Engineering vehicle carbon emission monitoring system and method

Abstract

The invention relates to the technical field of engineering vehicle carbon emission monitoring, and discloses an engineering vehicle carbon emission monitoring system and method.The engineering vehicle carbon emission monitoring method comprises the steps that a multi-layer data sensing and collecting system is constructed; carrying out preprocessing and standardization on the obtained multi-source data; extracting environmental geographic features from the preprocessed data; establishing a unified mathematical framework of a vehicle working condition, a geographical environment and a meteorological condition; analyzing the three-dimensional coupling model, identifying a key coupling relationship and dynamically updating model parameters; generating a carbon emission calculation result by using the updated model parameters, and providing real-time monitoring, trend prediction and decision support; the adaptability and accuracy of carbon emission monitoring are improved, and a scientific basis is provided for engineering vehicle carbon emission management and optimization decision making.

Description

Technical Field

[0001] This invention relates to the field of carbon emission monitoring technology for engineering vehicles, and more specifically, to a carbon emission monitoring system and method for engineering vehicles. Background Technology

[0002] With global emphasis on carbon peaking and carbon neutrality goals, the monitoring and management of carbon emissions from engineering vehicles, as significant mobile sources, has become a key technological requirement for environmental protection and green construction. Engineering vehicles are widely used in infrastructure construction, mining, and urban construction, operating in complex and variable environments involving diverse geographical and meteorological conditions. Traditional carbon emission monitoring methods, often based on fixed emission coefficients or single operating parameters, struggle to comprehensively reflect the actual impact of different environmental and geographical factors on carbon emissions, resulting in limited adaptability and accuracy of monitoring results.

[0003] In practical applications, the carbon emissions of engineering vehicles are influenced by a combination of multiple factors, including the vehicle's own operating conditions, the geographical environment (such as altitude, slope, and terrain complexity), and environmental parameters (such as temperature, humidity, air pressure, and wind speed). These factors exhibit complex interactions, and changes in a single factor often non-linearly affect carbon emission levels through coupling with other factors. For example, in high-altitude and low-temperature environments, engine combustion efficiency and emission characteristics change significantly, and traditional models struggle to accurately capture this multi-factor coupling effect.

[0004] Furthermore, with the development of sensor technology and data acquisition methods, engineering vehicles can acquire a large amount of multi-dimensional operational data in real time, providing a data foundation for carbon emission monitoring. However, how to effectively integrate and mine this multi-source data, establish a unified mathematical model, and dynamically reflect the combined impact of the environment, geography, and operating conditions has become a technical challenge for improving the accuracy and applicability of carbon emission monitoring.

[0005] Therefore, there is a need for a method for monitoring carbon emissions from engineering vehicles that can integrate environmental geographic information, dynamically adapt to changing working conditions, and have high-precision prediction capabilities, in order to meet the actual needs of green construction and carbon emission management. Summary of the Invention

[0006] This invention provides a carbon emission monitoring system and method for engineering vehicles, which solves the technical problems in related technologies such as the inability to accurately integrate environmental and geographical factors, the difficulty in dynamically adapting to changing working conditions, and the resulting insufficient monitoring accuracy and applicability.

[0007] This invention provides a method for monitoring carbon emissions from engineering vehicles, comprising the following steps:

[0008] Construct a multi-layered data sensing and acquisition system to obtain vehicle operating parameters, geographic parameters, and environmental parameters;

[0009] The acquired multi-source data is preprocessed and standardized in preparation for feature extraction;

[0010] Environmental geographic features are extracted from preprocessed data using principal component analysis and nonlinear manifold learning algorithms.

[0011] A three-dimensional coupled model is constructed based on the extracted features, and a unified mathematical framework for vehicle operating conditions, geographical environment and meteorological conditions is established.

[0012] Tensor decomposition and sparse learning algorithms are used to analyze the three-dimensional coupled model, identify key coupling relationships, and dynamically update model parameters;

[0013] The updated model parameters are used to generate carbon emission calculation results, and real-time monitoring, trend prediction and decision support are provided.

[0014] In a preferred embodiment, the steps of constructing the multi-layer data sensing and acquisition system include:

[0015] Collect vehicle operating condition parameter data, including engine speed, load, speed, acceleration, and fuel consumption rate;

[0016] Collect geographic parameter data, including altitude, slope, and terrain complexity;

[0017] Collect environmental parameter data, including temperature, humidity, air pressure, and wind speed;

[0018] The collected raw data is preprocessed, including outlier filtering, missing value handling, and data standardization.

[0019] In a preferred embodiment, the step of extracting environmental geographic features from the preprocessed data includes:

[0020] Principal component analysis was performed on environmental parameter data to extract key feature vectors characterizing the environmental state.

[0021] A nonlinear manifold learning algorithm is applied to geographic parameter data to construct terrain feature manifolds and extract low-dimensional representations.

[0022] Construct an environmental geographic feature space and map environmental feature vectors and geographic feature manifolds to a unified feature space;

[0023] Feature engineering is performed on vehicle operating data to extract key features that characterize the engine's operating status.

[0024] In a preferred embodiment, the three-dimensional coupled model represents the interactive relationships among operating parameters, geographical parameters, and environmental parameters using a tensor network, including:

[0025] A three-dimensional tensor structure is established to characterize the intensity of interaction effects under different parameter combinations;

[0026] The operating parameters, geographical parameters, and environmental parameters are respectively mapped to the three dimensions of the tensor;

[0027] The numerical value of a tensor element represents the coupling strength of the corresponding parameter combination.

[0028] In a preferred embodiment, the step of identifying key coupling relationships includes:

[0029] Dimensionality reduction analysis is performed on the coupling tensor to extract key interaction patterns;

[0030] Tensor decomposition is used to transform high-dimensional tensors into low-dimensional representations;

[0031] Using sparse learning methods, we can identify combinations of factors that have a significant impact.

[0032] Construct a coupling relationship diagram to visually demonstrate the interaction strength and influence direction between different factors.

[0033] In a preferred embodiment, the step of dynamically updating the model parameters includes:

[0034] Construct a parameter update algorithm based on gradient descent to adjust model parameters according to the deviation between real-time data and prediction results;

[0035] The design includes a multi-scale parameter update strategy that incorporates rapid response adjustments and long-term trend learning.

[0036] Establish parameter optimization libraries for specific vehicles and regions to improve prediction accuracy under similar conditions;

[0037] Implement a model self-validation mechanism based on performance metrics to continuously evaluate model performance and trigger parameter recalibration.

[0038] In a preferred embodiment, the parameter update algorithm includes:

[0039] Set error assessment indicators and calculate the deviation between the model's predicted values ​​and the measured carbon emissions;

[0040] The gradient descent method is used to update the model parameters based on error backpropagation;

[0041] An adaptive learning rate adjustment mechanism is introduced to dynamically adjust the parameter update step size according to the error change trend.

[0042] In a preferred embodiment, the steps of providing real-time monitoring, trend prediction, and decision support include:

[0043] Calculate real-time carbon emissions and provide accurate emissions monitoring data;

[0044] Predict carbon emission trends under different conditions to support emission management decisions;

[0045] Generate operating condition optimization suggestions to guide the operation of engineering vehicles in order to reduce carbon emissions;

[0046] Generate a carbon emission analysis report, including emission composition, influencing factor analysis, and optimization directions.

[0047] In a preferred embodiment, the method for constructing the geographic influence function in the three-dimensional coupling model includes:

[0048] Collect emission test data of engineering vehicles in different geographical environments;

[0049] Comparative analysis of emission differences under the same operating conditions and different geographical conditions was conducted to establish the relationship between geographical parameters and emission correction coefficients.

[0050] By using piecewise linear or polynomial fitting, a function reflecting the influence of geographical factors can be constructed.

[0051] In a preferred embodiment, a carbon emission monitoring system for engineering vehicles is used to perform a carbon emission monitoring method for engineering vehicles, comprising:

[0052] The multi-layer data sensing and acquisition module is used to acquire vehicle operating parameters, geographical parameters, and environmental parameters;

[0053] The feature extraction module is used to extract environmental geographic features from multi-source data using principal component analysis and nonlinear manifold learning algorithms;

[0054] The three-dimensional coupling model construction module is used to establish a unified mathematical framework for vehicle operating conditions, geographical environment and meteorological conditions. The coupling relationship identification module is used to identify key coupling relationships using tensor decomposition and sparse learning algorithms.

[0055] The parameter dynamic update module is used to dynamically update model parameters based on historical data and the current state;

[0056] The results generation and decision support module is used to generate carbon emission calculation results and provide decision support.

[0057] The beneficial effects of this invention are as follows:

[0058] This invention utilizes an environmental and geographically integrated method for monitoring carbon emissions from engineering vehicles, enabling precise monitoring and dynamic prediction of carbon emissions under diverse geographical environments and variable weather conditions. Compared to existing technologies, this invention effectively overcomes the limitations of traditional methods that treat environmental and geographical factors in isolation, significantly improving the adaptability and accuracy of carbon emission calculations.

[0059] The three-dimensional coupled model can uniformly describe the comprehensive impact of three types of factors—vehicle operating conditions, geographical environment, and meteorological conditions—on carbon emissions, and fully explore the interaction relationships between multi-source data. Through algorithms such as principal component analysis, nonlinear manifold learning, and tensor decomposition, it can identify key influencing factors and their coupling relationships, and achieve effective modeling of carbon emission change trends under complex operating conditions.

[0060] This method possesses the capability for dynamic updating of model parameters, continuously optimizing the model based on real-time and historical data, thereby improving the monitoring system's response speed to environmental changes and the reliability of predictions. Applicable to various engineering vehicle operation scenarios, it can provide a scientific basis for applications such as carbon emission management, emission optimization decision-making, and carbon trading.

[0061] In summary, this invention not only improves the accuracy and applicability of carbon emission monitoring for engineering vehicles, but also provides technical support for achieving intelligent and automated carbon emission management, demonstrating promising prospects for widespread application and significant social and environmental benefits. Attached Figure Description

[0062] Figure 1 This is a flowchart of a method for monitoring carbon emissions from engineering vehicles according to the present invention. Detailed Implementation

[0063] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, some features described in the examples may be combined in other examples.

[0064] At least one embodiment of the present invention discloses a method for monitoring carbon emissions from engineering vehicles, such as... Figure 1 As shown, it includes the following steps:

[0065] Step 1: Construct a multi-layer data sensing and acquisition system to obtain vehicle operating parameters, geographical parameters, and environmental parameters;

[0066] Specifically, the following steps are included:

[0067] Step 1. Collect vehicle operating condition parameter data, including engine speed, load, speed, acceleration, fuel consumption rate, etc., through the vehicle OBD interface or dedicated sensors;

[0068] The specific implementation method is as follows:

[0069] The on-board diagnostic interface (OBD-II) collects the operating parameters of the engine control unit (ECU) in real time, including engine speed (RPM), throttle position, intake air volume, fuel injection quantity, etc.

[0070] Install acceleration sensors to collect acceleration data in the three axes of the vehicle to determine the vehicle's motion state and road conditions;

[0071] Deploy fuel flow sensors to directly measure fuel consumption and improve the accuracy of fuel consumption rate calculation;

[0072] Design a data acquisition protocol to ensure that the sampling frequency of operating parameters matches the dynamic changes of the vehicle, typically 10 to 100 Hz.

[0073] In some implementations, a torque sensor can be installed to directly measure the torque value of the engine output shaft, thereby improving the completeness of operating parameters.

[0074] Step 1.2: Collect geographic parameter data, including altitude, slope, terrain complexity, etc., through high-precision GPS, tilt sensor, terrain database, etc.

[0075] The specific implementation method is as follows:

[0076] It integrates a differential global positioning system (DGPS) to achieve sub-meter level positioning accuracy and accurately obtain the vehicle's geographic coordinates and altitude;

[0077] Install electronic tilt sensors to measure the vehicle's pitch and roll angles, and calculate the slope and lateral tilt of the road surface.

[0078] Preload a high-precision terrain database and combine it with real-time location information to extract terrain features and complexity indicators for the current location;

[0079] Implement a geographic information caching mechanism to maintain the continuity of geographic parameters through inertial navigation systems and map matching technology when GPS signals are weak or interrupted.

[0080] In specific engineering environments, this system can also integrate lidar or ultrasonic sensors to obtain more accurate micro-topographic information and further improve the accuracy of geographic parameters.

[0081] Step 1.3: Collect environmental parameter data, including temperature, humidity, air pressure, wind speed, etc., through the vehicle-mounted environmental sensor group or by connecting to local weather station data.

[0082] The specific implementation method is as follows:

[0083] Temperature and humidity sensors, air pressure sensors, and wind speed sensors are installed on the exterior of the vehicle to collect environmental parameters around the vehicle;

[0084] Design sensor protection devices to ensure normal operation of sensors in harsh engineering environments;

[0085] Establish wireless data access channels with local weather stations to obtain a wider range of environmental parameters and weather forecast information;

[0086] A multi-source environmental data fusion algorithm is implemented, which performs weighted fusion of locally collected and remotely accessed environmental data based on sensor accuracy and data timeliness.

[0087] In special working environments, this system can also selectively integrate air quality sensors to monitor the concentration of pollutants in the environment as a supplement to environmental parameters.

[0088] Step 1.4 involves preprocessing the collected raw data, including outlier filtering, missing value handling, and data standardization, to ensure data quality.

[0089] The specific implementation method is as follows:

[0090] Moving median filter and Kalman filter are used to filter noise and outliers in sensor data;

[0091] Design a missing value interpolation algorithm based on historical data to handle data loss caused by temporary sensor failure or communication interruption;

[0092] An adaptive data standardization method is implemented, which selects an appropriate standardization strategy based on the distribution characteristics of different parameters to make various types of data comparable.

[0093] Establish a data quality assessment mechanism to score the quality of collected data in real time and select high-quality data for subsequent calculations.

[0094] The data preprocessing of this system can also include data compression, which reduces the burden of data transmission and storage by using lossless or low-loss data compression algorithms, thus adapting to the resource-constrained vehicle environment.

[0095] Step 2: Preprocess and standardize the acquired multi-source data to prepare for feature extraction;

[0096] Specifically, the following steps are included:

[0097] Step 2.1: Perform principal component analysis on the environmental parameter data to extract key environmental features, such as the comprehensive index of temperature and humidity, and the rate of change of air pressure.

[0098] The specific implementation method is as follows:

[0099] The collected multidimensional environmental parameter data is organized into a matrix form, with each row representing the environmental state at a point in time and each column representing an environmental parameter.

[0100] Standardize the data matrix to eliminate the influence of different units of measurement;

[0101] Calculate the covariance matrix and solve for its eigenvalues ​​and eigenvectors;

[0102] Based on the magnitude of the eigenvalues, the main eigenvectors are selected to form a projection matrix, which projects the original environmental data onto a low-dimensional feature space.

[0103] In some embodiments, nonlinear principal component analysis methods, such as kernel principal component analysis (KernelPCA), can also be used to better capture the nonlinear characteristics of environmental parameters.

[0104] Step 2.2: Apply a nonlinear manifold learning algorithm to the geographic parameter data to extract terrain features, including terrain complexity index, elevation change rate, etc.

[0105] The specific implementation method is as follows:

[0106] Construct a local proximity graph of geographic parameters and use the K-nearest neighbor method to determine the neighboring points of each geographic point;

[0107] The ISOMAP algorithm is applied to map high-dimensional geographic data to a low-dimensional feature space by maintaining the geodesic distance between data points.

[0108] For terrain regions with high curvature, the Local Linear Embedding (LLE) algorithm is used to achieve nonlinear dimensionality reduction by preserving the local linear structure.

[0109] Based on the dimensionality-reduced feature representation, key terrain features such as the terrain complexity index and elevation change rate are calculated.

[0110] When computational resources are sufficient, this system can also use the t-SNE (t-distributed random neighborhood embedding) algorithm, which can better show the global structure of the data while preserving the local structure and is suitable for feature extraction in complex terrain.

[0111] Step 2.3: Construct an environmental geographic feature space, mapping environmental features and geographic features to a unified feature space;

[0112] The specific implementation method is as follows:

[0113] Design a feature fusion architecture to connect the environmental feature vectors obtained from principal component analysis with the geographic feature vectors obtained from nonlinear manifold learning;

[0114] By applying normative correlation analysis (CCA), the correlation between environmental and geographical features is learned, and a unified feature representation is constructed.

[0115] By setting weighting coefficients, the influence of environmental and geographical features in a unified feature space can be balanced.

[0116] Furthermore, in some embodiments of this application, a deep autoencoder can be used to achieve nonlinear fusion of the feature space, further improving the feature representation capability.

[0117] Step 2.4: Perform feature engineering on the vehicle operating data to extract key features that can characterize the engine's operating status;

[0118] The specific implementation method is as follows:

[0119] Extract time-domain features from the raw operating data, such as average speed, load change rate, and acceleration distribution.

[0120] Fast Fourier Transform is applied to extract the frequency domain features of operating condition data and analyze engine vibration characteristics;

[0121] Construct a condition descriptor to comprehensively represent the vehicle's operating status, which serves as the input for carbon emission calculation.

[0122] For engineering vehicles that operate for extended periods, this system can also extract the temporal characteristics of the operating conditions. For example, by using recurrent neural networks or long short-term memory networks, it can capture the patterns of changes in operating parameters over time, thereby further improving the accuracy of carbon emission prediction.

[0123] Step 3: Extract environmental geographic features from the preprocessed data using principal component analysis and nonlinear manifold learning algorithms;

[0124] Specifically, the following steps are included:

[0125] Step 3.1: Construct a unified theoretical framework and establish a mathematical expression for carbon emission calculation;

[0126] The specific implementation method is as follows:

[0127] Define a unified theoretical framework expression for carbon emission calculation:

[0128] E(v,g,e,t)=B(v)·G(g)·A(e)·C(v,g,e);

[0129] Where E represents the output value of the carbon emission calculation function; V represents the vehicle operating condition parameter vector, including parameters such as engine speed, load, speed, and acceleration; g represents the geographical parameter vector, including parameters such as altitude, slope, and terrain complexity; e represents the environmental parameter vector, including parameters such as temperature, humidity, air pressure, and wind speed; t represents the time parameter, used to identify the calculation time; B(v) represents the basic emission function, which calculates the emissions considering only the operating conditions under standard conditions; G(g) represents the geographical influence function, which calculates the correction coefficient of geographical factors; A(e) represents the environmental adjustment function, which calculates the adjustment coefficient of environmental factors; and C(v, g, e) represents the coupling term function, which calculates the interaction coefficient of the three types of factors.

[0130] Define the specific expression for the coupling function:

[0131] C(v, g, e) = ∑ i，j，k T ijk ·v i ·g j ·e k ;

[0132] Where C(v, g, e) represents the coupling term function, which calculates the interaction coefficients of the three types of factors; T ijk Let v represent a tensor element used to characterize the interaction strength coefficients of the i-th operating condition parameter, the j-th geographical parameter, and the k-th environmental parameter; i This represents the i-th component in the operating condition parameter vector v, such as specific parameter values ​​like engine speed and load; g j This represents the j-th component in the geographic parameter vector g, such as the specific parameter value of altitude, slope, etc.; e k This represents the k-th component in the environmental parameter vector e, such as the specific parameter value of temperature, humidity, etc.; ∑ i，j，k This represents summing over all possible combinations of parameters (i, j, k).

[0133] Step 3.2: Construct the basic emission function and geographical impact function;

[0134] The specific implementation method is as follows:

[0135] Collect emission test data of engineering vehicles under standard conditions and establish a fuel consumption rate mapping relationship;

[0136] Calculate the carbon emissions per unit of fuel consumption based on the carbon content and combustion efficiency of the fuel type.

[0137] Emissions test data were collected under different geographical conditions, including different altitudes, slopes, and terrain complexities.

[0138] The geographic influence function G(g) is constructed by piecewise linear or polynomial fitting.

[0139] Step 3.3: Construct the environmental regulation function and coupling term function;

[0140] The specific implementation method is as follows:

[0141] Collect emission test data under different environmental conditions and analyze the relationship between environmental parameters and combustion efficiency;

[0142] An environmental regulation function A(e) was constructed using multivariate regression analysis.

[0143] Design a third-order tensor network structure, and define the tensor shape and connection topology;

[0144] Organize multi-source data into tensor form and calculate the intensity of interaction effects.

[0145] Step 3.4: Perform the calculation and optimization of the coupling terms;

[0146] The specific implementation method is as follows:

[0147] The parameters of the tensor network are initialized, and the parameters are optimized using a hierarchical training strategy.

[0148] The tensor element values ​​are updated by minimizing the prediction error using the stochastic gradient descent method.

[0149] Implement the forward computation process of tensor networks and process parameter mapping through multilinear interpolation;

[0150] Tensor decomposition techniques are used to reduce the number of parameters, thereby reducing computational complexity while maintaining modeling capabilities.

[0151] Step 4: Construct a three-dimensional coupled model based on the extracted features to establish a unified mathematical framework for vehicle operating conditions, geographical environment, and meteorological conditions;

[0152] Specifically, the following steps are included:

[0153] Step 4.1, couple the tensor T ijk Perform higher-order singular value decomposition to extract the main interaction patterns;

[0154] The specific implementation method is as follows:

[0155] Applying the Tucker decomposition algorithm, the third-order tensor T is decomposed. ijk It is decomposed into the product of the core tensor and the three factor matrices;

[0156] Specifically, it is expressed as follows:

[0157] T ijk ≈∑ p,q,r G pqr ·U ip ·V jq ·W kr ;

[0158] Among them, T ijk G represents the original third-order tensor, where i, j, and k correspond to the indices of the working condition, geographical, and environmental dimensions, respectively; pqr This represents the core tensor, where p, q, and r are the dimension indices after dimensionality reduction, and their dimensions are usually smaller than the original dimensions; U ip This represents the factor matrix for the working condition dimension, where i is the original working condition dimension index and p is the working condition dimension index after dimensionality reduction; V jq The factor matrix represents the geographic dimension, where j is the original geographic dimension index and q is the reduced geographic dimension index; W kr The factor matrix represents the environment dimension, where k is the original environment dimension index and r is the environment dimension index after dimensionality reduction.

[0159] The calculation process of Tucker decomposition is as follows:

[0160] The original third-order tensor T ijk Decomposed into a core tensor G pqr and three factor matrices U ip V jq W kr The product of; where the core tensor G pqr The dimension (p,q,r) is usually much smaller than the dimension (i,j,k) of the original tensor, thus achieving dimensionality reduction and compression of the data;

[0161] The three factor matrices capture the main patterns in the working condition dimension, geographical dimension, and environmental dimension, respectively.

[0162] This decomposition method allows for the extraction of the main structure and patterns from the original tensor, significantly reducing the number of parameters that need to be stored and computed, while maintaining good representation of the original data.

[0163] This decomposition process is similar to a generalization of principal component analysis to multidimensional tensors, and can effectively identify key features and interaction patterns in multidimensional data.

[0164] By setting a singularity threshold, the main interaction patterns are retained while patterns with less impact are ignored;

[0165] Analyze the structural characteristics of the core tensor and identify the main interaction patterns among the three types of factors.

[0166] In some embodiments, other tensor decomposition methods, such as CANDECOMP / PARAFAC(CP) decomposition, may also be used, and the most suitable decomposition algorithm shall be selected according to the specific data characteristics.

[0167] Step 4.2: Apply sparse learning algorithms to identify combinations of factors with significant influence, thereby reducing model complexity;

[0168] The specific implementation method is as follows:

[0169] Introduce an L1 regularization term to impose sparsity constraints on tensor elements;

[0170] Solving sparse optimization problems using the alternating direction multiplier method (ADMM);

[0171] Based on the sparsified tensor structure, identify combinations of factors with significant influence.

[0172] This system can also employ group sparsity constraints to ensure that related factor combinations have similar sparsity patterns, further improving the interpretability of the model.

[0173] Step 4.3: Construct a coupling relationship diagram to visually demonstrate the interaction strength and influence direction between different factors;

[0174] The specific implementation method is as follows:

[0175] Environmental geography constructs a three-dimensional coupling relationship diagram based on tensor decomposition results, where:

[0176] Nodes represent different factors such as operating parameters, geographical parameters, and environmental parameters;

[0177] The thickness of the edges indicates the strength of the interaction between factors;

[0178] The color of the edge indicates the direction of influence (red for positive correlation, blue for negative correlation);

[0179] The size of a node indicates the importance of that factor.

[0180] Environmental geography employs a force-directed graph layout algorithm to optimize node positions and reduce edge intersections.

[0181] The elastic force of the edge is simulated using a spring model;

[0182] Automatic node placement is achieved through iterative calculations;

[0183] Keep nodes with high relevance close to each other.

[0184] Environmental geography enables interactive visualization:

[0185] Supports dragging and scaling of nodes;

[0186] Clicking on a node will display detailed parameter information;

[0187] Hovering over the edge will display the specific interaction coefficients;

[0188] Provides slice views in different dimensions.

[0189] Environmental geography uses coupling relationship diagrams for quantitative analysis:

[0190] Calculate the degree centrality of nodes and identify key influencing factors;

[0191] Analyze the weight distribution of the edges to identify the main interaction patterns;

[0192] Extract community structure and summarize factor groups that influence similar patterns.

[0193] Step 4.4: Identify the critical coupling paths to provide guidance for subsequent parameter optimization;

[0194] Furthermore, according to some embodiments of this application, causal discovery algorithms, such as the PC algorithm based on conditional independence tests, can also be applied to explore the causal relationships among the three types of factors, further deepening the understanding of the coupling mechanism.

[0195] Step 5: Tensor decomposition and sparse learning algorithms are used to analyze the three-dimensional coupled model, identify key coupling relationships, and dynamically update the model parameters.

[0196] Specifically, the following steps are included:

[0197] Step 5.1: Construct a parameter update algorithm to adjust model parameters based on the deviation between real-time data and prediction results;

[0198] The specific implementation method is as follows:

[0199] Set error assessment indicators and calculate the deviation between the model's predicted values ​​and the measured carbon emissions;

[0200] The error evaluation function is defined as follows:

[0201]

[0202] Where Error(θ) represents the error evaluation value under model parameter θ; N sample Indicates the total number of samples; This represents the predicted carbon emissions for the i-th sample. θ represents the actual measured carbon emissions of the i-th sample; θ represents the set of parameters of the model.

[0203] The calculation process of the error evaluation function is as follows:

[0204] Get N sample Predicted values ​​for individual carbon emission samples and actual measured value

[0205] Calculate the prediction error for each sample, which is the difference between the predicted value and the actual value;

[0206] Squaring each error value eliminates the effect of positive and negative errors canceling each other out.

[0207] Sum the squared errors of all samples and divide by the total number of samples N. sample The mean square error value is obtained.

[0208] The smaller the error value, the more accurate the model prediction, making it an important indicator for evaluating model performance and guiding parameter optimization. The physical meaning of this function is to measure the average deviation between model predictions and actual observations, with units equal to the square of carbon emissions.

[0209] The gradient descent method is used to update the model parameters based on error backpropagation;

[0210] The parameter update formula is:

[0211]

[0212] Where, θ new Indicates the updated model parameter values; θ old This represents the model parameter values ​​before the update; η represents the learning rate, which controls the step size for each parameter update. The error function with respect to parameter θ old The gradient is the rate of change of the error function at the current parameter point.

[0213] The calculation process for parameter updates is as follows:

[0214] Calculate the current parameter θ old gradient of the error function This gradient represents the rate of change of the error with respect to each parameter, and indicates the direction in which the error increases the most rapidly;

[0215] Choose an appropriate learning rate η to control the step size of parameter updates;

[0216] The product of the learning rate and the gradient is used as the parameter adjustment amount, and the parameters are updated in the opposite direction of the gradient (the direction of error reduction).

[0217] Obtain the new parameter value θ new .

[0218] Through multiple iterations, the parameter values ​​gradually converge to the position that minimizes the error function, thus optimizing the model performance. This gradient-based optimization method can efficiently find the optimal solution for complex functions and is a widely used parameter optimization technique in machine learning.

[0219] An adaptive learning rate adjustment mechanism is introduced to dynamically adjust the parameter update step size according to the error change trend.

[0220] In some embodiments, Bayesian optimization methods can also be used to achieve a balance between exploration and utilization by constructing a probability distribution model of the parameters, thereby accelerating the parameter optimization process.

[0221] Step 5.2: Design parameter update strategies for different time scales, including rapid response adjustment and long-term trend learning;

[0222] The specific implementation method is as follows:

[0223] A dual-timescale update architecture is implemented, with a fast response layer to handle short-term environmental changes and a long-term learning layer to capture seasonal trends.

[0224] For rapid response adjustments, a sliding window method is used to process the most recent N data points to achieve rapid parameter adaptation;

[0225] For learning long-term trends, an incremental learning method is used to gradually integrate historical data and update the model's basic parameters.

[0226] This system can also employ a reinforcement learning framework to model the parameter update problem as a Markov decision process, and optimize the update policy through dynamic programming or policy gradient methods.

[0227] Step 5.3: Environmental geography establishes a parameter optimization library for specific vehicles and regions to improve prediction accuracy under similar conditions.

[0228] Environmental geography constructs a hierarchical parameter optimization library structure:

[0229] Vehicle layer: Records the optimal parameter configurations for engineering vehicles of different models, power, and service life;

[0230] Regional layer: Stores environmental characteristic parameters of different geographical regions (plains, mountains, plateaus, etc.);

[0231] Working condition layer: Stores parameter optimization schemes for various working conditions (excavation, transportation, loading, etc.).

[0232] Environmental geography implements parameter retrieval and matching mechanisms:

[0233] Based on similarity calculation, quickly locate the best matching historical parameter configuration;

[0234] A weighted fusion method is used to integrate parameter information from multiple similar cases;

[0235] Dynamically update and optimize parameter values ​​in the library to maintain data timeliness.

[0236] Step 5.4: Implement a self-validation mechanism for the environmental geography model to continuously evaluate model performance.

[0237] Environmental geography establishes a multi-dimensional performance evaluation system:

[0238] Prediction accuracy: Calculate the mean relative error (MAPE) and root mean square error (RMSE);

[0239] Response speed: The time it takes for an assessment model to adapt to environmental changes;

[0240] Stability: Monitoring the degree of parameter drift and prediction variance.

[0241] Environmental geography design adaptive threshold triggering mechanism:

[0242] Define the warning threshold and alarm threshold for performance metrics;

[0243] A warning message is issued when the performance is below the warning threshold for N consecutive cycles.

[0244] Automatic parameter recalibration is triggered when performance indicators fall below the alarm threshold;

[0245] The recalibration process includes data resampling, feature re-extraction, and global parameter optimization.

[0246] Step 6: Use the updated model parameters to generate carbon emission calculation results and provide real-time monitoring, trend prediction and decision support;

[0247] Specifically, the following steps are included:

[0248] Step 6.1: Calculate real-time carbon emissions using environmental geography.

[0249] Environmental geography calculates emission rates in real time based on current operating parameters and environmental conditions;

[0250] Environmental geography cumulatively calculates total emissions at different time scales (hours, days, months);

[0251] Environmental geography analysis reveals the spatiotemporal distribution characteristics and variation patterns of emissions.

[0252] Step 6.2, Environmental Geographic Prediction of Carbon Emission Trends;

[0253] Short-term environmental geography forecasts: Predicting emission changes over the next 4 to 8 hours based on current conditions;

[0254] Medium-term environmental geography forecast: Predicting emission trends from 1 to 7 days by combining weather forecasts and other information;

[0255] Long-term environmental geography forecasts: monthly and quarterly emission levels are predicted considering seasonal variations.

[0256] Step 6.3, Environmental Geography generates optimization suggestions for working conditions;

[0257] Environmental geography identifies high-carbon emission operating conditions and provides optimization solutions;

[0258] The optimal engine parameter settings and operating modes are recommended based on environmental and geographical factors.

[0259] Environmental geography provides route planning suggestions based on environmental conditions.

[0260] Step 6.4, Environmental Geography Formation Carbon Emission Analysis Report;

[0261] Environmental Geographic Emissions Composition Analysis:

[0262] Emissions contribution at different operating stages;

[0263] Additional emissions due to environmental factors;

[0264] Statistics on abnormal emission incidents.

[0265] Analysis of environmental and geographical influencing factors:

[0266] Sensitivity analysis of key parameters;

[0267] Quantifying the degree of impact of environmental conditions;

[0268] Operating condition-emission correlation analysis.

[0269] Recommendations for environmental and geographical optimization:

[0270] Short-term operational optimization measures;

[0271] Recommendations for medium- and long-term equipment upgrades;

[0272] Carbon emission reduction potential assessment.

[0273] Step 6.5, Environmental Geographic Visualization Display Function;

[0274] Multi-dimensional visualization of environmental geography data:

[0275] Emissions over time curve;

[0276] Geographical distribution heat map;

[0277] Radar chart of influencing factors.

[0278] Interactive exploration of environmental geography data:

[0279] Supports time window selection and scaling;

[0280] Provides multi-level data drill-down functionality;

[0281] Implement custom chart configurations.

[0282] The embodiments of the present invention have been described above. However, the embodiments are not limited to the specific implementation methods described above. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art can make more equivalent embodiments under the guidance of the present embodiments, and all of them are within the protection scope of the present embodiments.

Claims

1. A method for monitoring carbon emissions from engineering vehicles, characterized in that, Includes the following steps: Construct a multi-layered data sensing and acquisition system to obtain vehicle operating parameters, geographic parameters, and environmental parameters; The acquired multi-source data is preprocessed and standardized in preparation for feature extraction; Environmental geographic features are extracted from preprocessed data using principal component analysis and nonlinear manifold learning algorithms. A three-dimensional coupled model is constructed based on the extracted features, and a unified mathematical framework for vehicle operating conditions, geographical environment and meteorological conditions is established. Tensor decomposition and sparse learning algorithms are used to analyze the three-dimensional coupled model, identify key coupling relationships, and dynamically update model parameters; The updated model parameters are used to generate carbon emission calculation results, and real-time monitoring, trend prediction and decision support are provided.

2. The method for monitoring carbon emissions from engineering vehicles according to claim 1, characterized in that, The steps for constructing the multi-layer data sensing and acquisition system include: Collect vehicle operating condition parameter data, including engine speed, load, speed, acceleration, and fuel consumption rate; Collect geographic parameter data, including altitude, slope, and terrain complexity; Collect environmental parameter data, including temperature, humidity, air pressure, and wind speed; The collected raw data is preprocessed, including outlier filtering, missing value handling, and data standardization.

3. The method for monitoring carbon emissions from engineering vehicles according to claim 1, characterized in that, The step of extracting environmental geographic features from the preprocessed data includes: Principal component analysis was performed on environmental parameter data to extract key feature vectors characterizing the environmental state. A nonlinear manifold learning algorithm is applied to geographic parameter data to construct terrain feature manifolds and extract low-dimensional representations. Construct an environmental geographic feature space and map environmental feature vectors and geographic feature manifolds to a unified feature space; Feature engineering is performed on vehicle operating data to extract key features that characterize the engine's operating status.

4. The method for monitoring carbon emissions from engineering vehicles according to claim 1, characterized in that, The three-dimensional coupled model uses tensor networks to represent the interactive relationships among operating parameters, geographical parameters, and environmental parameters, including: A three-dimensional tensor structure is established to characterize the intensity of interaction effects under different parameter combinations; The operating parameters, geographical parameters, and environmental parameters are respectively mapped to the three dimensions of the tensor; The numerical value of a tensor element represents the coupling strength of the corresponding parameter combination.

5. The method for monitoring carbon emissions from engineering vehicles according to claim 1, characterized in that, The steps to identify key coupling relationships include: Dimensionality reduction analysis is performed on the coupling tensor to extract key interaction patterns; Tensor decomposition is used to transform high-dimensional tensors into low-dimensional representations; Using sparse learning methods, we can identify combinations of factors that have a significant impact. Construct a coupling relationship diagram to visually demonstrate the interaction strength and influence direction between different factors.

6. The method for monitoring carbon emissions from engineering vehicles according to claim 1, characterized in that, The steps for dynamically updating model parameters include: Construct a parameter update algorithm based on gradient descent to adjust model parameters according to the deviation between real-time data and prediction results; The design includes a multi-scale parameter update strategy that incorporates rapid response adjustments and long-term trend learning. Establish parameter optimization libraries for specific vehicles and regions to improve prediction accuracy under similar conditions; Implement a model self-validation mechanism based on performance metrics to continuously evaluate model performance and trigger parameter recalibration.

7. A method for monitoring carbon emissions from engineering vehicles according to claim 6, characterized in that, Parameter update algorithms include: Set error assessment indicators and calculate the deviation between the model's predicted values ​​and the measured carbon emissions; The gradient descent method is used to update the model parameters based on error backpropagation; An adaptive learning rate adjustment mechanism is introduced to dynamically adjust the parameter update step size according to the error change trend.

8. A method for monitoring carbon emissions from engineering vehicles according to claim 1, characterized in that, The steps to provide real-time monitoring, trend forecasting, and decision support include: Calculate real-time carbon emissions and provide accurate emissions monitoring data; Predict carbon emission trends under different conditions to support emission management decisions; Generate operating condition optimization suggestions to guide the operation of engineering vehicles in order to reduce carbon emissions; Generate a carbon emission analysis report, including emission composition, influencing factor analysis, and optimization directions.

9. A method for monitoring carbon emissions from engineering vehicles according to claim 1, characterized in that, Methods for constructing geographic influence functions in 3D coupled models include: Collect emission test data of engineering vehicles in different geographical environments; Comparative analysis of emission differences under the same operating conditions and different geographical conditions was conducted to establish the relationship between geographical parameters and emission correction coefficients. By using piecewise linear or polynomial fitting, a function reflecting the influence of geographical factors can be constructed.

10. A carbon emission monitoring system for engineering vehicles, used to execute the carbon emission monitoring method for engineering vehicles according to any one of claims 1-9, characterized in that, include: The multi-layer data sensing and acquisition module is used to acquire vehicle operating parameters, geographical parameters, and environmental parameters; The feature extraction module is used to extract environmental geographic features from multi-source data using principal component analysis and nonlinear manifold learning algorithms; The three-dimensional coupling model construction module is used to establish a unified mathematical framework for vehicle operating conditions, geographical environment and meteorological conditions. The coupling relationship identification module is used to identify key coupling relationships using tensor decomposition and sparse learning algorithms. The parameter dynamic update module is used to dynamically update model parameters based on historical data and the current state; The results generation and decision support module is used to generate carbon emission calculation results and provide decision support.

Images