamic matching between transportation demand and actual equipment operation.

[0057] Step S33: Transmit the real-time mixing tank control and operation characteristic data and the real-time environmental perception data to the transport mixing tank carbon emission prediction model to perform real-time carbon emission prediction processing of the transport mixing tank and generate real-time carbon emission data of the transport mixing tank.

[0058] In this embodiment of the invention, real-time stirring tank control operation characteristic data and real-time environmental perception data are input into the carbon emission prediction model of the transport stirring tank to perform real-time carbon emission prediction. The prediction model adopts a neural network architecture combining multi-branch convolution and temporal regression. The input layer receives multi-dimensional feature vectors from two types of data, where the stirring tank control operation characteristic data mainly reflects mechanical energy consumption and thermal effect consumption, and the environmental perception data reflects the additional impact of external conditions on energy consumption. The intermediate layer extracts energy consumption coupling features at different time scales through multi-scale convolution, and then models the temporal dependency relationship through bidirectional long short-term memory units to capture the changing patterns of carbon emissions under short-term fluctuations and long-term trends. The output layer is a single-value regression node, corresponding to the carbon emission value of the transport stirring tank within a unit time window. During training, mean squared error is used as the loss function, and a Monte Carlo dropout mechanism is added in the prediction stage to provide a confidence range for the prediction interval. The final output of the real-time carbon emission data of the transport mixing tank is a time-series carbon emission prediction value. Each moment includes the actual carbon emission estimate, the upper and lower limit range of the prediction, and the residual correction value to support the carbon emission control of the real-time correction process.

[0059] Further, step S4 includes the following steps: Step S41: Perform vehicle operation state sequence analysis on the transport vehicle operation data to generate vehicle operation state sequence data; In this embodiment of the invention, the transport vehicle operation data is subjected to operation state sequence analysis. The operation data includes vehicle speed, acceleration, throttle opening, engine speed, braking status, road gradient, and vehicle load weight, etc. An operation state coding mechanism is established to convert continuous operation parameters into a state sequence. For example, when the speed is higher than the rated threshold and the throttle opening is greater than 70%, it is defined as "high-speed heavy-load acceleration state"; when braking is frequent and the road gradient is greater than 5%, it is defined as "downhill braking state". Subsequently, a hidden Markov model is used to fit the state sequence in segments, and the state transition probability matrix and duration distribution of each time period are output, thereby forming vehicle operation state sequence data. This data not only preserves the physical continuity of the operation process, but also provides the transition rules between states, which can be used for subsequent energy consumption accumulation feature analysis.

[0060] Step S42: Perform cumulative energy consumption feature analysis on the vehicle operating state sequence data to generate cumulative energy consumption feature data for the operating state sequence. In this embodiment of the invention, cumulative energy consumption feature analysis is performed on the operating state sequence. First, the instantaneous fuel consumption data collected by the fuel flow sensor and the engine speed curve are mapped to each operating condition interval. The calculation formula is to integrate the instantaneous fuel consumption within each operating condition interval and normalize it by unit time to obtain the operating condition energy consumption rate. Second, the cumulative energy consumption of the operating condition is calculated, that is, the cumulative value is obtained by multiplying the operating condition duration by the energy consumption rate, and then superimposed on each operating condition to form a time series. Further, the cumulative energy consumption features are extracted, including the proportion of operating condition energy consumption, the energy consumption fluctuation amplitude of acceleration and braking conditions, the time gradient of energy consumption growth rate, and the duration of high energy consumption intervals. To identify abnormal energy consumption segments, a sliding window energy consumption mean and variance sequence is constructed. If the energy consumption exceeds twice the standard deviation of the mean, it is marked as a high energy consumption segment. The output operating status sequence energy consumption cumulative feature data has a multi-dimensional structure, including cumulative energy consumption value, energy consumption rate, fluctuation characteristics and anomaly labels, providing input for operating condition correction processing.

[0061] Step S43: Based on the mixture parameters of the transportation mixing plant and the transportation road data, the cumulative energy consumption feature data of the operating state sequence is corrected for the influence of operating condition differences, generating corrected cumulative energy consumption feature data for operating conditions. In this embodiment of the invention, the cumulative energy consumption feature data of the operating state sequence is corrected for the influence of operating condition differences based on the mixture parameters of the transportation mixing plant and the transportation road data. The mixture parameters include mass, viscosity, and thermal sensitivity; the road data includes slope, curvature, road surface friction coefficient, and traffic density. First, an operating condition difference correction factor matrix is ​​constructed. Taking the slope influence factor as an example, when the slope is greater than 4%, the additional energy consumption increase is calculated as 10% of the operating condition energy consumption; when the traffic density index is greater than 0.7, the cumulative value of idling and low-speed energy consumption is increased by 15%. Second, the mixture mass parameter is added as a weight to the correction model. The greater the mass, the higher the amplification factor for accelerated operating condition energy consumption. The correction calculation process is as follows: the original cumulative energy consumption feature value of operating conditions is multiplied by the corresponding weight in the operating condition difference correction factor matrix to obtain the corrected cumulative energy consumption sequence. The process outputs corrected cumulative energy consumption characteristic data for operating conditions. The data includes the corrected energy consumption sequence, the source of the correction coefficient, and the range of uncertainty.

[0062] Step S44: Based on the corrected energy consumption cumulative characteristic data of the operating conditions, perform real-time carbon emission analysis and processing of the transport vehicle's operation to generate real-time carbon emission data of the transport vehicle.

[0063] In this embodiment of the invention, real-time carbon emission analysis of transport vehicles is performed based on cumulative energy consumption characteristic data corrected for operating conditions. The standard carbon emission factor method is used to convert fuel consumption into carbon emissions. Specifically, the total fuel consumption is calculated based on the corrected cumulative energy consumption value, and then multiplied by the fuel's carbon emission factor to obtain the total carbon emissions. To achieve real-time performance, a sliding window calculation method is used, with each window lasting 10 seconds, outputting the carbon emissions within that time period. Further calculations are made of the cumulative carbon emissions curve, instantaneous carbon emissions change rate, carbon emissions per unit distance, and the contribution of operating conditions. The output real-time carbon emissions data structure for transport vehicles includes a timestamp, cumulative carbon emissions, instantaneous carbon emissions rate, emissions per unit mileage, and the contribution of operating conditions, providing a basis for real-time carbon emissions correction and operation and maintenance strategy optimization.

[0064] Furthermore, step S43 includes the following steps: analyzing the impact characteristics of vehicle transportation based on the parameters of the mixed material at the transportation mixing plant and the transportation road data, and generating vehicle transportation impact condition characteristic data; analyzing the driving behavior characteristics of each vehicle transportation condition based on the vehicle transportation impact condition characteristic data and the vehicle operation state sequence data, and generating vehicle transportation condition driving behavior characteristic data; and performing condition difference correction processing on the cumulative energy consumption characteristic data of the operation state sequence based on the vehicle transportation condition driving behavior characteristic data and the corresponding vehicle transportation impact condition characteristic data, and generating operation condition corrected cumulative energy consumption characteristic data.

[0065] In this embodiment of the invention, the influence characteristics of vehicle transportation on operating conditions are analyzed based on the parameters of the mixture and the characteristics of the transportation road. The mixture parameters include mixture mass, viscosity, moisture content, and heat sensitivity. The transportation road data includes road slope, curve radius, road surface friction coefficient, traffic flow density, road segment length, and road surface material. The road data is segmented, dividing the entire road into several continuous segments, each corresponding to different slopes, curvatures, and friction characteristics. Then, combined with the mixture parameters, the influence coefficient of each road segment on vehicle load and power output is calculated. For example, the greater the mixture mass, the higher the proportion of energy consumption increase under acceleration conditions on slopes; when the road surface friction coefficient decreases, the vehicle's traction demand increases, leading to an increase in energy consumption. A multivariate regression method is used to establish a road segment load-energy consumption mapping relationship to obtain the energy consumption influence factor for each road segment. Finally, the energy consumption influencing factors of each road segment are fused with the driving time series of continuous road segments to generate vehicle transportation impact condition characteristic data. This data includes the slope factor, friction factor, traffic congestion factor, mixture load factor, and comprehensive operating condition impact index for each road segment, providing basic input for subsequent driving behavior characteristic analysis. The vehicle transportation impact condition characteristic data is time-aligned to ensure that the vehicle state at each moment matches the corresponding road segment's operating condition influencing factors. Driving behavior patterns for each operating condition are identified through time series segmented analysis, such as acceleration, constant speed, deceleration, and braking. Average acceleration, instantaneous torque change, fuel flow, and engine efficiency features are extracted for each segment. Furthermore, driving behavior is combined with influencing factors such as road segment slope, curve radius, and traffic flow density to calculate the driving behavior operating condition adjustment coefficient (i.e., setting the impact weight of driving behavior on each slope or traffic condition). For example, in high-slope sections, the increase in energy consumption due to acceleration is recorded as the gradient condition increase; in congested sections, the fuel consumption due to idling is quantified as the traffic congestion condition increase. Driving behavior characteristics and condition adjustment factors are integrated into vehicle transportation condition driving behavior characteristic data, which includes the average acceleration, energy consumption correction coefficient, and driving behavior adjustment weight for each condition segment, providing a basis for energy consumption correction in operating states. After generating the vehicle transportation condition driving behavior characteristic data, the cumulative energy consumption characteristic data of the operating state sequence is corrected for the impact of condition differences. Based on the condition driving behavior characteristic data, the energy consumption increase or decrease for each condition segment is calculated, and the original cumulative energy consumption for each condition segment is weighted and adjusted with the driving behavior correction coefficient and the vehicle transportation impact condition characteristic index. For example, in sections with steep gradients and high vehicle loads, the original cumulative energy consumption of acceleration condition is multiplied by the gradient factor and the load factor to obtain the corrected acceleration condition energy consumption; in congested sections, idling energy consumption is corrected based on the duration of idling condition and the congestion factor. The corrected energy consumption of all operating conditions is accumulated to form a corrected vehicle operating energy consumption sequence. The cumulative energy consumption, energy consumption per unit mileage, and the proportion of energy consumption for each operating condition are calculated to ensure that the corrected sequence is complete and continuous.The system generates cumulative energy consumption characteristic data corrected for operating conditions. This data includes the corrected energy consumption value for each operating condition segment, the operating condition type, the source of the correction factor, and the corresponding influence weight. This provides accurate energy consumption input corrected for operating condition differences for subsequent real-time carbon emission analysis of transport vehicles. The entire process ensures that road, load, and driving behavior factors during the transportation of the mixture are accurately considered, achieving dynamic correction and precise characterization of carbon emissions.

[0066] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0067] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A method for real-time correction of carbon emissions during the transportation of mixed materials at a mixing plant, characterized in that, Includes the following steps: Step S1: Utilize the multi-source sensors built into the transport vehicle to collect and synchronously analyze multi-source data of the mixed material transport vehicle, generating synchronous multi-source data of the transport vehicle. This synchronous multi-source data includes environmental perception data, transport road data, transport vehicle operation data, and auxiliary operation data of the transport vehicle's mixing drum. Step S2: Obtain the parameters of the mixed material at the transport mixing plant. Based on the environmental perception data, the auxiliary operation data of the transport vehicle's mixing drum, and the parameters of the mixed material at the transport mixing plant, establish a mapping relationship between the operating characteristics and carbon emission of the transport mixing drum, generating a carbon emission prediction model for the transport mixing drum. Step S3: Perform real-time data analysis on the environmental perception data to generate real-time environmental perception data. Based on the parameters of the mixture at the transportation mixing plant and the auxiliary operation data of the mixing drum of the transportation vehicle, the real-time control operation characteristics of the mixing drum are analyzed, and real-time control operation characteristic data of the mixing drum are generated. The real-time mixing drum control operation characteristic data and real-time environmental perception data are transmitted to the transportation mixing drum carbon emission prediction model to perform real-time carbon emission prediction processing of the transportation mixing drum and generate real-time carbon emission data of the transportation mixing drum; Step S4: Based on transportation road data and transportation vehicle operation data, the real-time carbon emission analysis processing of transportation vehicle driving is performed to generate real-time carbon emission data of transportation vehicle. Step S5: Perform real-time correction of carbon emissions from the transportation of the mixing plant's aggregates using real-time carbon emission data from the transport mixing drum and the transport vehicles.

2. The method for real-time correction of carbon emissions from the transportation of mixed materials at a mixing plant according to claim 1, characterized in that, Step S1 includes the following steps: using the multi-source sensors built into the transport vehicle to collect multi-source data of the mixed material transport vehicle, generating multi-source data of the transport vehicle; performing multi-source data time synchronization and preprocessing on the multi-source data of the transport vehicle to generate synchronized multi-source data of the transport vehicle.

3. The method for real-time correction of carbon emissions from the transportation of mixed materials at a mixing plant according to claim 1, characterized in that, Step S2 includes the following steps: Step S21: Obtain the parameters of the mixture at the transportation mixing plant; Step S22: Analyze the property characteristics of the mixture at the transportation mixing plant based on the parameters, and generate property characteristic data of the transportation mixture; Step S23: Analyze the structural characteristics of the transportation mixture based on the property characteristic data of the transportation mixture, and generate structural characteristic data of the transportation mixture; Step S24: Analyze the global carbon emission characteristics of the transportation mixture in its normal state and structural changes based on the structural characteristic data of the transportation mixture, and generate global carbon emission characteristic data of the mixture structure; Step S25: Mix the mixture based on the auxiliary operation data of the mixing drum of the transportation vehicle. Step S26: Based on environmental perception data and the specific data of the mixing tank's operating mode, perform energy consumption characteristic analysis of the mixing tank's operating mode to generate energy consumption characteristic data of the mixing tank's operating mode; Step S27: Based on the energy consumption characteristic data of the mixing tank's operating mode, perform carbon emission characteristic analysis of the mixing tank equipment to generate carbon emission characteristic data of the mixing tank equipment; Step S28: Based on the global carbon emission characteristic data of the mixture structure and the carbon emission characteristic data of the mixing tank equipment, establish the mapping relationship between the operating characteristics and carbon emission of the transport mixing tank, and generate a carbon emission prediction model for the transport mixing tank.

4. The method for real-time correction of carbon emissions during the transportation of mixed materials at a mixing plant according to claim 3, characterized in that, Step S24 includes the following steps: Step S241: Analyze the carbon emission characteristics of each mixture structure based on the transport mixture structure characteristic data to generate mixture structure carbon emission characteristic data; Step S242: Analyze the mixture structure change based on the transport mixture structure characteristic data to generate mixture structure change data; Step S243: Analyze the mixture structure change sensitivity characteristics based on the mixture structure change data to generate mixture structure change sensitivity characteristic data; Step S244: Analyze the mixture structure change time series characteristics based on the mixture structure change data to generate mixture structure change time series characteristic data; Step S245: Analyze the mixture structure change carbon emission trend characteristics based on the mixture structure change sensitivity characteristic data and the mixture structure change time series characteristic data to generate mixture structure change carbon emission trend characteristic data; Step S246: Analyze the mixture structure global carbon emission characteristics based on the mixture structure carbon emission characteristic data and the mixture structure change carbon emission trend characteristic data to generate mixture structure global carbon emission characteristic data.

5. The method for real-time correction of carbon emissions during the transportation of mixed materials at a mixing plant according to claim 3, characterized in that, Step S26 includes the following steps: Step S261: Extract the corresponding auxiliary operation energy consumption data of the mixing tank through the specific data of the mixing tank operation mode, and perform preliminary energy consumption characteristic analysis of the mixing tank operation mode based on the specific data of the mixing tank operation mode and the corresponding auxiliary operation energy consumption data of the mixing tank to generate preliminary energy consumption characteristic data of the mixing tank operation mode; Step S262: Perform unsteady heat flux density analysis on the structural characteristic data of the transported mixture to generate unsteady heat flux density data of the mixture; Step S263: Based on the environmental perception data and the unsteady heat flux density data of the mixture, perform environmental impact parameter adjustment processing on the preliminary energy consumption characteristic data of the mixing tank operation mode to generate energy consumption characteristic data of the mixing tank operation mode.

6. The method for real-time correction of carbon emissions during the transportation of mixed materials at a mixing plant according to claim 5, characterized in that, Step S263 includes the following steps: designing environmental perception impact quantification parameters based on environmental perception data; performing environmental perception impact benefit quantification analysis on the unsteady heat flux density data of the mixture based on the environmental perception impact quantification parameters, generating environmental perception impact benefit quantification data on the mixture's thermal effect; designing environmental impact factors for the energy consumption of the mixing drum operation mode based on the environmental perception impact benefit quantification data on the mixture's thermal effect, generating environmental impact factors for the energy consumption of the mixing drum operation mode; adjusting the environmental impact parameters of the initial energy consumption characteristic data of the mixing drum operation mode using the environmental impact factors for the energy consumption of the mixing drum operation mode, generating energy consumption characteristic data of the mixing drum operation mode.

7. The method for real-time correction of carbon emissions during the transportation of mixed materials at a mixing plant according to claim 4, characterized in that, Step S28 includes the following steps: Step S281: Using the specific data of the mixing tank operation mode corresponding to the carbon emission characteristic data of the mixing tank equipment, perform cross-feature analysis of the transportation mixing tank operation on the global carbon emission characteristic data of the mixture structure, as well as the time series characteristic data of the mixture structure change, to generate cross-feature data of the transportation mixing tank operation; Step S282: Using the cross-feature data of the transportation mixing tank operation, perform full-connected fusion feature processing on the global carbon emission characteristic data of the mixture structure and the carbon emission characteristic data of the mixing tank equipment to generate full-connected fusion feature data of the transportation mixing tank carbon emission; Step S283: Using a preset multi-scale convolution kernel, perform carbon emission fusion feature analysis on the full-connected fusion feature data of the transportation mixing tank carbon emission based on the differences in the mixing tank operation scale to generate multi-scale mixing tank operation carbon emission fusion feature data; Step S284: Based on the multi-scale mixing tank operation carbon emission fusion feature data, establish the mapping relationship between the operation characteristics and carbon emission of the transportation mixing tank, and generate a carbon emission prediction model for the transportation mixing tank.

8. The method for real-time correction of carbon emissions during the transportation of mixed materials at a mixing plant according to claim 1, characterized in that, Step S3 includes the following steps: Step S31: Analyze the real-time transportation demand of the mixture based on the parameters of the mixing plant to generate real-time transportation demand data; perform real-time data analysis on the environmental sensing data to generate real-time environmental sensing data; Step S32: Analyze the real-time mixing drum control operation characteristics based on the real-time transportation demand data and the auxiliary operation data of the mixing drum of the transport vehicle to generate real-time mixing drum control operation characteristic data; Step S33: Transmit the real-time mixing drum control operation characteristic data and the real-time environmental sensing data to the carbon emission prediction model of the transport mixing drum to perform real-time carbon emission prediction processing of the transport mixing drum, generating real-time carbon emission data of the transport mixing drum.

9. The method for real-time correction of carbon emissions during the transportation of mixed materials at a mixing plant according to claim 1, characterized in that, Step S4 includes the following steps: Step S41: Perform vehicle operation status sequence analysis on the transport vehicle operation data to generate vehicle operation status sequence data; Step S42: Perform cumulative energy consumption feature analysis on the operation status sequence based on the vehicle operation status sequence data to generate cumulative energy consumption feature data of the operation status sequence; Step S43: Perform energy consumption cumulative feature correction processing on the cumulative energy consumption feature data of the operation status sequence based on the parameters of the mixture at the transport mixing plant and the transport road data to generate cumulative energy consumption feature data of the operation condition correction; Step S44: Perform real-time carbon emission analysis processing on the transport vehicle operation based on the cumulative energy consumption feature data of the operation condition correction to generate real-time carbon emission data of the transport vehicle.

10. The method for real-time correction of carbon emissions during the transportation of mixed materials at a mixing plant according to claim 9, characterized in that, Step S43 includes the following steps: Analyzing the impact characteristics of vehicle transportation based on the parameters of the mixing plant and road data to generate vehicle transportation impact characteristic data; Analyzing the driving behavior characteristics of each vehicle transportation condition using the vehicle transportation impact characteristic data and vehicle operating state sequence data to generate vehicle transportation condition driving behavior characteristic data; Correcting the impact of condition differences on the cumulative energy consumption characteristic data of the operating state sequence based on the vehicle transportation condition driving behavior characteristic data and the corresponding vehicle transportation impact characteristic data to generate corrected cumulative energy consumption characteristic data for the operating conditions.

Images